Question 14: Please describe your Linux desktop/client deployment experiences or offer any other suggestions (optional)
Deployment of community distros still requires scraping up all the components that don't come with the distro including codecs and other media support.
I have beta tested and used for several years first mandrake, then fedora, per Novell suse, and now my first choice for regular users is ubuntu and second choice is sled 10. I use ubuntu every day for the work i used to do on windows xp and only launch my windows partition if there is not other way to do the work using open source OR a colleague is a MS office user and needs my openoffice files with identical formating and appearance. The filter to transfer from MS Office to OpenOffice and back is show stopper.
KDE
Very smooth. I had no previous Linux experience, successfully installed the OS and enjoy using it.
no problems with Xandros
Painless. Xandros just works.
Drivers are a lot easier than with Windows.
I'm perfectly happy with Linux desktop.
There is a lot of inconsistency in quality and usability between different OSS applications that poses confusion to users. Some applications are very good, others feel like they still need more testing and polish. Some applications are well integrated with the desktop environment, but others like Mozilla, OpenOffice, etc. seem to have their own style of doing things.
The recent releases of SLED 10 and Ubuntu, have brought the Linux Desktop up to a competitive level with other operating systems for end users. However, some barriers still exist, such as multimedia support, and program integration. Installation of new software is a fair sized obstacle as well. The FOSS community has all this great software out there but you need to hunt through hundreds of packages to install them. A unified installer would help a lot for the end user. Whether it is an apt-get style system or something else, installation of software needs to be a 1 or 2 click process for the average PC user.
Our biggest challenge is with OpenOffice and acceptance with reading MS Office formats
Until there is Microsoft Office, it's not even a consideration. Every time OSDL and the like go on about how Open Office is "close enough", it just makes us laugh at how out of touch they are.
deployment was always smooth in our environment.
The single largest barrier is the lack of a robust, works-as-well-as-outlook MS Exchange client.
Unfriendly until Mandrake/Mandriva, but it was too KDE-centric. Ubuntu's ease of installation and management was key.
my distro is very slow! on normaly hardware (1.5 GHz/256MB RAM IBM laptop)
:(
LTSP is awesome. It would be even better if we could run Audacity on a thin client.
Did not detect, and automatically adjust to a change in hardware, but the Windows XP system on the same computer did.
current problems with memory over runs cause by the Xorg Xwindows operations. It loves my GB of RAM and 2GB of swap until it stalls the computer. Using Mepis 6 kept current weekly with all updates on an AMD Sempron 2800 processors. Xorg has been taking bad lessons from "Microsoft"
Deployment has been nearly pain free- now looking to replace Win4Lin 9X/Win98 virtual machine with a WinXP capable solution (for a single application).
Other than networking hardware issues (non-OS related), Linux has just worked.
My experience is good, although there remains much to be done: fonts, device driver availability, sound issues. Development tools are usually NOT an issue, it is more an issue in home use ...
Linux is used for a couple of servers, all over the place for security, but for a desktop... almost not at all. We're a media shop primarily (though I'm a programmer) and none of the tools are sophisticated enough at this point to make Linux a priority, not to mention the limited support for Active Directory and VPN.
There is a lack of good quality providers/SoftwareHouses willing to develop for Linux.
I use Ubuntu at the desktop, use Evolution with evolution-exchange and Firefox, most things work but the Windows apps that I need to run on Vmware cause wine can't do it. Please help Wine.
Kernel dependencies and device support are critical barriers.
most users had no issues, until we had to use sharepoint or outlook web access.
Generally interoperability with entrenched products such as Exchange, Project, and Visio are the key inhibitors. KDE makes a good replacement desktop for Windows users, GNOME lacks polish and features. Cross platform IDEs such as Eclipse make this far easier for engineering staff.
generally great. excellent packages such as inkscape and gimp. wish openoffice was better at handling word documents. wish fonts and displays were clearer / less fuzzy.
At my current job, approx 10 employees, everyone has a Linux workstation, running GNOME and KDE desktops.
Even the people who had never used Linux before have really no problems using GNOME/KDE and all the various apps, OOo, Firefox, Thunderbird, Evolution, Gnumeric etc. In fact they tend to prefer it over Windows.
Very bad support of WiFi driver support from Broadcom, Atheros and Ralink.
Developers use it.
We ship product with it.
Quickbooks has been our only sticking point.
Issues with usability of Wine has slowed adoption, inability to run common Windows apps (e.g. Quicken) that many employees need for perosnal productivity in the consulting business.
Linux just works, is a fabulous development environment for Java desktop applications and the best part is knowledgable vendors do not separate the desktop from the server -- it's all in there the way it should be!
A couple of key Windows-only applications prevent widespread deployment.
very smooth, surprisingly trouble-free
Printing is still a major disaster. Using KDE to configure CUPS to a printer where the vendor supplies open source drivers, I still have trouble. OpenOffice.org works great. Adobe Acrobat doesn't - I have to print to kprinter. Firefox 1.5 didn't print right, 2.0 works better, but still not quite right. I can't imagine my end users dealing with the technical options and obstacles I face in printing.
I use SLED 10 on an IBM Thinkpad T42. This should be one of the better, more supported laptop configurations. The basic functionality is fine, but power management is still very sub-par.
6 yrs. of Desktop/Server/Thin Clinet experience..., have installed many business setups. One major obstacle is the lack of DVD-film playback...
With the release of Novell SLED it has taken away the apprehension to move to linux and the look is much more appealing to the enduser than previous versions and distros
One of the biggest obstacles in our environment is how hooked on .NET and Internet Explorer our programmers are. Microsoft makes like so easy for developers that they become super dependant on the .NET framework and the MSDN stuff that they become protective of it. The Mono Project needs to get more public exposure and needs to become a major focus if you want to see more Windows shops use Linux.
Need Adobe support (#1) for Adobe/Macromedia apps... need webcams, need better wireless (ndiswrapper is not a user option...) and need to fix the nvidia/ati driver issues... and printing - can't print from say canon printers - applies to others as well. Canon and Adobe are biggest issues for me.
it's a little buggy at times. Which hurts deployment. Meaning firefox might stall for no reason with a user. Or they might come back from lunch and it's frozen
Used *many* linux distros, run Solaris almost exclusively, run windows on rare occasion
The only problem is developer support for mainstream business applications.. Banking, Insurance, etc. Developer adoption is the only problem I have.
Users are considering switching to Linux for private use because of the Vista hardware requirements and EULA.
SLED 10 has been great on laptops and desktops.
I, personally, have been MS free at home for several years now. I have had very good luck with Most of the Debian clones and Debian proper. I would like to see continued work done on Samba and LinNeighborhood, making them even easer to use. Would sure like to see better printer support, and I fully realize that problem lies with the printer manufacturers. USB support has gotten pretty decent of late, had very good results with SuSE, Debian and Gentoo.
N/A
My organization does not want to deal with Linux, even for servers (and the Windows servers are not that reliable). Most do not understand Linux/Unix well, though the organization's original central server was SCO. Application (including server applications) integration, educational software, training, security, and end user support are all absolutely central concerns, though some in management might be moved by cost-of-ownership issues.
Focus on getting NVIDIA, ATI, etc. to open their specs and release non-binary drivers should be a high priority.
The only obstacle are drivers. Many devices doesn't has a working driver. Some examples: touch screens (only 5 controllers are supported by xorg), smart card readers (many usb models doesn't support linux), scanners (all parallel models, more or less), usb xDSL modems (in many cases you cant connect to internet but you aren't able to browse web or do anything else if you don't have working synch bin), etc.
We are an R&D startup (started 2001).
We used Linux wherever possible from
the beginning. Laptops for sales
and administration are
about the only exception.
It was a "no-brainer" to use Linux
as founders are all old-hands at UNIX.
We have found that as much as we want to get away from windows, there is always that one application or the one chunk of data that restricts our options. Migrating from Outlook can be painful if the users are saving data in local folders for example. There is presently no real substitute for Visio, Kivio is getting close but is marginally useful with the free stencils and building a large library at the prices they want is prohibitively expensive. Just a couple of examples.
We can't deploy Linux desktops to our staff because our core application, AutoCAD, and all it's add-on products require Windows.
Moderate terminal services success -from linux connecting to win2k3 terminal server.
Currently have ltsp running on suse, switching over to debian/ubuntu due to software development of a new project/support. One area that was mentioned previous was fingerprint readers, this would be great for our industry in order to identify clients for medication purposes.
InACCESS (Affordable Classroom Computers for Every Secondary Student) is a big success in Indiana. Schools are experiencing tremendous success using open-source software and Linux.
Suggestions:
Installations of new programs/apps must be clickable installs - must be easy.
There needs to be a recognition that millions are accustomed to MS Windows...as near as possible Linux should conform to that user experience. Changes can be introduced later.
Technical staff must have GENERIC training opportunities. RedHat, Suse/Novell, Ubuntu need to create a general course that they all offer. They are more alike than different.
Availability of certain applications is an issue (like certified financial software)
The main reason that users are keeping to Windows within my organisation is because there isn't a robust MS Exchange Client to rival Outlook. Evolution is good, but our experience is that it is very slow and crashes often - And this seems to be the only product that has full (i.e. calendar and mail) Exchange functionality.
Desktop deployment has been only an informal prototyping and testing by experienced developers. No usage by an average user has been attempted yet.
Linux clients are 90 percent ready for deployment, but remaining 10 percent imposes a significant hurdle. Lack of adoption by our peers raises questions from management, wireless support issues need resolution for laptops, Administrative tools need solidified for mass management (ZENworks for Linux helps). There are also concerns about OpenOffice compatibility. We can't risk a single incompatible document.
Using LTSP with thin clients on manufacturing floor. Internal web-based systems developed to handle production/inventory/maintenance. System is rock-solid, centrally managed, easily upgraded and secure. Systems is 3 years old and we rarley upgrade/touch it.
Still use Windows for 2 applications: Accounting - Peachtree and Engineering - SolidWorks.
We deployed one client when a temporary employee was hired.
It can be substantially harder to deploy things developed on Linux (e.g. iFolder) than on Windows. Client computer management is key to have (and wasn't listed)
Deploying the modern Linux desktop/clients is much easier than in the past.
Also considering support for Solaris and Apple desktops
Windows Media is a problem. DVD playing is still illegal. Little support or interest from small hardware vendors with Win-only drivers and apps. Basically, there are gaping holes in Lin coverage.
The desktop simply HAS to be easy and familiar for end users! Packaging, interfaces, graphics and marketing have always has been crucially important to the end user. No one who says he cares about the average user can give these things anything but his full attention.
Hardware-vendor support for new hardware (especially laptops) is really essential. Eg. try to setup your widescreen laptop to give a presentation through a projector with 4:3 800x600 resolution with clone-display. This is simply impossible for an average user with the available tools.
Challenges worthwhile the effort.
I/We have been using linux for the past three years in our organization. It looks great and keep up the good work
We've decided to standardise all our file and print sharing on Samba which works fine for the Windows XP/98 clients but for Linux clients we have to manually configure the smb/cifs shares through /etc/fstab. We've yet to find a reliable 'network neighbourhood' type smbbrowser; most of them read fine but writing to shares is erratic at best. This severely impacts on adoption of Linux on the desktop. Nobody is going to manually configure workstations to access shares. We see this as the major obstacle and where development should be focused.
Have tried various distro's but most people seem to be happy with SUSE for it's 'clean' desktop (all distros running KDE) and it's ease of manageability. Major problem was getting people to use software (Open Office, etc)that they had not grown up with.
After many attempts to build a supportable desktop environment, I'm always held back because of applications that only run on Windows. Open source alternatives are not acceptable to the end user or to management. Wine is not a supportable solution (when it works) and virtualization is ok for technical staff, but why would I want to deploy 2 OS's to a standard client machine to have to manage/patch, etc.? After about 3 years of trying to deploy a successful Linux desktop, it's looking more and more like I'm going to be stuck implementing Vista.
application packaging is the key - click-to-install support is an absolute necessity for the average user to accept Linux on the desktop, especially at home. We've gotten used to installing applications from an executable package on Windows & Mac - the same needs to be in place for Linux.
