Modify Colors
Wine In Ubuntu: Implementation Is Key
Scott Ritchie, an Ubuntu MOTU, posted a suggestion on the Ubuntu Development mailing list today that Wine (what?) be put into Ubuntu’s main repositories (what?).
Wine is a piece of software that allows Windows applications to run natively on Linux. The problem is that not all applications work perfectly through Wine. Some applications work perfectly, while others don’t even install. Furthermore, as it stands, Wine is something of a power user tool. It is usually poorly integrated with the system and the hit-and-miss application support , despite the help provided by the AppDB, can all be a pain to deal with. This is not to say that Wine is not improving. Recently, Wine celebrated their 1.0 release after 15 years of development. You have to give them credit for their work (reimplementing pieces of Windows on top of Linux is no easy task ), but that does not change the fact that there are still some rough edges.
These rough edges are exactly why the current proposal is not to just drop Wine into Ubuntu’s main repositories suddenly and then move on. Instead, Wine would be gradually eased in, much as Compiz Fusion (what?) was, along with integration work. The eventual goal would be out-of-the-box recognition of .exe’s and Windows software discs and invisible integration between Wine applications and regular applications, particularly in terms of files.
This sort of integration could mean that new Ubuntu users could easily install their existing Windows software inside Ubuntu, significantly easing, and in some cases enabling, the transition to Ubuntu. Unfortunately, there is also a risk that it frustrates new users instead, since not all applications will work. To ensure that integrating Wine is a purely positive move, it is important to remember three things:
- Implementation is key. This is the biggest thing to keep in mind. If you just dump Wine’s code into Ubuntu it might technically “work,”but no one will like it. If instead, you gradually ease Wine in and integrate it wellwith the OS, it will be received much better. It is absolutely essential that, if Wine is chosen to be included in main, its implementation will be well thought out.
- Don’t over promise. Wine is great, but it can’t work miracles. If it is advertised as a way of allowing you to run any Windows application perfectly on Ubuntu, people will be disappointed and feel that it doesn’t work.
- Give users some idea of what will/will not work. The user’s experience could be greatly improved if they were able to see if an application will work easily. Perhaps an Ubuntu-specific version of the AppDB could be created. The information from there could then be used to add some sort of indication of the application’s compatibility status to the application’s icon. For example, I might download an EXE and find a green check on its icon, indicating that it was expected to work.
If the decision is made to include Wine, it is important that everything is done with caution and the end user’s experience in mind. That is not in any way intended to say that it is a bad idea, in fact it is a great idea, but it is very important that Wine’s integration is done correctly.
Related posts:
- Using Wine + Wubi For A Smooth Transition Yesterday, I discussed the possibility of Wine’s inclusion in Ubuntu....
- Steam on Linux I’ve always avoided Wine, a Windows-compatibility layer for Linux, based...
- Will Ubuntu Be The First to Integrate Web Applications With the Desktop? Yesterday, Canonical’s founder, Mark Shuttleworth, announced the next version of...
- A Tour of the Ubuntu Software Store About a month ago, Canonical announced plans for an Ubuntu...
- Microsoft Office for Linux – It Just Might Happen I bet you never thought you would hear the term...
Switching back to Windows is not a bad option either. I did; after two years; got tired of waiting for Linux to mature. Wine is getting better and better (very very very slowly though) but it is not like the real thing (PSPad is my main application ((text editor situation is really exemplary where Linux stands – way too many of them and not a single one is really usable)) ). On the desktop even Vista (with aero and UAC disabled) is way better than the best of the Linux distributions as far as a normal user is concerned. OOo, Firefox, Thunderbird and even Gimp are much more responsive and smother on windows. On windows almost all applications pretty much appear before I finish clicking my mouse button; on linux I have stare at the spinning ball every time.
On Ubuntu, it they want to become a viable desktop alternative, I agree – Wine is the key. It was for me while I was using it.
I thought this discussion was about how to do it not why to do it. The why has been discussed way way too much, its the same debate that’s been going on for the 15 years of WINE.
How to do it? This is what I’d like to see:
* Themes: Windows apps should look like they belong on the Linux desktop. At the moment they look like underdressed step-children. This could be extended to be dynamic but I think it was already discussed on WINE lists that simply making it work for the default install of Ubuntu, Fedora, Mandriva and openSUSE would be great.
* Native dialogs: In the same way that OpenOffice.org got native dialogs (File, Save, Print) and suddenly felt like it belonged on your Linux desktop it should be possible for Windows apps to use your GNOME or KDE dialogs.
* Make configuration integrated into the platform. The wine-config application looks like it doesn’t belong. We should aim to eliminate much of that or work out how to integrate it into the general Linux environment.
* Multiple WINEs by default. Some way to ensure that you install each WINE app in a separate container should ensure that working WINE apps aren’t destroyed by that horrid game you just installed.
* AppDb: please don’t make an Ubuntu one blimey hasn’t anyone learnt anything! Its hard enough maintaining the WINE one and you suggest creating a whole new one??? Rather enhance the one that exists to do what is needed for more seemless integration. It would be nice to be able to ping AppDb and based on the .exe signature get the status of that file and show that on its icon sounds cool.
