Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Why port?

By hardburn in Op-Ed
Sat Mar 24, 2001 at 08:39:46 AM EST
Tags: Software (all tags)
Software

Why does the GNU/Linux community bothering with asking companies to port their software? This is an inefficient way of going about the problem, as you have to ask for each and every new piece of software to be ported. This goes against the hacker philosophy of "only do it once, then have the computer do it for you". How about we ask those companies to help out WINE instead?


In theory, with a sufficently mature WINE, any Windows program can run on any *nix box. This means you only have to create one kind of software to get all the rest, and not just on GNU/Linux, either, but for all of *nix. If another Free solution that a lot of people would like to move to (perhaps GNU/HURD will be one day), but it doesn't have binary compatibility with GNU/Linux, we have to ask for all those programs to be ported all over again.

OTOH, if WINE can run it, all we have to do is port WINE to GNU/HURD and we get the whole bundle.

So my suggestion is that we don't ask for companies to port their program, but rather, we should be asking them to put the programmers, who would have been working on porting the program, to be put to work hacking on WINE. This way, the duplication of effort is minimized.

This isn't perfect, of course. As stated before, in theory a sufficiently mature WINE will run any Windows program. As always, turning theory into practice is very difficult. Some programs have a hard enough time running between versions of Windows, much less an "emulated" version of Windows. Still, the effort to port a single program will be greatly reduced. If a program needs only minor patching to run on WINE, then I think WINE has done it's job.

Still, you will get a slow down with the emulated API layer. Just how much (and if it's actualy slower then Windows) may depend on a lot of factors. Games might suffer the hardest, depending on their features. I wouldn't want to run Unreal Tournament on WINE, although I have no problem running Starcraft and maybe somewhat older shooters like Half-Life. WINE does have facilities available to help in porting the program to a native binary, so games will still benifit, though perhaps not as much.

In all, helping WINE can only help but smooth the translation of those moving to GNU/Linux. Right now, I beleive that it is one of the last places that GNU/Linux needs to take the desktop. Ease of use is almost a non-issue (just look at Mandrake 7.2, which I shamefully admit to installing once), although we still need to be able to buy GNU/Linux boxes off the shelf at CompUSA and Best Buy. However, getting WINE mature may help eliminate these last few gaps.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Poll
Do we need WINE?
o Yes, it's essential 16%
o Yes, but it's not a priority 25%
o No, not really 12%
o No, it's useless 9%
o We don't need a bunch of non-Free programs ported 10%
o red red WINE makes me feel so fine 9%
o whatever 6%
o -1: Flamebait 8%

Votes: 104
Results | Other Polls

Related Links
o WINE
o Also by hardburn


Display: Sort:
Why port? | 48 comments (45 topical, 3 editorial, 0 hidden)
Give it another go (3.50 / 4) (#1)
by regeya on Fri Mar 23, 2001 at 11:58:11 AM EST

It's something that has the possibility of generating some discussion) but it's a little weak and rambling at the moment. :-/

Oh and here's an obligatory screenshot. :-)

[ yokelpunk | kuro5hin diary ]

That screenshot... (4.66 / 3) (#4)
by farmgeek on Fri Mar 23, 2001 at 12:04:14 PM EST

is just wrong. ;-)

[ Parent ]
wheres the incentive? (2.50 / 2) (#2)
by rebelcool on Fri Mar 23, 2001 at 12:01:19 PM EST

whats the economic incentive to spend the thousands of dollars in labor hours required to help out if you're going to give wine away, and considering the relatively tiny linux community hates paying for anything anyways.

I recall john carmack talking about the linux version of q3A..stores would *not* order it, simply because they couldnt get rid of the damn things once they did. So they are all but abandoning linux as a platform, simply because its not cost-effective to develop for it.

COG. Build your own community. Free, easy, powerful. Demo site

Q3A and Linux (4.66 / 6) (#8)
by Anonymous 242 on Fri Mar 23, 2001 at 12:32:10 PM EST

The retail Q3A for Linux box was an atrocity of marketing and not seeing the forest for the trees.
  1. It was priced at twenty to thirty dollars higher than the Windows version.
  2. It was released well after the Windows version. Even worse, IIRC, it was released well after the holiday buying season while the Windows version was released prior to Xmas.
  3. It was hard to come and when available harder to find mostly due to most retail software outlets not having a Linux section at the time.
  4. The Linux binaries were available on id's ftp and http servers and worked with the data files from the Windows version.
I don't know that the Linux retail box of Q3A would have been a blockbuster if things had played out differently. I do know that the way that the sales and marketing were carried out virtually guaranteed that sales would be minimal. Why would any sane Quake fan have bought the Linux version when the Windows version was (1) cheaper, (2) came out sooner, (3) was easier to find and (4) "worked" with Linux?