Needs to be faster and more stable. X hangs more than almost any other application.
We run a GNOME Linux (Fedora 5 with custom kernel) thin-client based system across multiple cities.
No real operational issues. Training issues have been neglible. In most cases, no special Linux training required at all.
Our biggest issue has been dealing with companies which either only support Windows or tie software to Windows unnecessarily. Examples:
LabCorp, Inc. : supporting only the Windows version of the citrix client when Linux version works fine.
Kroll Laboratory: Web site claims to only support Internet Exmplorer but works fine in Mozilla/Firefox when user agent changed to say IE.
Compass Bank: On-line banking only functional in Internet Explorer.
Medicare: Electronic remittance printing software written in .Net with windows-specific functions which serve no purpose to the function of the program. Not useable under Linux (mono or wine).
Dymo: no support for Linux although works fine under cups.
Ricoh: no linux support for scanning function in multi-function lasers.
Voice Dictation: no linux desktop options
Looking up. Less problems than in the past. We may try an actual Desktop Distro next (e.g. SLED or Xandros)
We use mixed 32 bit and 64 bit Debian sarge and etch distros.
Wish it were more up to date.
Video from DVD is madness, but we hack (lib-dvd-dcss)... sad.
need better VIDEO card drivers...
Wish for better tools for:
Analog electronics design
Speaker design tools
IO cards like serial devices else Internal Modems, soundcards
Wish we have better cross Linux Distro usability solutions
Synaptic, apt-get, RPM, Yast, tarballs, deb... whacky!
Just last night I got an AdobeReader to work in Debian 3.1
Not for the novice to do.
I am trying alien for Official StarOffice 8 install, Yikes!!
I am not a programer, but I am the only one who can build Linux desktops for our company.
Never give up, Linux will win.
GNU/Linux continues an effort, however, vastly improved hardware-software arrives EVERY six (6) months.
OPEN 'Global' development is the BEST as it trumps commercial vendors every time. A little known FACT!
All considered, Penguins are one FINE breed of bird!
We are predominantly introducing Linux at the application server level.
as of latest version of SUSE, great, earlier versions were a bugger.
Novell SuSE have been amazing. Both product, and company. And frankly are the only company making a product ready for the business desktop with full support.
Need standard hardware configuration/troubleshooting/performance analysis tools and stable drivers for at least most popular hardware (e.g. ATI video cards, Wireless adapters).
LTSP is the way to go!!!!!!!
no suggestions
We have a very positive experience with the LTSP-project (4.2)
testing one desktop (Xandros), and using one as a file server (Suse). Must be able, to load new applications to multiple distributions - easily, like Windows.
Have had problems running MYOB Accounting or finding a Linux alternative suitable for Australia.
We are please to be using Linux on the desktop for quite some time now.
Four yours of evaluating various emerging deskop Linux distros. Nothing breakthrough ever happened--burned me out, so I bought three Apple Macintosh computer.
I often find the need to compile my own kernel and tweak each and every individual system. Why can't Linux detect the hardware support needed and make a Kconfig file on the fly and custom compile a kernel...that would be cool. Also, a hibernate feature as full-featured as Windows XP's would be very useful....currently, all Linux-based solutions are not as capable as the Windows XP version (but it is getting there...and that's good to know.).
Has gone very well but NVidia's proprietary driver has held us back from upgrading the Linux kernel, an upgrade needed for full i965 and i975 compatibility
After using Debian I migrated to Ubuntu which works like a charm and is recommended by me to all of my friends.
standarize on one tool, say smart for example
It takes significant effort from staff, we are luckily a technical group and used the migration as a way to build linux skill in all the technical staff.
Apps should run under any window manager if I have the libraries (GTK2/QT/etc). We run Windowmaker, and don't want to have to install KDE or Gnome to run apps. The base libraries should be enough.
VPN usb keycard support non-existant, Exchange integration using evolution has been problematic
The transition from Windows was actually a bit easier than expected with little to no disruption.
early trial
Ability to use current non-Linux apps and end-user education/evangelism seem to me to be most important.
We tried an LTSP deployment and thought that it was working good, but troubles came when specific applications were needed. For general purposes, LTSP is very good, but now we have windows-only applications and web-based java applications that need sound output. Sound is a very complicated issue in LTSP and Windows virtualization is not feasible for a terminal setup nor with low-end machines as we have.
I find Linux far better than Windows 9X, but XP is
still ahead concerning R.A.D. programming Tools
Authentication across windows domain controllers, NIS servers and local systems is a problem.
Availability of commercial security applications, remote access, and authentication solutions (smart cards) is a large issue. Lack of good quality, comparable tools to Microsoft Visio and Project are also an issue.
Linux is easy to deploy, secure, and extremely stable. Windows servers have an uptime of days or weeks in our org. Linux servers have an uptime of months and years.
We're a small company that has always run Linux on the server, so it wasn't that hard once we found a distribution that would work on our existing desktop hardware without much tweaking. Since we do web design, we still need to keep Windows machines around to test with Internet Explorer, but the rest of our operations are carried out on Linux.
We were pleasantly surprised by the quality of even the most basic tools--the basic text editor provided with the desktop environment basically fills the roles we used to have to buy Visual Studio and VisualPerl to do. We also liked how easy it was to talk to our existing Linux servers with protocols like NFS.
main obstacles to migration:
1. ignorant and FUD victms management.
2. user resistant to learn new things.
In the face of the Microsoft/Novell agreement, protecting the GPL should be the first priority of the entire community!
Administer my own workstation at the university +
3 Ubuntu machines at home.
Excellent operating system, no problems since two years we starrted using it
I've run Linux on my machines for +8 years now, I came because Windows was no longer interesting but stayed because I fell in love with the freedom, stability and security given to me by free software. I've learned so much from the free software community and I struggle to pay them back for the good memories and kindness. As it not for the heaps of talented and friendly but direct people I've met here I would never have understood the need for things like open standards, open specifications and open code. I think that's the finest most precious gift anyone has ever given me.
Went very well. No major issues.
Disappointed by Novell going to the dark side....
Was all OSS while we were poor. When lots of cash appeared, management decided to switch to MS.
The most difficult part is to find same or equivalent applications or find web application because application are to often lock with Windows.
The second difficulty is printing driver that are very poor and not very friendly to use as they are in windows.
Only shopfloor computers run Linux since they can get by with limited functionality. We use Qt to develop apps which run on Linux, Windows, MacOSx :) We plan to switch to Linux clients, department by department.
NEEDS
-Must have directory services, like eDirectory.
-Need centralized printer management, auto install printers on clients or easy user install.
-Need central user login management. Must not require someone to manually create user accounts on each client machine.
-Clients must have file access Netware, Linux file servers and maintain proper user access rights. (Login scripts per OS) NFS cannot maintain Netware user rights very well.
Debian is too hard to install/configure for our normal users. They can handle "SuSE or better". The name "Ubuntu" can't be used in this country (it's hard to sell a name that makes people laugh).
Server deployment is easy. Desktop deployment suffers from lack of hardware drivers. Better then windows as a platform for open-source based computing tools.
We are all very happy for switching to Linux - the workflow is now better.
Hard to get applications to work reliably. Video Driver issues. Networking to non unix/linux machines quirky.
Linux is all there is!
We only use Linux Desktops and can recommend them to anybody. We use OpenOffice.org as our Office package. After getting used to it, we found that it was more effective them MS Office.
Generally good. Bad experiences from application vendors refusing to support system neutral applications.
Desktop (home) a lot of people have kids that want to use a lot of features of IM services that aren't yet available on any of the *nix ports, or they might be implemented but are buggy. For instance kids want the webcam and voice features to work, they want to be able to send winks and things like that, if they don't have it then they feel like their friends who use MS products have a better product. This is not an issue to us linux lovers, but to the people that aren't technical users, they think its not as good because their friends all use MS things that arent yet compatible on *nix, making them just want to switch back.. running MS products in virtualization shouldn't be an issue if linux ports were designed to be compatible with their counterparts.
The physical part was easy even to data migration. Staff training has been a pure pain. Windows users seem have zero knowledge of where their data is stored or what an application really does. They do expect the system to do everything for them regardless of the security factors. Linux works and works very well but the the process is totally alien to the general Windows user.
not yet viable due to lack of industry specific application support.
Wireless is a royal pain and what works in one distro or on one version suddenly doesn't work. Some Corp Apps need IE to work - Active X, etc. Lotus Notes sucks on Linux, esp. doesn't work with KDE, still should be considered Beta software. Need some recommendations on File Systems since Reiser is going away (ReiserFS 4 ?).. XFS, Ext3? SAN recommendations for large amounts of data? Need LDAP admin client for Linux !!, Need Antivirus Client(s) that work with Corporate Solutions esp. if they have kernel hooking modules - need to sync with updates from vendors i.e. Rug from Novell or YUM or whatever it is now. Network tools that allow you to switch from Wireless to Wired connections (and work consistently). Would like to be able to easily play mp3's, DVD's and CD's in a Corporate environment where appropriate. Need tools that run on Windows that have Linux versions - lessens the learning curve.
We deploy Kanotix LiveCDs as a single image with no hard disk requirement.
"Trim Clients" - non-volatile storage is provided by fileservers and usb flash drives.
It runs on generic PC hardware (mostly A64/2GB ram), loads into a ramdisk (to avoid LiveCD speed issues), and the hardware is supplied with a 5yr nbdos warranty.
In the case of hardware failure, pull the cd and usb stick out, and drop in a replacement machine.
Mostly positive; messaging (Evolution) sucks, wireless is so-so, vpn is hit and miss
I have looked at many distros and have found Suse (Open Suse) to be consistently the best for my application set and needs. 3D graphic support would be nice. In an educational environment, while the deployment of linux clients is no real drama, using LTSP, the real issue I faced was user training and user acceptance, particularly among users who were fairly illiterate but used wincrash at home. Because linux was perceived to be so different they did not want to use it. ( for fear of appearing illiterate! but kids are amazingly perceptive!!) Fortunately the kids had no problem with it and basically took over the computers and were quite happy using them.
Have been using Linux exclusively for 3 years now. Linux has come a long way in non-nerd usability but has a way to go. Hardware and PDA issues need major attention.
Better printer and scanner support. Keep GNU/Linux core drivers and utilities free software/open source. Not use blobs like nvidias driver, develop own.
remarkably smooth and trouble-free
High featured thin clients could become a key factor.
Virtualization can be so interesting too.
desktop linux applications needs more robustness and confiability.
Obstacle is X window configuration changes. Our firm changes monitors as prices fluctuate. There are many combinations deployed. A wizard to quickly setup/change X config would be a great help.
Mixed. I like the idea, but have always had to go back to Windows... Either a device vendor doesn't have an working linux driver, or a needed software package isn't available. Dual-boot is a silly joke, it's fine for computer-geeks, but I can't multitask thru a dual-boot cycle. It's wonderful that linux can read thru NTFS, but the write barrier is insurmountable. As long as Microsoft has the stranglehold on the desktop, Linux will be just like MAC-OS, better but irrelevant.
Sorry folks, I wrote software for 17 years, only did one linux project and I had to pull teeth to get the customer to see the advantage. Now I onw the company, and I don't have the time or inclination to "play" with the computer anymore. I own tens of thousands of $$$ of software on each on my PCs and it just has to work. TCO includes each day of downtime, and despite the possible long term benefits, Linux goes far into the red in the first few days; so much so that I just can't afford to keep playing. As soon as there is a "guaranteed" glichless crossover, we'll try again...
In case it's not obvious, it's been a while since I've tried a linux deployment, maybe a year or more ( I think it was Suse 10.0 )
Thanks for reading.
ldap central auth for both linux and windows machines requires lots of config.
gnome which appears to be the default in most distros is too simple and different for average users, kde is more configurable and in some distros it is a second class citizen.
For some printers and cameras, the support is lacking, incomplete or difficult to configure.
It is not possible to change all the desktpos because of apps that only runs on windows.