* Taking the wine-doors etc approach might also be the right strategy. This might introduce an interesting concept for WINE – WINE packages. I.e. you can run anything but a WINE package will ensure that it runs correctly on your platform.
Oh my £^”&!n3 eyes!
CTRL+W
Associate .exe files with Wine. What’s so hard about that? I’m sure there may be ways to do more “integration” but I can’t think of really anything more that would be needed.
In any case, this better not be some lame thing that makes Wine be some “custom Ubuntu package” more than it already is. If you, say, want to integrate Wine into Gnome more by making some Gnome Wine GUI, then make that be a separate package. Programs are supposed to work together, not turn into some giant exclusive software stack so that users are forced even *more* than they already are to rely on your repository instead of being able to easily upgrade directly from the developers. (which is possible only because the devs make a Ubuntu package in the first place)
Package standards where aaaaaare yooooooou…
The very minute that anyone can load and execute .exe files is the day that Microsoft begins to fade. Most of my clients that were reluctant to switch to Linux were attached to a single program that they couldn’t leave behind. If they could execute them in Linux I would have about 60% more clients.
qt/gtk only solves a very small part of the problem for
interesting apps. Look at Chrome; there is very little
use of OS dialogs in that app, nearly all the porting work
is elsewhere.
If you spend much time on the Ubuntu forums helping newbies, you know that newbies seem to have no problem getting their hands on WINE as it is; you also know how much frustration it causes them trying to get it to work.
One of the most painful things to watch a newbie go through is trying to get some crappy little application they used in Windows to work in WINE when there is a perfectly viable alternative available in the repositories. WINE integration OOTB will only exacerbate this.
It sounds silly, but there are two things that seem very hard for new users to grasp:
– Linux is not Windows, and doesn’t natively run applications designed for Windows.
– The repositories are the FIRST PLACE TO LOOK when you want to install software to accomplish some task.
If those two points could be effectively communicated to new users, it would eliminate the bulk of “software install” woes people post about on the forums.
IME Wine works well for those big mainstream apps like Office or Photoshop. But your run-of-the-mill “I bought this game for the kids off the clearance rack” software generally fails to run for some reason. If you give people the impression that these things should run, and they don’t; then you’ve just made Ubuntu (and Linux in general) look that much more like second rate junk.
In this imperfect world, denigrating those who cling to the tattered threads of their proprietary apps will convert far fewer than a nice, supportive, Rogerian approach will.
The issue also remains of niche apps not paralleled in Open Source offerings. Linux is still the new kid on the block, not every itch has been addressed by its adherents, nor every niche represented in its fold.
Finally, Wine can offer salvation when a bug in an OSS program might otherwise hamstring the user. A bug in K3b once stopped me cold. Running a Windows burner in Wine saved, (or, at least, substantially expedited), the day. In the real world, what works is what works and that’s what matters.
I fully support the concept of rolling Wine into K/Ubuntu. It might be appropriate to give it an opt in/out check-box at installation, though.
“The eventual goal would be out-of-the-box recognition of .exe’s and Windows software discs and invisible integration between Wine applications and regular applications, particularly in terms of files.”
What could possibly go wrong?
To use Wine on Ubuntu right now you only need to add the Wine repository. This can be done by cutting and pasting two commands into a terminal. I think that’s easy enough for anyone to do.
I don’t use Wine to run Microsoft programs. I use Wine to run programs written for the Windows platform. Some of them are even FOSS released under the GPL from Sourceforge.
The way I see it; Wine allows Linux to steal from Windows.
But then, I’m an imbecile…
I am concerned about malware and viruses if this happens. Yes *most* Linux users are savy enough to avoid these pitfalls. However, including the ability to run Windows .exe’s is going to attract more Windows users (a good thing). But, these newer users are less likely to sidestep that “you have a virus, click here to fix it” BS you get from the web. If Wine just kicks in and runs that .exe, you have an instant malware problem. I could cite other examples but those reading this should get what I mean.
It is important that in integrating WINE that Ubuntu NOT allow auto execution of exe and similar files except in well defined circumstances and then after ensuring some fundamental awareness exists in its community esp newbies. People, especially newbies, using Ubuntu may well be under the impression that they continue, with WINE in their environment, to be as secure as without it. The simple fact is that WINE will execute malware as readily as Windows though its impact may simply be to totally destroy the users home directory. Malware folk will not be ignorant of this potential for their endeavours and may well craft exploits designed to be a nuisance in the Ubuntu environment specifically.
See https://bugs.launchpad.net/ubuntu/+source/wine/+bug/85338
Wine should only be used for legacy support of windows apps that no longer are in development. The fact that people are using wine as an excuse to say that a certain app does not need to be ported because it works perfectly in wine are imbeciles.
The fact that Google took this path with Picassa and plan on doing so with more of their apps upsets me even more. Why not just use qt/gtk libs from the beginning, its easier to go from linux to windows then vice-versa.
Especially since they are one of the companies that claim to be sooo supportive of OSS. True that doing so they did contribute to wine, but just that doesn’t make it right.
wine integration should be done along with wine-doors & playonlinux, etc.
i get a list of apps/games that work and even offers auto-install some of them