Using Q3A as anecdotal evidence of the non-viability of Linux as a gaming platform is like using WordPerfect as an example of how Microsoft used nefarious tactics to dominate the Office Suite market. Microsoft may very well be guilty of nefarious tactics, but the WordPerfect Corporation, then Novell and then Corel have all demonstrated incredible incompetence in the way they handled WordPerfect. Chances are that the way the WordPerfect product was handled would have resulted in the same sinking ship whether or not Microsoft was guilty of abusing their OS monopoly to the detriment of WordPerfect.

[ Parent ]

nevertheless (4.00 / 1) (#12)
by rebelcool on Fri Mar 23, 2001 at 01:06:39 PM EST

theres many other instances of companies pulling out of the consumer linux market, adobe, corel and im sure others. Theres a good reason for it.

For practical purposes, there is no "consumer linux" market.

The 1% of few linux hobbyists, simply dont warrant the resources to develop for the platform.

Now, in terms of server software, where linux makes up a much larger percentage, oh yes. Thats where any money going into developing for linux is going, and should be.

COG. Build your own community. Free, easy, powerful. Demo site
[ Parent ]

Could be (4.00 / 2) (#13)
by Anonymous 242 on Fri Mar 23, 2001 at 01:32:11 PM EST

theres many other instances of companies pulling out of the consumer linux market, adobe, corel and im sure others. Theres a good reason for it.
Correct me if I'm wrong, but I do not believe that Corel has abandoned the Linux consumer market. Did they not just release Draw for Linux? Were they not the ones that contracted with Microsoft to deliver a .net implementation for Linux?

Even if I conceded that Corel was entirely abandoning the Linux market, given Corel's difficulty in targeting the supposedly lucrative Windows market, I'm not about to draw any conclusions from their performance in the Linux market.

Adobe, AFAIK, was never really in the Linux consumer market. The closest they came was a beta of Framemaker and the Acrobat Reader. Framemaker is hardly a "consumer" oriented app. And let's just say that the Acrobat Reader is hardly a cash cow. Did I miss the delivery of Photoshop for Linux?

I'd be more interested in how well Loki is doing and Applix. It will also be interesting to watch Sun's Open Office. IBM is also porting much of its consumer software to Linux. It will interesting to see what happens there. Linux is a very young market. It takes time to determine whether or not it is a viable market. How many companies with Windows products when v3.1 came out are still around? Over 70% of business startups are no longer in business five years later. Let's talk about this five years from now. It is quite likely that the lanscape will be a bit different and it will be much easier to gauge examples of success and failure.

For practical purposes, there is no "consumer linux" market.

The 1% of few linux hobbyists, simply dont warrant the resources to develop for the platform.

That IDG gave Linux a 1% marketshare in retail sales for OS implies that the actual Linux market is far higher than 1%. Even if it 1% is an accurate measure of Linux's presence on the desktop, that speaks of a very large opportunity. MacOS only has something like 4% retail marketshare and plenty of companies are making money off of that. 4% is even large enough for Microsoft to be interested in keeping Office alive on the Mac.

[ Parent ]
Heh (none / 0) (#25)
by regeya on Fri Mar 23, 2001 at 04:10:41 PM EST

I came about this close to buying Q3A in a story, but the clerk kept saying, "Now, you know this isn't for Windows. This runs under Linux. You know that, right? I mean, if you get it home and it doesn't work, you know you're stuck with it, right? I mean, not too many people run Linux. You run Linux, right?" I literally said, "Y'know, fuck it, I don't want it now that you've harassed me," and sat it down. I'd already played the demo under Linux, had a blast with it, knew what I was doing, and the clerk was too busy trying to save me from myself to even let me look at the fucking box.

Sorry, John Carmack, but maybe there's yet another reason the Linux boxed versions didn't go too fast...

[ yokelpunk | kuro5hin diary ]
[ Parent ]

WINE is a bad idea. (4.60 / 5) (#5)
by Faulty Dreamer on Fri Mar 23, 2001 at 12:11:12 PM EST

WINE, in my opinion, has always been a bad idea. And it always will be.

Sure, the concept might not be that bad of an idea. But, from the moment I found out about it I thought that all this will do is lead people to believe that we do not need native software on Linux/Unix. Companies will not be interested in porting efforts and will instead use the worst possible programming method (creating DLL hell) and transporting those terrible programming habits over the Linux/Unix. Does anyone really want to see DLL hell re-created on the Unix desktop?

My fears were confirmed when Corel released Word Perfect Office 2000 for Linux. It is a sloppy implementation of what is/can be a really awesome piece of software. Why go for emulation (I know, I know WINE is not an emulator) when you could port? Now, you suggest that we should not be asking for porting at all? The entire point of getting people interested is to get them writing native software. It runs faster and smoother if it is native. Telling companies to throw resources at a emulation layer is not going ot improve desktop productivity.