Major obstacle is getting our independent software/hardware vendors (telephone PBX, CRM, business software, etc) to comprehend that we really want nothing to do with Windows. For example, our telephone switches run Linux but the voicemail and directory are stuck using proprietary Microsoft protocols running on a Windows 2003 server with no way but Outlook plugins to talk to it. This sort of madness has to be overcome somehow. Worse, the vendors won't tell us (or maybe they don't even know) how the protocols work so we can write our own software.
As a consultant, I use Linux desktop software in my personal business at home. In the client offices, I more typically use Linux either as a server or as a remote desktop system through VNC or a similar service. I also frequently use remote VPN connectivity between home and office sites.
It's very hard to keep up with new demands, so there's a big necessity of the organization to ensure there will be a cross-desktop policy to avoid Windows users from getting ahead.
Loaded it a few months ago but have had a very difficult time getting it to work properly. Open office under windows is OK but the Linux 10 I am trying to use is a nightmare. I am a user not a computer master.
Thanks to open-source model, have been able to build a completly automated and robust installation system of my own. Cut maintenance time to almost nothing.
I use it for all machines in my business.I dual boot 1 machine fr the few things that need Win, but that is minimal.
I'm impressed of how easy it is to install, maintain and update linux desktops.
And no more viruses and system register errors and bluescreens.
We use allot of old DOS applications and theres no good enough emulator.
We are very small. Our major wish is for easy wireless capability, otherwise we love Linux.
It's still a problem. I had problems getting my wireless working (6 months to fix). Sound it still not realiable, etc etc. Windows just worked out of the box :-(.
very good
Todays Linux Distros Like ubuntu are easily deployed with significant licensing and cost reductions. A good deal of our consulting work with SMB's is based on these easy deployments.
I have had problems with some distros on some hardware. Until pre-installed Linux is commonplace this is going to be an issue for many people, and a reason not to risk migration. Currently SUSE 10.1 is performing well for us, apart from an Audacity bug.
The default appearance of KDE and Kontact/Kmail on SUSE 10.1 was ugly, and wasteful of space, to the extent that I considered using Thunderbird instead of Kmail. Both KDE and Kontact/Kmail can be configured to be neat and functional, but why not get it right by default? Kmail is a great email client, it should not present like a kid's toy.
Open office is not user-friendly.
SUSE Linux needs to get past all the tweaking the customer has to do in order to play videos, listen to rtsp, get wireless connection, etc.
We are a 100% Microsoft-free company. Do I need to say more? The biggest problem we see is that some vendors still don't provide open source drivers. This hits us quite hard in several cases (nVidia graphics cards, Highpoint RAID controllers, etc.).
good and bad, printing is always a bitch and someone always wants to keep windows.
All our machines (about 500) either Linux or dual-boot with Windows. In-house support and expertise, easy to handle this many machines with a single distro.
printing problems are the #1 issue - time to figure out and install print and scanning drivers are the #1 headache. Ubuntu (my distro of choice) has been seamless for most other requirements of my office as we use primarily internet/intranet based applications for everything else.
Too much of a build your own mentality. Not ready for typical end users, period. It still is a nerd only tool, a good one to be sure but not for humans.
Deployed Fedora to ~400 in-store pharmacy desktops without incident. Saved hundreds of thousands of dollars in equipment upgrades and software licensing costs. Driver: HIPAA compliance.
faultless
just make it so I don't have to spend more time writing scripts or calling my buddies for help than I do actually using the apps to finish my projects
SuSE should NEVER release a product whose installer is broken! I was very angry to find the installer broken with OpenSuse 10.1 - very bad look. Wine should NEVER produce releases which have LESS functionality than their previous versions - nobody should ever need to do a version downgrade, it's total nonsense! We were very happy with Parallels and VMWare products. Ubuntu should come with NetworkManager out of the box. Suse should be seamlessly upgradeable like Ubuntu.
A single software update/install GUI that is easy to configure and use is a must. Synaptic, rpm, yum, yast, kpackage et all is just too much.
Very Easy. Most of the concerns are FUD, we wun an equivelant number of linux workstations as pc ones. The linux install base is easy, and simpler yto maintain and easier to setup.
Mixed, depends on hardware
All is well as is, so far.
The Linux desktop has come a long way, there a quite a few things that need some work though. Better integration with AD and/or Samba and/or SSHFS based authentication would be nice. Also, there needs to be a way to run wireshark (formerly ethereal) as a non-root user.
audio, flash, dvd, support needs to be stronger out of the box... what about conversion of formats on the fly to open formats?
Need to convince more commercial vendors to write native Linux apps (i.e scientific and engineering software etc.)
Beryl shows a lot of promise, should take advantage of such technologies (Xgl graphics etc.) for games, useful apps and accessibility
Also, need to have virtualization such as xen and openvz more built-in so that developers can utilize it easier and develop more secure apps for common users.
-Prohibitive for non technical users
-Generally painless with experience
A smaller company that isn't oriented to computer technology would not be able to deploy Linux effectively because of the absence of coherent documentation.
Fedora is fairly bleeding edge, so all of the issues with deployment are expected. Most of our issues revolve around applications such as OOo, not the desktop experience as an abstraction.
consistency is key. stop using k in the name for kdesktop applications, it just sounds stupid.
We have succesfully been using Linux for over 5 years and up to now the only problem has been the unavailability of some drivers for old devices, especifically an optical card reader.
Strongly dislike the installing of non-packaged software from other venders. This task is very cumbersome and unacceptable if Linux wants to be a real player in the OS market. For now, we are staying with Microsoft Products because of the easy of installs and upgrades of products.
At first, It's better to be used as a development tool and a access tool to web applications.
rather painful; and for crying out loud, quit partisan wars between distros.
I would like to support better graphics, there exists molecular viewer graphics that consume a lot of CPU time to process and display molecules.
I have deployed linux with mixed results on desktops at work and at home. The main problem areas have been wireless networking support, hardware support (ease of installation), and multimedia compatibility (not being able to play mp3 music and most if not all video formats out of the box is a major problem. I also run linux on my laptop computer with satisfactory results, but hardware support is incomplete and there are teething problems with user interfaces for areas such as power management and wireless networking. All in all I can live with linux, but it's a bit of a chore. Windows and Mac are still the best choices for an efortless desktop experience. With linux you need to want it really bad or you'd quit on first boot. BTW I use SUSE Linux Enterprise Desktop (v.10)
Overall, pretty good. Whenever an issue or question arose, the community quickly provided an answer/solution.
I'm very disappointed by some hardware manufacturers which do not want to release Linux drivers or at least release specifications so that Linux developers can write quality drivers themselves. I think big companies supporting Linux should put some pressure on hardware companies that do not want to cooperate with Linux community.
We must standardize linux desktop environment.
Used Ubuntu since Oct05, very good experience. Using both 6.06.1(main) and 6.10(projects), good updating, both run well in my office. Freespire has some image handling ability that I like, weak on setting up printers/scanners; still in beta.
The only option we really have. Sun/SGI are dead meat.
They are getting better!!, but we want to have applications like Adobe Family Software, AutoCad, etc.
I'd like to see an accounting application that could both import & export data so that my Quick Books using Accountant can use what I send them. Ideally, it would be a MySQL database so that the associated management & backup tools could be utilized.
I am currently testing out Linux deployment off site, in a Linux / Windows environment. Still learning :)
Documentation is a major issue at times.
Overall very good experience. Major obstacle: lack of hardware support standards.
Poorly written documentation in web, print or telephone formats is a Linux hallmark. Language is "high geek" with an emphasis on the theoretical while ignoring the practical. Vocabulary eludes even advanced users and programmers. 3D and wireless drivers that can be installed simply are essential. The drivers must be packaged to privide all dependent functions and files. Please: no more late night treasure hunts for driver fragments. Lose "Make", "Configure" and "Install". Businesses don't have time for that. Those are the main issues. The rest of Linux is ready to be productive.
Very difficult to get some software tools such as Oracle Developer running on Linux
Fantastic. Ltsp and thin client is the only choice for cost conscious non profits
Major *COMPATIBILITY* problem area for linux application dstribution across different distros: Fonts! Fonts! Fonts!
PDF management tools - functional equivalents of Acrobat in terms of comments, security, etc.
Academic citation management tools such as Reference Manager, Z39.50 clients, etc.
We had video issues with intel chipsets, other than that basically printer support and p&p support
Problems in Microsoft Office files.
Marketplace has few qualified developers; majority practice unstructured design, coding, documentation.
User backlash when internet multimedia (MSNBC, CBSNews, FoxNews, youtube, vendors web sites) does not work properly is a huge problem. Vendor supplied windows-based (*.exe) product catalogs and training cds are a show stopper and have forced me to take linux off several desktops.
It has been a long transistion to Linux due to lack of crossplatform support for core applications but after recently purchasing a MacOS PC I've come to realize just how far Linux has come. Linux is not perfect, but it has sure come a long way in meeting the needs of users and not just companies who want to make profits rather than providing users with what they need. Well done!!! :-)
Finding alternatives to MS Office, and changing the entrenched perspective that Office is the only solution, is key to Linux desktop success in large corporations.
Needs to be consolidation in distros. Get more wood (unison/consistency) behind the desktop arrow.
Relative to the Mac desktop (and Windows) it is still quite primitive. The Help system for different parts of the desktop is also awful - which wouldn't be bad if the desktop were intuitive and easy to use, and had all the functionality required, but it isn't. For example, in GNOME - how do you blank a CD-RW? I always end-up using the command-line.
When companies develop websites, they user-test them to get feedback on useability. The same kind of thing should be done for the Linux desktop.
Our development software is ran in VMWare (until Parallels improves) for all of our MS dev tools which is most of our development. We are attempting to migrate to an open enviroment with Coldfusion, which will enable us to be free of the MS constraints. Wine is improving well, and we already are testing our MS dev tools now with some success. Perhaps we may soon be able to move to Linux in spite of our MS dev tools with Wine or even Crossover Office.
Personal Desktop requires faster graphics to play games.
In an elementary school with carts and laptops, I have dual boot systems with Windows, and Linux. 1st grade through 5th grade have adapted well to SuSE linux. They prefer to use SuSE rather than Xp. KDE has better graphics. SuSE has some really good educational software right out of the box. In an Apple G4 lab, we replaced Mac OS 9.3 with Yellow Dog Linux. It worked beyound expectation on a power pc platform. At least for a short while. Although it ran well, there was no flash plugins for the browsers. The school pays for some service that requires flash and shockwave. We had to buy licenses and upgrade to Mac OS X.3. On the x86 platform, keep up the good work and keep them plugins comming.
Rod
The only real problem has been getting streaming video to work but has not been a serious issue.
we have deployed Linux desktops with a variety of distros to roughly 8% of our userbase aver the last 18 months. We are now ready to start deploying on a wider scale and have settled on Ubuntu as our primary distribution with some RedHat/Centos where software design compatibility is an issue.
Overall, relatively good. Upgrades and mass installs are still more complex than they should be.
Have tried numerous live CD desktops. A simple sturdy, efficient one like Puppy is most attractive. But I have to keep a windows system to maintain my large catalog of books written in WordPerfect!
I've been using Linux for 2 years now. I've completly replaced Windows with it. The only reason I still have windows is for 3d acceleration requirements for playing games, etc. I wine IE and emulate for critical apps like Photoshop and Dreamweaver (mind you, native linux versions would be great!).
It's all we have ever used. We're all skilled and experienced linux users and that was a job requirement. All users modify their distro.
Need better user interface and performance
No perfect replacement yet for Quickbooks
Some printers are only widows only
it used for word processing.........and audio
Latest version installs easily but still has trouble recognizing periperals
an easier way to choose default applications is needed. graphical tools for setting up X is needed. Better WIFI support is a must. Easier to install VMware server for virtualization is needed to boost the will to switch. Switching from FF to Opera as the main browser is also a top concern to gain security. Easier to install XGL/AIGLX is a must. And out-of-the-box support for mult-buttoned mice and keyboards.
I had a good experience with Linux desktop, but i consider that Linux desktop is not ready yet for the masses
Linux works very well.
Learn from Ubuntu. Easy to install, and great community support.
First of all, I would like to plug the Lazuli Mockups at http://desplesdadotcom.nfshost.com/?p=75. They show an intuitive interface that Linux needs to be on track with XP, Vista, and OS X.
Second, having used Linux on several IT workstations, I can honestly say that Linux is not ready for large desktop deployments. The operating system is very stable and secure, but it lacks quality drivers for mainstream hardware and client management tools like those available to Windows users.
Finally, software installation and dependency fulfillment need serious work. Tools like apt and yum are great for installing open source software, but there seems to be no uniform way to install 3rd party programs. Until I can click on "setup.exe" and go "next, next, finish", I can't truthfully say that Linux is ready for wide desktop deployment.
since most m/c are preinstalled with some OS, the firts important tool/deterrent is presence/absence of partition resize tool. pls mandate it in install CDs.
Novell SLED - Simply Great.
Very smooth, with some custom scripting required to frontend CLI applications.
Several models of laptops and desktops and servers.
Good
Fantastic! I am looking forward to move more clients onto Linux
Standard interfaces across all distro would go a long way to improve userhip and userbility
So far Wine has not come close enough in being able to run a particular VB app that one of my clients relies on. With it supported, they have already said they are otherwise ready to remove Windows from all of their machines and replace them with Linux/Wine based operating systems. This would also allow greater control over access to the internet, system setting, and overall has one the one test machine lead to better uptime on their slightly aging hardware.
Multimedia support in browsers leave a lot to be desired. Windows pop up but don't play video file.
Needs support for Video Cards
Test deployments have been problematic - especially with HP and other printers. Ease of setting up SAMBA has also been a problem.
Poor support for audio/video drivers; little support for fixing bugs in previous releases of kernel and app's; projects with hundreds, even
thousands, of open bugs years old; each upgrade seems to break something that worked before; progress seems to negate stability of workstations. Users are looking for progress with
increasing functionality and stability; we're not seeing that today. Considering going the mac route. need reliability, stability and function.
Great - that's what I'm using now.
Integation with Windows AD environment is major problem. Complex and highly skilled people required. Need more low skill IT friendly tools to easily merge Linux into Windows AD environment.
I deployed linux desktops on some offices, where work is generally based on documents and internet applications.
Great, except that the video acceleration for my laptop's intel card is horrible. I've got it working, but it's a pain to get it working if I have to reboot.
One thing is that Intel has given the source code for their video cards to the public, so I think that the drivers should be really good.
Ubuntu Linux changed everything, and it is a very interesting case study. Prior to Ubuntu there was some adoption, mostly amongst the engineers. When others started to see the elegance of Ubuntu, specifically package management, upgrades across versions, hardware support and general smoothness across the board, even those non technical folks began to give it a try. The free CDs just laying in the office became the catalyst for them to jump in. No longer do I have to push Linux, it just seems to speak for itself.
I was a avid fan of Linux but gave up after trying hard to get the device drivers working but in vain.
In my opinion Linux has been ready for the Desktop for at least two years. Linux does, however, currently require at least someone in the company to supply support/knowledgs/tweaking/training or IT outsourcing. As usage increases and understanding of OSS/FOSS culture grows, though, a growing number of PC users will get by without additional support.
User training is quite a biggie, getting people to realise things do work but just look a little different
Linux desktops are only deployed to IT workers whom have elected to handle this themselves. However we do have a full 25 computer lab of sun workstations running RHEL 3.
Easy setup, but disc encryption don't work. Wireless setup/client should be betterand have the possibility to switch between profiles.
its easier now compared to years before. just need to show FOSS alternative to what they're using now
IMHO the best desktop Unix is Mac OSX
Lately, we have found that the installation process is even simpler than Windows. But we have hit obstacles with a few device drivers such as onboard LAN, Wireless, PDA synchronisation etc.
Get me out of command line hell, I can never remember which shell commands work with which distro depending on what it is based on and I can hardly ever remember with out some serious goole searches - how to setup wireless, printing, particular packages. I definately have had great experiences with the autopackage system that is used by Xara LX, it has worked on Suse, Mandrake, Ubuntu and a few others. easy and simple. Thats the message - Keep it simple, and simplify it for those of us who never really want to recompile our kernel but just want to use blender or inkscape and do our work.
Distros are fairly easy to install these days, the browser support for Flash etc. should be better, better free multimedia support would be nice. Stability is very good and desktop features are good too.
lack of new drivers and documantation
Biggest obstacle is software. We need InDesign, Photoshop, Dreamweaver and some others Win-only programs. None of those run natively in Linux (Wine is not a solution for us).
For now I am the only one running a Linux desktop, but my colleagues are highly interested. There are no company-wide plans of deploying, though. Main problem: our sine-qua-non application (SDL Trados) is Windows only (.Net-based). I am currently running Windows XP under VMware in order to use it and some lower importance programs.
great
i'm a gentoo user. nuff said :)
About 100 clients already deployed to meet demand of technical users. However we are now in a pilot to create managebale (sutomatic) linux desktop environment in order support much larger user base.
Major proprietary license issue is the cost of MANAGING bizarre and complex licence schemes.
User resistance mostly is a result of brain wahing in government funded "computer literacy = MS Office familiarity" training schemes. These people tend to be resistant to other good practice methods. and thinking in general. Refusing to give them a MS environment is a good encouragement for them to go.
We have some government mandated software that requires MS Windows. We use rdesktop to access this on a single remote virtual instance of WinXP per application. There is a similar solution for Shipper supplied label printing/requisition software. This allows outside organisations to install their software on a virtual machine without giving them access to our systems.
Get ahead don't catch up - put some effort into Project Looking Glass before January product launches...
Three years in a small business environment.
With proper due diligence, migration is not terribly hard.
Tried Mandrake (Mandriva) free issue but lack of support put me off. Have received Ubuntu and have yet to try it. Obviously painless and preferably AUTOMATIC (!)upgrades would make life easier... Openoffice is a brilliant credit to Open Source and a wonder. It should be much more heavily marketed. Firefox ditto. I believe the future is with Linux.
awesome
Our users have been quite happy, even curious, but we did underestimate the amount of end-user training needed.
Many companies have special software running only on Windows and these don't run perfectly with e.g. Wine.
the focus on pre-installed systems from oems like dell, lenovo is not important, but feature complete open source device drivers for hardware componants used by such vendors i there systems is absolutly critical
Can't get management workstation fully supported with every tool i use now on windows. And the wide installbase of windows applications in the organisation that doesnt have linux support
Clients run well nice graphics are in corporate environment not necessary, so no focus is needed there. Development tools are excellent. More focus should be put on changing the mindset that it's not bad to have commercial software. Get rid of the zealots!!!!
The Ubuntu partition manager destroyed my win2000 partition (NTFS) when I wanted to move it to create a linux boot partition.
Linux desktop only for IT staff, the other people are dependent on come Windows only 3-rd party software
-Pleasant graphics improved the transition experience for the end user.
-Closed source plugins for browsers were an overlooked necessity.
We are a small Open source firm, so Linux and *BSD Unices are in everyday use by 80% of the workers.
The biggest problems with Linux desktop adoption have been related to the "Windows way" of doing things. Some minor (fixable) problems have risen from technical reasons.
Company infrastructure is based on Microsoft technologies. Moving toward Linux is very slowly (hard). Main problems (stop factors) are Active Directory, MS Office, Exchange and interoperability issues with them in Linux.
Concentrate on KDE for new users.
For new sys admins, concentrate on "webmin".
For developers concentrate on "desktop standards". Create them and publish them in a similar way to the internet RFC's.
While desktops such as Apple's OS X show major innovation the Linux desktops seem rather clones of existing desktop features, with some "nerd" extensions. Linux clearly requires Desktop UI innovation.
satisified with the results, though the progress seems to be slowing down during the last 2 years
We have seen problems with user restrictions. We do want our personell to try oout new software. This often requires root access which we do not want to grant them. It is also the same issue when handling graphis resolution changing, especially on the move.
The biggest problem we have in our organization is to have Microsoft Active Directory and MS Exchange integration. This is a necesisty for us due to corporate requirements. The Evloution solution is crap in KDE. Korganizer is taking excgange support away and we have had limited success with NS Outlook on Wine.
AD integration has improved tremendosly lately.
OpenOffice is the key for success of Linux in the office. For the labs most software packages are now available for Linux too.
Linux needs to have more applications and games in order to be major player on the desktop
Make Linux desktop more responsive by reducing bloat.Other than few application like OpenOffice.org and gaim I do not see any other kill application on Linux.
There is a need for good development tool.Do not tell me you can use Eclipse.Qt desinger is good but it is still not there.
I am not looking for Visul Studio clone on Linux.
Ubunutu is great, but when it comes to management I'm not satisfied...
Very good
High priority should be given on improving the interoperability between linux-based mail clients (Thunderbird, Kontact) and Exchange server (specially its calendaring features).
I used to use linux desktop for my day to day job on succesively RedHat, Ubuntu and Gentoo. Gentoo is the most difficult to deploy but the most configurable one, no extra packages installed. I'm not using Linux anymore due to a MS Exchange compatibility issue.
No good default config, needs much time to install all needed stuff (driver, programs...)
Using Suse/OpenSuse as primary Desktop
Linux deployment is not very important for a company where there is a healthy and working Windows environment (Active Directory, OS Imaging [Norton Ghost], established procedures).
Deployment is limited to a small fraction of users not connected to the main corporate network. Used for web access and custom development and modeling.
Currentley we deploy our desktops by booting from cd/usb with the distro on the net (ftp, http, nfs) and configuring with AutoYast or KickStart
We're thinking about PXE deployment
Suggestion: some form of chain-boot system to make deployment distro agnostic, i.e.
1) boot from floppy to load a network-boot image
2) boot from network image to load a PXE image
3) boot from PXE image to load an iso-boot image
4) boot from (network connected) iso-image to load the distro
The Govt. If India the central Govt is the most major obstacle to deploying Linux, Check out their websites , Especially the Ministry of Company Affairs They do not allow any other browser even to search the site, Apart from that there is no major obstacles, The state governments are doing well and most of their sites are vendor independent,
multiseat systems
Desktop Linux is quite ready for prime time. Availability of third party applications is hindering the adoption. Pre-installed linux desktop/PC needed and marketing policy to increase awareness. Availabilty of good games to increase adoption of linux as it is a big industry.
- actual distros (especially SUSE) requires substantial computer resources (CPU, RAM, HD). Get a lean 'basic installation' and allow to add what is required (for installations on older H/W)
- Several required apps are not available on Linux. Especially in the Graphics Business (prof. Vector-Graphics Programs like CorelDraw / Adobe Illustrator), special-apps for sign-making business (Gerber Scientific OMEGA)
- WLAN support is presently still poor. A "remote" desktop could not be linked to the WLAN due to poor hardware support (Several components (PCI-card, USB-stick) tested. There might be a way to get it running, but not easy enough for a newbee.
- In order to make a substantial impact (against MS Windows), all distros have to work together to get more standardized.
e.g. get a COMMON next-generation package managent system which replaces .deb- and .rpm-packages.
There must a be more COMMON base for all the distros. So more resources could be used more widely.
Too much work is done multiple times. Therefore much LINUX energy is wasted in this duplication work.
Using Ubuntu distro since mars 2005 and migrate all our laptops, desktops and servers in may 2005.
Before that, numbers of test with Mandrake, Red Hat, Suse between 2000 and 2005, but none responded to our needs as Ubuntu did in 2005.
Subsist one windows machine for accounting, and two windows test machine where replaced in october 2006 by VM Player.
Compatibility issues... might consider compiling more with static libraries for this reason?
Significant problem with Microsoft Planning Server. The online timesheet system requires ActiveX. Incompatibilities between Clearcase and VMWare.
The Tech-Staff uses stand alone Workstations.
The Students use diskless remoteboot Clients. Updating is an issue here, but we are working on better solutions.
Virtualized several Linux distributions on Windows - works fine, generally easy to install and use. I plan to switch to virtualised Windows on Linux platform.
To hard to install new external applikations - need grafic interface and one click command
A good downloadsite with the most popular linux apps and all large software producers- not only OS projects softwares
Startup of Linux-OS is generally too slow
Keep up the good work!
THE MAJOR OBSTACLE in deploying the GNU/Linux environment as a replacement for Windows is the lack of support from the top company management. They are used to luxury and overlooked mindlessness and that's why they absolutely do not understand the whole picture. They think implementing GNU/Linux is just a matter of installation and they do not see the need for further education of those who should use it. Then, when the employees complaint, the management says "OK, return to Windows". This is absolutely the worst factor influencing the adoption of GNU/Linux in enterprises.
more of "it just works" would be nice. Many things are possible, but require quite some manual work and experience.
Working in a scientific research environment - already many major tools ported. Little need for windows legacy support, so perhaps not typical. However main issues are still proprietary hardware support.
It is difficult to provide an automated installation
mechanism which gives the user with diverting hardware a simple way to install. It would be important to update and install the machines with a common set of tools
everything went fine there where some issues with hardware but they got resolved by googling for hot-to's
better integration of windows application in to linux and proper mime type mapping
Due to earlier desicions we are bound to MS Products (our Application, telephone/outlook integration, some applications run on win only)
perfectly good distro install takes an eternity, but for the rest of the pc's its just a multicast issue... so if you have some brains its time consuming but not difficult at all
Only at trial stage for one desktop. Wine MUST be the top priority for companies where, like ours, use a Windows server accounting package that requires Windows clients to run. I cannot stress this point enough.
in general all efforts have been stopped by lack of support fro some important application, feature or standard - the latest being proper wireless management and support for wpa & wpa2
It's also a mayor issue that developers do not realise how important these wireless issues are - it's simply make or brake for a linux deplyment.
Good but some applications are far as quality from competitors or are missed (ie Ms access)
Hard to convince hierarchy
very good, but need some polishing. Only big problem is woth some native windows apps
Technical staff sympathetic but fear they lack the knowledge, skills to transfer to Linux. Also, hardware leased from Research Machines who are still tied to MS-Windows.
It great! But we sorely miss integration with our handheld devices, other than bluetooth. Support for smart phones(such as Nokia, Sony Ericsson with Symbian)and palm leaves a lot to be desired(probably windows mobile too but we won't touch those devices with a firepoker).
Focus on making the Linux desktop the best you can, not on making it like another OS to make migration easier.
We successfuly deployed Ubuntu after training the entire team to get used to Linux softwares on Windows, then when felt ready we started deplying Linux progressively.
Easy integration in microsoft AD
The linux desktop is wonderfull
If a company is motivated enough, there are no major drawbacks in migrating, and a lot of value to gain. We have developed a thin client solution that almost solved all the desktop deployment issues. We can manage security and stability with virtually no dedicated personnel.
Linux Desktop is the greatest, stable as H^ll no viruses everything works just out of the box.
Have had challenges but with the right skill available overcome that. A secure and manageable environment.
The Linux deployment experience was good, but we think there is a need for long term support (security/bugfix patches).
1) Too much hassles in connecting and updating peripherals support/drivers and configuring them.
2) too many application versions - but not enough comparative info about them to help decision making
laptops seems to be the future for our students. And with windows preinstalled, windows on laptops will dominate. Some students might prefer mac's or even linux laptops.
I've deployed 5 linux desktop to the secretary (mail/office suit) and they are happy with it (it's faster/easier).
I'm using a linux workstaion to manage the network, and developing web application.
Laptop power management still needs to be improoved. Let notebooks sleep when they are not working...
Further work in non English spell checking is needed.
Primarily ease of use, reasonable speed (OpenOffice.org is slow) and stability to entice users to switch to Linux readily.
Switching to Fedora for Web development (php mysql apache) has been 100% success.
Shockwave development is so far held back by Adobe Director MX suffering a small but significant bug when running under WINE. Windows still used here.
However, shockwave plugin runs fine under windows firefox, running under WINE.
3D desktop /eye candy is important for attracting new users. I was first attracted to Linux by SuSE 9.1 which was better looking that XP.
We are a company of 40,000 consultants - there are many consultants who contribute to OSS and the company sponsors OSS software.
However, the approach to desktop is that we have a "supported" build which is windows xp and ms office on IBM/Lenovo or HP equipment (choices for cost and capabiltity). About 200 consultants use Linux partitions for their day to day work with about 1000 more who will touch a linux server as part of day to day development. The company takes a sympathatic approach to consultants using their own OS but does not offer support if it goes wrong. We have a company community that help each other out to configur Linux to work with the latest Microsoft service. the latest problemn we have is that they are rolling out an internal version of Microsoft Internet messanger (microsoft live communications server) that is not supported by gaim etc.
N.A.
Sometimes to often problems installing distris on current hardware (PCs and server)
none
Runnibf desktop environment in kiosk mode providing only few different user profiles (menues, desktop behavior, rights)
The device, manufacturer, reseller and related support for Linux should be increased! Microsofts' influence on hardware manufacturers is far to big.
use of FAI deployment tool
Only linited Linux deployment to date.
We heavily depend on Excel and that completely beats the open source competition for our power users
So Far, Linux on the desktop is not a viable alternative for Windows because of the following reasons:
1. Development on linux is much harder due to scattered developer documentation and the lack of a really good graphical development environment, like visual studio.
2. There are no alternatives to many mission critical software components like various departmental databases and billing and accounting software packages.
3. Lack of a Group Policy type of configuration utility to remotely lock down and manage Linux desktops.
4. Lack of a good set of graphical based management tools that will work with multiple versions of linux to make management easy.
5. Complex distribution specific libraries and dependencies. Linux needs something like the Windows registry to make it easy to install, develop, and manage software on an individual desktop.
Linux needs this:
1) Linux should have ONE COMMON, CENTRALIZED easy_to_use MIME (file association) registry. All applicatins should use the same MIME registry when opening a file. Each filetype can have several choises on what application to start & run. Firefox, various email clients, file browsers etc should USE THE VERY SAME MIME registry + it must have management GUI for easy maintenance.
2) One common fileformat for EMAIL database. All email-clients MUST ! use the same fileformat to save contact, emails, drafts and resources.
Why the hell are the filformats used by Thunderbird, Evolution etc so different !
3) Some Linux distrubutions ask during installation my name, address, email and phone-numbers. Whats the f* point? It is never used for anything useful at all because ALL email and contact applications have their own filformats and method. What a messs!
UNIFY , CENTRALIZE, UNIFY !
Nil todate
The have been obvious and refreshing improvements during the last two years to deploy out of the box and ready to go after an hour or so plus time for individualisations on per user basis.
Not many problems, but application support is still poor. Wine is a big help in some areas, but not enough. Desktop as itself runs fine on most machines I've seen, but until there are important applications available (like Photoshop, SQL editors, ...) not many will use it. Also more competent dev. tools should be available.
Peripheral support is number one priority, especially display/monitor support. Easier configuration.
We currenly use Kubuntu as a web development / programming environment and would like to see improvements in the tools available (& their integration). So far it's been pretty good, but a true dreamweaver replacement is still needed.
We are a small organisation and I am technically capable enough to do a rollout, however we have no need to do so, however much I would like to. My employer and I use Apple notebooks for portable working (I also have Ubuntu Linux at home and on my own notebook). Our developers use Windows desktops primarily down to my employer's choice, plus we need to test websites in Internet Explorer.
We are a web hosting and E-Commerce company, we host only on Linux. I see no reason why we couldn't use Linux on the desktop also, but we have little desktop hardware turnover and a software investment in certain applications (Photoshop and labelling software).
- Windows culture is a problem
- People not engaged with the Linux project is a major problem
- Linux/Windows interoperability is a major concern
Not enough management support, due to some systems (PowerBuilder/VB6) used for internal applications that are not available under Linux.
one word, multimedia !
Excelent in what respects to stability and bad in what respects to training.
Companies often don't move desktops to Linux because there is a lack of trainers or there isn't training in afordable values (in Portugal).
Missing key Applications (read: Adobe Photoshop and "Final Cut Pro"-class video editor) is holding deployment back for us. People are very interested in using Linux, but it's "easier" just to use Windows, because of available applications on that platform. Linux should be treatet like a 1. class citizen by ISV's, and not an afterthought (Flash 9 still in beta, no Quicktime support, etc.)
Hopefully by this time next year, +80% of our IT is Linux.
Better handling of removable USB devices. ie, when user request safe removal of a USB device when there is a program still using the device, instead of telling the user the device is in use, pop up a dialog box showing the user which applications are accessing the device and give the option to switch to listed applications or to close the applications (with appropriate warnings about information loss)
Using SuSE and Ubuntu worked well, better peripherals support would be nice
I have had no technical difficulties deploying Linux.
I have had political issues deploying Linux. Some people are intensely committed to to using other systems.
Lack of knowledge about Linux and the fear resulting from it have been the largest obstacles to Linux deployment. Linux needs a major marketing/PR campaign.
lack of active x for proprietary browser apps is a major obstacle
very good
I am a full-time student and I casually use Linux on my laptop at home and I also work with it in a server environment at work. The majority of my daily work duties have me dealing with the average desktop PC user. In my personal use of desktop Linux and my professional interaction with desktop users, there are a few areas of desktop Linux I have come to find are weak.
The first weak area of desktop Linux in my opinion is package management. I feel that there are too many options, as well as lack of cross-platform options. Average PC users want to be able to double click something and have it install without question. Package managers such as apt-get and yum are somewhat good at that; however, they require command line interaction, and they are not cross-platform options (such as you can't run apt-get on Fedora). Common desktop PC users don't know anything about the command line and will get turned off of a distribution if you have to interact with one.
The next is weak area I feel desktop Linux has is multimedia support. I believe there are too few Linux distributions that support common multimedia such as MP3 and DVD right out of the box. Sure, these distributions can be customized to allow for these formats to be supported, but this goes back to the fact that common PC users donﾒt know how to interact with a command line. The average desktop PC user wants something to ﾓjust workﾔ without having to do complicated configuration.
Finally, I feel that Linux as a whole needs to work on marketing. There are two things that feed into marketing and those are advertising and information dissemination. If you ask they average desktop PC user, they pretty much know of 2 different operating systems: Windows and Mac OS. That is because there is no advertising effort on the Linux front, either in the server or the desktop market. And if someone does learn of Linux and they want to be able to find out more information on it, there is no central repository that can disseminate information to those people. Because of this, users will start to feel lost when they come upon such large amounts of information that they donﾒt know what to do with.
With these observed weakness, I suggest the following. First, develop a package management system that requires little user interaction that is capable of installing different types of packages (.deb, .rpm, etc.), all of the dependencies required, as well as build the package from source if needed. Next, find some way to get open source drivers for MP3 and DVD support. Iﾒm not sure if they can be developed or if an agreement has to be reached with the license holders, but it would help desktop Linux greatly if they were available. And finally, get money together to start advertising Linux to common PC users and create some kind of central information database or a central repository that points to other places that contain the data.
This is a secret state, not public.
Dual-monitor support needs to be better auto-configed, or better documented. Resistance by people still running Windows who see no need to migrate even though they've been hit by viruses.
Applications seems to start slower on Linux then on Windows. So first user impression is "why linux is so slow?".
A lot of aplications while not function correctly, do not show feed-back to the user (MessageBox etc.). They just write to stdout/stderr, so technical staff must always resolve even simpe problems.
The window for killing application in KDE is not higest priority window like in Windows, and is too sophisticated for avarage users than windows one.
Other problems:
- five Printer selections dialogs on Linux, several File Open Dialogs (it should be one selected by administrator, and all applications, desktops should use it)
- several mime/type connections to applications
- still some problems with drag and drop beetween applications
Directory server support is lacking, should be supported out of the box, but difficult to set up currently. We need something better than Active Directory, yet easier to set up and use.
package management is far too complex, there really needs to be one standard adopted by all major distros ...
Video card configuration needs better support and better drivers -- some are too unstable and too difficult to set up (i.e. requires a lot of tweaking)
i would like to see more emerging standards regarding interaction with the user, standards that will be used by all distros
Initial deployment on selected hardware presented few issues. Vendor-released patches are largest cause of support issues - incomplete patches or cause configuration changes which interfere with functionality. Inconsistent or incomplete documentation makes problem diagnosis difficult.
SMB support needs to be cleaner. Such as open office needs to be able to open documents from an smb url.
Need Flash, Java, WIN32 codecs, MP3 as standard equipment in every Distribution. Company wide use will not happen until a Linux can do everything that Windows can do.
hard to sell to users: change
Good !
It is still not possible to run w/o Windows applications when integrating with the corporate enterprise computing environment.
We currently have one staff member running openSUSE 10.0, unfortunatly with some difficulties with programs and especially printing. Also looking at the opportunities to re-purpose older equipment running Linux.
I'm an avid user of OpenOffice.org and recommend it exclusively to everyone. So this request should in no way be construed as a slur upon this fine office suite; this request is a plea for a supplemental or adjunct program (or extension) for a single purpose: a dedicated snail-mailing list program similar to the MS Windows compatible "MyMaillist".
CAMS and Freemailer were the only projects (on SourceForge) that I discovered which come close to my appeal for a snail-mail Linux-compatible program; and they both appear to be dormant projects, no downloadable files or activity since 2004.
The Franklin County Humane Society, for whom I'm a full-time volunteer, needs to do an occasional snail-mail funding-raising appeal, sending out letters and envelopes the 'ole-fashion' way. For the general mass-mailing, we can use a commercial service with a bulk-mail permit, but for the smaller, supplemental mailings (50-300 or so), it falls to me to use some sort of program to print the addresses on envelopes. Right now, I'm 'hobbling along' with Mail Merge in OfficeOffice.org (even the Animal Shelter Management program depends on exporting to MS Word or OpenOffice.org Writer).
But it would be far better if I could maintain the complete mailing list in a dedicated mail-list program comparable to MyMaillist (by Avanquest)(Windows format); and yes, I know it's sometimes advertised as compatible with Linux, but this is false.
This is precisely the kind of small-utility program that Microsoft Windows users miss when they switch to Linux, the ONLY reason I still even have MS Win XP installed on any computer!!
Thanks very much, Walter.
I appreciate that Linux distros configure themselves with little assistance from me. Good job.
Wireless is very important. The more automated the configuration, the better.
Interoperability with Windows/Novell Networks is an important issue around here. Clients to connect Linux to legacy Netware NOS, such as Netware 5.0, would be a big help.
Otherwise, we like Linux. I personally greatly appreciate its qualities. It's the only operating system I use at home. Of course, the kids still use Windows. They play more games than I do.
Hope this helps.
implement outstanding accessibility features in Gnome
Printing (KDE/CUPS) is not predictable - particularly WRT tray selection and image size (edges are frequently cut off);
Version of KDE does not support off-line viewing of messages;
Version of O/S does not fully support video card (i905) and does poor job of supporting our external monitor (1920x1200);
High resolution web-cam is not supported;
The single biggest pain point is dealing with Exchange. The Evolution Exchange connector needs a lot more work.
We have successfully rolled Linux on the desktop for all our technical staff, and are evaluating it for the rest of the enterprise. There are a few areas (finance and banking, mostly) where W32 support is required, and we have not had the opportunity to properly solve those issues.
Earlier attempts were difficult. Newer attempts have succeeded admirably. Peripheral support as gotten better.
SUSE 10.1 on Thinkpads, such as T40 or R50 takes about 30 minutes. WiFi took about 30 minutes the first time - experienced installers can configure in less than 1 hour.
Technical design complete but shelved due to deployment complexity.
We are very happy
I love linux, There is no reason to go back to windows. Easy install very reliable and easy for user to work with
none
binary only drivers have been a real issue. More assistance should be focussed to the efforts to reverse engineer and/or open source those drivers. Particularity nvidia video drivers, which are needed for decent 3d performance on our higher end desktops.
My experience is that Linux needs too much work for pulishing the desktop user environment. Multimedia codecs, fonts, drivers lack, appearence themes are key areas to improve experience. Masive deployments and standarization is very important.
we use debian as server os
we use debian linux as control unit for engine en we have linux desktop to develop
for the control unit
OpenOffice is great and has come a long way, but it is only one app that businesses need. Viso, Accpac, canon print drivers and a bunch of other things(or alternatives) need to be available before it's feasible to install Linux on more than just our IT department computers.
Network is Linux based, ie: Samba file server, FTP server, VMware server, FreeNX. Setup for remote login with FreeNX, and run Windows apps with VMware server (WinXP Guest).
Deployed over 100 Kubutu clients well rec ived no problems
One thing is often said by users: "it is not windows compatible". It is 99% not true but... May be a better communication topic?
The Windows market provides lots of knowledge about, for example, security threats and what products/techniques to employ to protect aginst them.
Linux may be inherently more secure, but management and non-I.T. staff (and even some I.T. staff) don't necessarily know this, or to the extent that it is true/not true, and can't often get a clear idea of what products/technqiues they need to employ to make everything safe.
The Linux evangelists running around screaming "Linux is the best" don't really help a reasoned debate, and don't inspire confidence by management, even if they are right.
SuSE since 6.2 (now 10.1)
Problems integrating with SingleSignOn, problems integrating with Active Directory, need ability on desktop for an admin user to "Run as", problems defining access control for executables, JAVA slow, lack of FLASH or Shockwave support, problems deploying updates/pushing out to desktops using Microsoft SMS
It has been a great experience so far. The emulation provides access to all of the apps the teachers need to run and a the same time we are getting them use to using Firefox, Open Office and such. It will be an ongoing adventure I'm sure.
1.Linux needs to improve the experience given to the end users(in particular desktop users).This is generally a matter with the usability of the Linux desktops.
2.We should first try to release an OSS app for windows and give the same experience on linux too. This will greatly reduce the fear of the users and they will feel releived when they hear that the app that they daily use is available as such on linux.
Conversion from Windows was EASY, since most work was on Macs. But proprietary audio/video formats leads to some problems. We use Windows only on rare occasions, so keeping a dual-boot machine is fine - if it gets a virus, it doesn't affect anything else.
The lack of preinstalled systems resulted in a long lag between hardware purchase and first use. Application deployment has also been a long process because individual users must request that the administrator install most applications (if they are to be tracked by the distros package manager). In a small research group like our own, the administrator may not be available for such requests for days at a time.
One major issue encountered has been Video desktop, and sound support for some users that are very comfortable with a windows environment.
Ease of interoperability and connectivity between operating systems is also an issue. Most of my users need to have things done "automagiacally" and have easy, obvious and recognizable connections to resources. they need to be able to click on "that link" on their desktop and be delivered to "that network connection" where they store "that stuff" and i need to be able to set this up in a manner that does not wate my time. i have other things to do than create links on a dozen deskotps.
I also need the ability to use programs that my users are familiar with. I need to be able to run quick books without the need to boot another operating system in a virtual space to do so. I need to be able to use all of the various program and web-based Peoplesoft applications in the same way.
I support users that could be compared to the opening scene in the movie " A Bug's Life" when the stick falls across the path. The ants in line are lost becasue of that simple change in the environment. the "foreman" has to take over and lead the panicked ant around the obstical.
To be able to successfully deploy Linux as a viable alternative to "other operating systems" I need to be aboe to roll it out without dropping any sticks on the path.
vmware ws -- good
I use FC6 at home. Here at hospital we use FC5/FC4 on all servers and a couple of desktops. I'm very satisfied. I think the best pathway is to migrate our home grown applications to web model in order to leave MS Windows.
The biggest obstacle is specialized applications (accounting, real estate, construction). Until these programs go web-based or see Linux as a needed platform, our options will be limited.
Development of a business class desktop for use in a terminal server/thin client environment is a priority for our company.
Difficult to migrate and deploy office productivity as general employees are not Linux savy and installation of Linux is not intuitive for non-technical audience
Documentation has been hell. Once a problem is understood, though, implementation is straightforward.
At this stage we have not deployed a LDT to any users; we are still at the planning level.
No issues but media makes it sound worse than it is.
LTSP has worked well for us for 4+ years and continues to improve.
Better SATA RAID support for RAID cards.
My Linux experience has been wonderful.
Trying to raise some interest with my students' using open source and non-Redmont OS
We have deployed on various systems, the largest area of concern is with the suspend modes and power management. This needs to be a joint effort with the main hardware vendors. Better customer experience with both windows and linux would be better - Novell seems to be doing this.
Learning curve too steep for the average user, lack of trained tech support, and lack of comercial application compatability are the main stumbling blocks.
Many problems with printer drivers...
None
Get the installation of new programs easier and standerdised. Please.
Wine and other emulation programs fall miserably short. Linux is my choice of OS however my business feels as if Linux doesnt have enough support for peripherals and lacks enough user base to make it usable in the mainstream of our corporation.
Still trouble with some homepages that uses shockwave, javascript or flash.
We use modules (modules.sf.net) as a complement to applications in our distro and also for proprietary applications. This have been critical to us, since we now are able to have a lot of applications and different versions of applications installed at the same time.
ok
Drivers areslow to materialize. My scanner driver broke when I went from Suse 9.3 to 10.1 and it's still not fixed.Commonly available printers and scanners need to be supported, period.
very good
Easy to use, good usability. Missing working ACL support in Fedora Gnome, and ACLs in NFS4. Would have liked something like Dreamweaver
Differences between rpm/deb type systems and Gnome/KDE are annoying. Compiling things is stupidly difficult.
Good so far, students dont notice much difference as we use samba to sync the server (also linux) so that the home areas are the same on both systems, not much software used so not a problem, had to change things since as the home icon to say my computer etc etc
Biggest problem is a lack of tools for syncronizing Windows Mobile 5.0 devices: that is the major showstopper.
Once I figure out how to use something, I can implement it. The flexibility of Linux is awesome, but finding out how to do things takes a while. The process of deploying Linux has enabled me to learn many things. I love the ability to "play" with software. At this point, I need certain applications or equivalents to run in Linux, improvements in ease of use for the end user (dialogs, steps to perform functions, etc.), and some improvement in driver support.
wireless & video driver issues have been huge. cross-distro packaging a must.
Currently running on one of our servers. Hopeing to spread out to our mail servers next
Seamless virtualization *could* enable IHVs to preinstall Linux on systems so that e.g. games could still be played. Then consumers buy, then hardware peripheral companies might be persuaded to at least document their hardware so we can write real drivers.
I have been a personal user of all amjor distros for over five years to study the level of usability. At present using Suse Linux SLED 10 on a couple of personal machines.
I would like to have more Desktops with Linux already instaled on the market
Better Windows environment integration needed
There are still some stability issues in even major distributions if newer functionality is required. A more extensive testing regime would no go amiss.
Harder than hell to get management to listen to any arguments in favor of anything but windows.
Many large corporations have Windows mandated for all their employees For them to make a decision to replace thousands or even tens-of-thousands of Windows PCs with Linux is a monumental decision. I've seen one platform dictated to remove interoperability issues. For the Linux desktop to make headway into these large corporations, seemless interoperability with Windows PCs is mandatory.
Applications are the most important factor.
The main obstacle is that the management and salesforce are more or less locked in to Excel and solutions developed spesific for this application. Norwegian softwarevendors in the economics area are also usually Microsoft only. If you try to touch Excel, they cut your hands off. Reporting, reporting and reporting, as well as dependency of MS SQL.
My experience is BETTER FONTS (best Hinting, Smooting, Antialiasing etc...) We need in the past and still need Better font Technology in Linux.
OpenOffice, used primarily to edit and view MSOffice documents takes too long to start.
Windows share compatibility is problematic sometimes (long timeouts...)
linux has reduced my time working on fixing broken machines. the greatest problem is finding people who know linux. I'm an art therapist and know more about computers then some of the staff my agency pays to fix things when things go wrong. I use to be able to use my e-mail with little problem. one day on the widowns server and some kid screwed it up. His boss knew what to do and now she is not there now. at home I could not get my PVR-150 tv card working. Called tech support and they were like hummm don't know. the advertise that they know linux. most hardware works unless it does not. we depend on donated computers. many do not have sound. only linux non linux program I need is quicken. at some point will try to get it to work with linux. gnucash is just not up to the task wish it was. have told people this is the last year I will work to fix a ms machine. just is not worth it anymore. linux has gotten sooo easy to install and use. took 9 minutes using mandriva 2006se on a dell. took me hour and a half to get the same results from MS.
Very painful in getting codecs and so on in order to do any video work whatsoever. We had to switch to Windows XP to do our video editing. We contacted SUSE but got helpful response to our problems.
Must increase USB support! Some USB Flash Drives won't work in Linux. AutoCAD and other important professional software must have Linux version or counterpart!
i'm using linux at home as only one OS since 2002. Now it is gentoo.
Minimal problems with basic requirements (mostly office and web aps).
If I could make one suggestion to the open source community as a whole I'd say that we need to focus on making Desktop Linux a solid choice...across all distros. We need better device support, better app support, but most of all we need to get the platforms stable and working cohesively with supporting desktop standards and subsystems that will allow companies and developers to start building compelling desktop applications
Very good! CD music players should default to FLAC - store bought music comes with FLAC in sub-directories.
many niggling and time-consuming issues, especially with convenience tools that change settings without notice; uninformative or hidden error messages that make diagnosis of problems difficult
In my experience the bigg hurdle is hardware and software availablity
I've seen a huge improvement in the last year in quality of major distros and hardware compatiblity. Those need to continue. Also, the following needs continued work:
- multimedia support easy to setup/or preinstalled
- wireless drivers and usability needs some further improvement
- 3D desktop & support needs to become more seemless in the user experience.
Those all apply to my personal prefernece as well as what I think will enable further growth in linux desktop usage.
Personally, I'm also bound by the need to MS development tools. I wish that wasn't the case, but I have to live with it and accept that will continue for some time.
Suggestions:
- ability to upgrade an application without upgrading the whole operating system
- a way to integrate third party software packages with OS managed packages (without breaking all sorts of dependencies)
needed programs won't run correctly,, set-up for multimedia cumbersome at best, Instructions, to difficult to sort through all the different makes and versions to decide which one you have, then have different instructions for the same proceedure.. neither of which work!! THe same problems will keep Linux from ever being a household desktop.. Joe public can not (will not) take the time required to MAKE everything work..!!
Have SUSE 10.1 x-86-64 well integrated into a Windows world, including email functionality via an Exchange server through Evolution, networking to half a dozen printers through CUPS and full share access through SAMBA. Most people in office are unfamiliar with Linux and are astounded to see a fully-functional computer with a sharp-looking desktop screen and a huge array of software, and all with no license fees, no purchase orders, no worries about software audits/tracking, etc.
Long-time Linux system-level developer (Wine, Texas-Instruments WLAN, ALSA, kernel, network, ...).
Company Linux boxes used for development mainly, amount of desktop use currently far less than desireable (need to play a more active role in decision-influencing). Company excessively Windows-based, only few "pockets of resistance".
WLAN drivers should be approached much more aggressively by OSDL (i.e. very actively inquire hardware manufacturers about their support status and thoughts/concerns, community input will often be ignored no how matter how frequent/big).
OPC ("Open" Process Control) seems to be a wide field of process automation (admittedly non-Desktop) where Linux support is awfully lacking, a similarly capable open standard for Linux use should have been established a long time ago already, it's very astonishing/depressing that nothing suitable developed in this area.
This failure most likely means that Linux is a complete non-contender in the strongly networked kind of process automation technology.
BTW, Google "OPC Windows": 1.460.000 hits!)
Has to be able to inter-operate with non-linux system. This means collaboration with MS office/outlook users. It needs to be as seemless as possible. It also means application-specific compatibility, financial software, etc.
Big company need, good desktop managment/policy tools, central application support (terminal service, citrix and web).
The more out of the box the systems become, the better. Users will have enough trouble with re-oriented menus and app level changes. We need to ensure that we have out of the box ability to sync with Palm, configure wireless devices and access, etc.
Good experiences, although tough competition from Windows-people who want to integrate everything in AD (and thus want to select the distribution accordingly).
WPA support for wireless devices is flaky, laptops in general are a problem if some kind of offline-use is planned (home directory etc).
We are developing analysis methods for clients that may be turned into applications if appropriate. Linux is the best development environment. If others wish to port to legacy systems ....
We've had very very good success with Mandriva 2006 in a 15 seat laboratory for a VLSI design class. The GUI of choice is KDE/Konqueror combination. In other labs though, people are reluctant to take up linux because perception that windows is easier to program for with visual studio and the like. Additionally with our many of our foreign students (particularly middle eastern) they prefer windows because of certain prayer time players that don't seem to have an equivalent in KDE/GNOME/X (ie. Applications), or in the case of some of our Chinese students, the ability to us video phone chats and IM with video to family and friends at home, keeps them with windows.
Personally after decades of using linux and Unix, GNOME and KDE need to be unified is someway (preferably making gnome more KDE like). Specifically I absolutely hate GNOME's file query browser when saving or loading. Having gnome bring it up one way and KDE another just increases the learning curve.
The other really huge hurdle, I believe is to bring some new technology to the way configuration files are designed. No one wants a windows like registry, but it would be nice to bring some sort of consistency to help users through configuration. Here is my suggestion. Design a GUI script language that can be embedded in the remark statements of the configuration file, that would direct a gui editor through the configuration options. That way, a when a file browser sees the file, it will have a magic that indicates it's a gui-editable configuration file and when editing it could help the guide the user through the configuration options. For example, if you ever sat down to configure samba, there are literally 1000's of options that will scare an expert. For a novice adding a host to the /etc/hosts file can be an experience. Even with years of experience with linux, I sometimes have to do a man-page to recall the options needed to configure NIS. If the instructions for editing configuration files could somehow be embedded into the configuration files, I think it would be a huge benefit towards the acceptance of linux and dispelling the myth the linux is complicated.
the priamry issue remains support for peripheral devices and web cam support...
Better ability to run Windows apps.
my major inconveniences besides missing drivers or an application only available for MS Windows came from (1) inconsistent ways that KDE, Gnome and WMs manage settings like startmenu entries, layout differences (try to change font/window settings for GTK2 apps if you don't have Gnome installed :( ) or the way the system tray is handled and (2) dependency hell which came from different apps using different versions of other apps, I am talking here especially of multimedia apps.
I would also like some consistent configurations across distributions, like save X-Server settings (for example disable tcp-listen by default), but I don't know if that is applicable. Anyway, in case this is the last question: good luck with the survey :)
Linspire / Freespire meet our needs very well. The CNR software delivery service is excellent and we wish that other Linux vendors would consider going this route to make it easier for the end-users.
office environment
My experience? Probably more or less intermediate :-)
Other propositions: CorelDraw on Linux, please ...
Make it easier to see trhough the vast majority of all the possible software-switches so end-users have a better chance of a initially working system. And a way to go back to a former setup.
Relatively smooth so far. Most important feature to us is the ability to either replace current win applications with equivalents, or if absolutely not available, the ability to run win application in something like a wine environment.
User still needs to configure things that should be discovered automatically.
QC of desktop applications not very good.
In the Education business, we are subject to book publisher's software supplied with their books, along with perpiratary software aps which are exclusivly Windows. We need a Samba-like bridge to run Windows software.
Users also need to have floppy drive, CD drive, home directory, etc, to be accessible similar to Windows "My Computer" and such. The trasformation from Windows to Linux needs to be as close as possible in allowing users to perform the "basic" functions they are used to.
Drop GNOME. Well, more seriously, two different main DEs is a big problem for ISV
Deploying Ubuntu and vmware player to use quickbooks has been flawless. We are looking into migrating to SQL ledger down the road.
Switched in one month time frame.
I got dual boot for 1 month then I moved to linux only
The desktop is ready - application support from third party vendors that are tied to vb and .net code from M$ is the biggest problem.
Very limited. However, installation was difficult.
The Deployment of the workstations went pretty strait forward, couldn't complain about the installation. The managing of client's workstations is a problem that I experience using linux. Also peripheral support is also a factor based on work type.
I have enjoyed my transfer to Linux, ever since I was first introduced to Slackware in 1998. I really have enjoyed the changes that have occurred and have allowed me to read documents and files from Windows systems without me having to request intermediate file formats. I have converted to Linux since I run a shop that fixes Windows Machines, and I need to be able to connect machines to my network without the risk of infections. I enjoy the security that each user has over his/her documents. Keep up the good work!
Linux on laptops is horrible with docking stations and/or going from multiple monitors to laptop screen without logging out of X--this is a major problem
Proprietary graphics drivers are a pain, certain obscure hardwares don't work. Kubuntu works great, except 64-bit support is lacking especially among proprietary programs.
From the technical aspect of getting it done it was pretty painless as we run all linux servers in the first place. The problems arose from users being unhappy with limitations they were encountering when browsing the web or trying to hook up the phone/pda/ipod/usb thumbdrive. We also heard complaints about interfaces being confusing. Items existing in more than one menu congruently, things not in a logical place. Items not being placed in menus after they were installed.
We need Windows apps to work on Linux (duh)
A major concern is user friendliness- Linux is set up too much like a server, and to be ready for the desktop, the file sorting needs to become more user friendly (eg. usr/bin, etc.)
k/ubuntu has made great strides in making the linux desktop more user friendly. although i use linux at home, i am often forced to dual boot to windows b/c of specific media hardware that doesn't have linux drivers. while ati and nvidia have been go at providing drivers for linux, most (like creative labs) have yet to make this available. while i completely understand that it's a HUGE amount of work to support all the hardware out there in linux, it has to be done. b/c for an avg computer user, they don't have the technical skills to figure out what's wrong and would go back to windows in a heartbeat b/c that's their security blanket. "this scanner/printer/usb storage/(any hardware example) doesn't work in linux, but it does in windows. i m going to use windows." that's what i hear a lot at work and among my non-technical friends.
I am using FAI on Debian to deploy my desktops. Most of my users are happy with their desktops, the evolution-exchange plugin could be better. I think that the quality of documents produced on OOo is passable, I don't know if it is because of the fonts or the printer.
Definite development of tablet devices and support is needed. Many of our employees are on the go and need more power than a PDA or phone can offer but with the convenience of being on the go, not sitting down to the notebook.
Biggest problem is lack of drivers for scanners and printers. Canon (and many others) scanners are a NO-GO, which represents a major hold back in my company's case.
Also out of the box support for PCMCIA GSM/UMTS modems (ex: Option GlobeTrotter). While there is support for USB modems in the kernel, there is no user friendly front-end to set-up a connection trough the card.
Biggest problem is senior management resistance to non Microsoft software/services.
in past 4 years I tried number of distributions (Suse, Xandros, Knoppix, Ubuntu) and after some time I was frustrated that sooner or later I had to go and change one or more conf files to achieve something. Admisntartion should be more user friendly and intuitive. Would be nice to have a plug and play or wizard based approcah to a lot of conf related tasks
Make fonts on the linux desktop look better or as good as on Macs or Windows _by default_. Please.
I think WINE is the crux to windows migration. For example, a web development company isn't going to touch linux without photoshop support which it can't get till wine fixes MSI compatability. However photoshop (and WINE) requires better fonts on linux also.
Please put a major focus on printing. Now days, printing should be simply 1 click away, however finding drivers for printers other than a select few is EXTREMELY difficult.
Biggest complaint from salesguys with Linux on laptops is the occasional "fsck after N mounts" that happens at startup even though the system was shut down cleanly.
Would be much nicer if the fsck happened at shutdown time so wouldn't delay their presentations.
Worth it, but steep learning curbe for some users.
Having to modify repository servers to add in restricted modules-though understandably not included; is a hassle. It should be given as an option under normal repositories even though FOSS is against blobs. Too much alienation of a migrating end user base can be detrimental.
Our company is a Windows shop, and I am enslaved to it. Please save me from Windows hell! He he.
The primary reason I am making this push is to simplify manageability and peace of mind. The only "objections" so far are just the typical "but its different/complicated" only because people haven't ever used it. The way I plan on rolling it out, from a useability standpoint, it should be far easier to use than WinXP which is what the company is currently using on most of their desktops.
Make sure the necessities are in the base installs. Ubuntu is missing a network network search tool and drive configuration tool (current).
Linux is trying to play catch up. Rather they should beat them to get adoption. Think of interfaces like Secondlife.com that will change the playing field.
Deployment is easy, but getting user buy-in is still hard. The Linux community needs to focus on developing a solid OS, and less on evangelism for Open Source, etc.
I <3 Linux
Linux desktop for myself since 1996 (Gave up trying to administer Unix machines from Windows 98 due to constant crashing). Ran dual boot until VMWare became available. Run Windows for Corporate apps in Virtual Machine since 1999. Really need linux clients for the Corporate apps supplied by large software vendors (Peregrine)
Generally deployment is quite smooth, but getting certain capabilities provided by specialty applications is difficult, and once again gaming on linux is poor.
Ubuntu is awesome for my purposes all except for its inability to run photoshop cs2
Better naming of applications - ie decribe what they do.
deployment experiences:
we are a market reaserch firm that is quite dependant on specific applications: SPSS, MsAccess, Excel. So we need equivalents for these type of software (the first is proprietary,so we'll have to run in in some kind of emulation or virtualisation - or maybe they'll release a Linux version, they already have one so it shouldn't be very hard; the second could be easily implemented opensource, great DBs exist, but not a friendly interface to them, for the last gnumeric or ooffice are ok). these being said we implemented succesfully linux only on servers, because only 1/3 key desktop software is available now. suggestions - an Access like interface to Postgres or some other good db implementation.
The "theme" of the desktop sometimes is important, the user-friendly is a major question, not the tecnical side.
Something has to be done about the level of game support for Desktop Linux to succeed. I own many older Win98 only games and can't play them on Win2000 or higher. When Vista comes out I fear I won't be able to play WinXP based games. Sigh, maybe it's time to forget a PC and go to a console only for games as long as PS and XBox are backward compatible.
In my experience, Linux needs to focus upon a second front, the end user. Experienced tech heads have no problem with RTFM but my greatest problem with deployment are the users who want windows "it just works" quality. Though this has been the greatest downfall for the microsoft environment, most end users really just want something they do not need to tweak or adjust at all. I can only hope that this does not disrupt the fine quality and control that tech-heads desire.
We use only small setups, basic web/mail servers, and individual desktop machines. The server stuff works like a charm, desktop needs a lot more help.
I am individual user working in scientific/engineering software development. Started to use linux due to a large performance increase of the compiled fortran code I develop and more-importantly to avoid a slower and slower running Windows XP with critical security issues and an IT department working very hard to limit the usability of the Windows systems for the sake of security...
Audio drivers need to work with all the ins and outs of a complicated card without the complicated controls.
Easy as pie, piece of cake,
Hardware support lacking for many video/wireless/modem devices is making it diffacult for us to support/recommend/deply linux
Hardware support is the biggest problem for me, especially wireless issues, suspend and support with my laptop - related to the graphics driver issues.
Gaming is a big problem for Linux there are a few games, but arranging support for a compatible format for Linux games would be ideal for future, than transgaming or wine wrapper. Needs to be easy for developers to transfer games quick. Make it look attractive.
!!!There is no marketing or knowledge of Linux to normal people who use Windows, they haven't heard of it or know any benefits of it. Get a marketing team or create an association between developers, distributions, desktop management to find a suitlable strategy
Things should be done in a more slow and controled method. New features going out with a log of bugs are causing more trouble then delayed features.
Ubuntu has been great... with great community support. Our biggest challenge is industry specific software that has no Linux alternative that is at a state worth using.
Everything is fine, except when it comes to web site using proprietary stuff. ie: wmp em.
mostly good
I have supported Linux servers in an enterprise environment (not in my current role), and I have found that although the Linux community is great about documenting problems and solutions, there often is too much information. Sifting through all of the solutions -- many of which are just duplicates -- can be time-consuming. And if there is one thing IT folks don't have much of, it's time. I would like to see one "official" support site for every distro. Although some distros already have something like this -- Ubuntu, for example -- many are lacking. Also, too many distros can scare away users and IT folks (although this isn't true for me).
thin client for students
simple transition took place four years ago. will not consider going back to Windows. found replacement software for all applications used under Windows.
Rething the paradigm from ground up. Don't imitate, innovate.
Please, file permissions is good. But make it easier to deal with it. Sometimes you install software and it messes with some file ownership an permissions and the only help you have to get the system running back is the community (forums).
People must stop spawning linux distros and concentrate on making existent distros better.
Work on wireless support. Perhaps this is just my card (bcm43xx). Device drivers are important. Engineering tools (Solid Edge) are important.
Prevent proprietary vendors from shipping shoddy Windows-like applications which trash the system, fail after updates and do not honor Unix conventions.
kickstart style installs are essential, printer driver support is essential - pick printers wisely, Windows sets the level of expectation - be familar with what users expect - know their "work patterns" (what tools they use, how they use them, who they interface (share files with, etc)), test test and test
Several abortive attempts to run desktop pc's - Mandrake and Red Hat. Ran into driver issues and difficultly installing apps (seems some arcane knowledge is required).
Had small trials (3 clients).
AutoCAD and MS Project are hard to replace
End user level standardization and simplification of following: printing (mish-mash of printing setups between OS and different applications, worst example: OpenOffice), file management (our users love "rox" application vs. native file managers [ex. Nautilus] 100:1. Simplify. Speed-up. Standardize), desktop/display customization (user should be able to switch resolutions, sync./freq., font sizes, etc. "on fly" and without any admin. priviledges).
Robustness: crashing application should never bring the desktop down.
Do not oversimplify customization: example newest Firefox (bad) vs. original Mozilla (good) preferences setup.
Beware of dramatic desktop operations slowdown when Linux-only environment is assumed: make sure that if Win OS, Samba, Sun OS, etc. are interacting/cross mounted desktop applications don't go into snail-pace (example, Gnome desktop and Sun OS filesystems mounted on Linux workstations,...)
The transition was very easy and most employees hardly notice any problems. My time has increased drastically because I don't have to clean the PC. With the higher security rights, our employees can't install software by them selfs. Making things much easier on me. If it wasn't for our Accounting software and our Shipping software, we would be off windows completely.
Not being able to simply install a program and expect it to work is the single largest factor. This is due to differences between distributions and even versions of the same distribution--in other words, the fact that each one depends on its own application dependency heirarchy.. A very, very bad idea.
Just started with s few ubuntu LAME servers, and experimenting with ubuntu and freespire and linspire desktops, looking to learn more about all and other distros.
Use Puppy Linux on my desktop and Ubuntu on my laptop, very successful with both.
Generally frustrating due to closed formats. Web browsing is problematic due to WMV format, and of course VERY LITTLE GAMES.
I have personally installed all the major desktop and server distro in order to road test them. While I use RHEL on my current server farm, my desktop of choice is SLED10. This is a recent move away from ubuntu due to SLED feeling far more stable in evey day use. Wireless networking is still a major obstacle!
with the ubuntu distribution, I having a good experience, no need for windows any more
With enhanced graphics, greater ease of application installation, and dramatically increased driver support for hardware and peripherals, Linux would easily overpower my usage of Windows. But it has a way to go before it gets there.
Love it.
deployment mostly for special applications that can not be run using MS Windows.
Generally good with frustration with multimedia and drivers being broken by kernel upgrades
very good
The backup systems need better configuration and automation.
Difficulty with certain web-based critical tools such as LiveLink and ADP payroll apps requiring activex controls to work like in IE6. Extreme reluctance by IT group to switch away from Windows on the desktop. At least 1/2 of users are Linux-savvy development staff. Corporate mentality problems, not end-user problems.
Switching both servers and desktops to linux has significantly improved productivity while eliminating problems commonly related to IT infrastructure.
For several years I (myself) does not notice any major obstacle. I develop for windows on vmware and my primary OS is Linux.
Setup and system upgrades are the major drawbacks. Setups at times can be extreamly lengthy to get drivers configured correctly. Additanally upgadeing linux systems seems to create more problems than it fixes. For example most applications need to be reconfigured after an upgrade leading back to the setup problems that I forgot how to solve.
Have tried several flavours of Linux, and worked extensively in a Debian environment.
Lacking the ability to use the required Windows development tools, and also not being able to tolerate a VM when running these apps, Windows has been used exclusively due to applications in use and being developed.
Printers, Drag-and-Drop, a version of Firefox that can handle just as well as the Windows Version, better spreadsheet program.
Cmon.. it's the 21st Century already..
Things should just work. Right now Linux takes too much dicking around and fidling and researching and tribal knowledge to get it working out of the box. More drivers are needed, I don't care if they're closed source or not. nVidia works great -- so what if they're not open source.
Iemjoyed Learning To Program With Kdevrlop And Intergrated Subprograms, Although I Had ASall Documentation Problem (9.2)
Need an easier way for linux clients to log into a Windows domain environment. Add this to the installation menu.
ubuntu smooth
More work needs to be done to integrate projects like Beagle into the standard desktop. There are a lot of cool applications floating around, but they need to be integrated so that the Linux desktop has feature parity with Mac OS X and Windows Vista.
The biggest headache I encouter is lack of drivers. For instance, my Blackberry phone/PDA doesn't charge because Linux doesn't recognize it and therefore underpowers the USB port.
Problematic support for USB Flash drives and network file sharing
More distributions should install as easy as Ubuntu, with the additional applications supported like SimplyMepis, along with the ability to add all the extras like you get from Automatix. If this came out of the box, user adoption on the unindoctrined level would be much quicker. Honestly Linux still needs a killer app, and some killer games. Honestly, you have the nerd and IT savvy business votes already, games and the major killer applications will bring many more users, and bring more talented eyes into the tweaking community. I'm a veteran of 10+ years linux and longer UNIX, I have no issues with my desktop, it's perfect for me, but I see the above 2 things as the biggest hurdle. Additionally, all networking ESPECIALLY wireless needs to work out of the box on as many distributions as possible. Biggest hurdles for new user is not being able to look up how to get help, because of no wireless access support. I'm glad (K,X, Mepis)Ubuntu work on the Centrino chipset (2100) but I know lots of people who won't waste their time getting the NDISWrapper working with the old linksys PCMCIA card, that used to just work under windows. Just my 2 cents. This is a great initiative. Good luck.
I use Windows XP under VMware primarily for Quicken and QuickBooks. I also use Windows for iTunes and Adobe Acrobat Professional, but those applications are less critical for me. I'm looking into GnuCash as an alternative to Quicken and QuickBooks, but that will be a tough change.
Need more positive approachs from PC vendors
Gentoo
professional apps required for our business are not available. linux is not mature enough. linux desktop is just a copy of windows instead of trying to do something new. why switch?
Ubuntu is really the most mature linux distro for the desktop out there, and most of the employees do prefer it against that other OS. What we miss is some standard packaging among all distros.
Lack of software needed when needed and lack of support by our vendors (who have their own programs running on Windows)
Need to emulate the "Double-click to install" functionality of windows/mac. Current installation methods are far too complicated for the average user.
need drivers to be easily installable/configurable. ati drives for example are very difficult.
None @ this time. But I'd like to say its great what you guys are doing, keep it up, and good luck!
The problem lies in the migration. There are many irreplacable Windows applications at the moment that will still need to be used when migrating to Linux. It is much more convenient to just run Windows instead of dual booting so the solution lies in making the Windows applications run REALLY efficiently in Linux until companies being coding for Linux as well or even eliminate the need to do so.
We have had very, very few problems. THANK YOU, GUYS!
I personally am running openSUSE 10.0 but when I tried installing it on other machines, the people was afraid to migrate as some applications which they are used to like Visio(this