If we could use it as an interim step, that would be one thing. But let's face it, if people get used to sloppy programming habits they will hang on for ages (for 'backwards compatibility's sake'). I do not want to see a desktop invaded by four hundred different versions of WINE just so that I can run each of my programs. Seriously, there are already a number of WINE run programs out there, and each of them requires a different set of WINE libraries and a seperate run-time, none of them are compatible with each other, and we run smack-dab into the thing that most of us were trying to escape: Windows like software incompatibilities, each program destroys another. I'm sorry, but I don't want to go back to fighting my computer to get something done. It's just not worth it.

Recreate Windows on Linux/Unix and lose the old-school Unix fans. The new people won't know any better because "computers never work" has become the motto of the industry. So we will have gained nothing in the transfer from Windows to Linux/Unix (if it happened at all.) Or, people will be sick of having all the incompatibilities of Windows, plus having everything run slower because it is emulated, thus, people run back to Windows screaming "UNIX SUCKS!"

I just don't see it as a good thing. As the poll says "-1 Flamebait".

--------
Faulty Dreams - Barking at the moon 24/7...

If you think I'm an asshole, it's only because you haven't realized what a fucking idiot I am. - Faulty Dreamer

WINE is also a porting tool (none / 0) (#15)
by regeya on Fri Mar 23, 2001 at 01:34:28 PM EST

WINE allows Win32 binaries to run under Unix-y systems (again, the obligatory screenshot) but it's also a porting tool, which is I think what Corel was helping work towards as well. At one point they made a vaporware promise, at least, of making PerfectOffice 100% native in Linux using libWINE. :-)

And really, afaik that's how it's done with some software. When IE was ported to Solaris it was done using a commercial Win32 API. Ditto for the Microsoft line of software on Macs. If nothing else, the existence of a free version of what Microsoft and others pay for makes not porting inexcusable.

[ yokelpunk | kuro5hin diary ]
[ Parent ]

Let me try again. (5.00 / 1) (#16)
by Faulty Dreamer on Fri Mar 23, 2001 at 01:44:25 PM EST

You are not talking about porting. Running Win32 on top of Unix is not porting. I was one of the unfortunate idiots that actually purchased some of the WINE running software that has been available up to this point (thinking it would show support to "smart" companies). They do not run well. They are slow, they are buggy, and they conflict with eachother.

In short I believe you are very, very confused about what "native software" really means. Running Win32 on top of Unix is a bad, bad idea. I doubt I am alone when I say I do not want to see Win32 duplicated on top of Unix. I cannot believe there are people out there that actually believe it is a good idea. Win32 is a mess on its native platform. How is adding in another layer of software beneath Win32 going to make it better?

--------
Faulty Dreams - Barking at the moon 24/7...

If you think I'm an asshole, it's only because you haven't realized what a fucking idiot I am. - Faulty Dreamer
[ Parent ]

I'll try again too. (none / 0) (#17)
by regeya on Fri Mar 23, 2001 at 01:56:59 PM EST

Here's a link to help you understand where I'm coming from:

http://wine.codeweavers.com/docs/winelib-user/winelib-introduction.shtml#AEN6

WINE is, indeed, for running Win32 executables. What I was really talking about was Winelib, which allows you to compile Windows programs on UNIX. And really...

Running Win32 on top of Unix is a bad, bad idea.

Couldn't agree more. Haven't done it since I tried VMWare.

[ yokelpunk | kuro5hin diary ]
[ Parent ]

I've seen it before. (none / 0) (#20)
by Faulty Dreamer on Fri Mar 23, 2001 at 02:32:53 PM EST

And I remain un-impressed. Winelib simply takes the sloppy coding style of Windows and makes it into a sloppy coding style on top of Unix. It is still a bad idea, and still sucks.

I'm sorry, no matter how many times you try you will not convince me that we need to have Win32 programs on Unix. Through Winelib, or through WINE itself, both are bad options.

Have fun with your delusions though.

--------
Faulty Dreams - Barking at the moon 24/7...

If you think I'm an asshole, it's only because you haven't realized what a fucking idiot I am. - Faulty Dreamer
[ Parent ]

Have fun with yours, too. (none / 0) (#22)
by regeya on Fri Mar 23, 2001 at 03:37:31 PM EST

So the answer is...what? Convince companies to go through painful ports from Win32/MacOS to, again, what, GTK/GNOME? As if GNOME isn't hackish (no offense, Inoshiro...) Or maybe we should just use all Free Software? Y'know, it's not laziness that's keeping RGB->CMYK out of The GIMP, but patent problems. If you can convince Adobe to port Photoshop to Linux/other UN*X systems without using a porting tool, good luck...you're going to need it.

Or perhaps we should just give up and just use Windows. Yeah, have fun with your delusions...

[ yokelpunk | kuro5hin diary ]
[ Parent ]

There are alternatives. (none / 0) (#23)
by Faulty Dreamer on Fri Mar 23, 2001 at 03:54:41 PM EST

There are always mid-line alternatives to any discussion. The problem is that I typically reside somehwere in this mid-line and am seen by both sides to be a hard-nose idiot.

Fine and dandy. My take has always been, and always will be, that the end goal should be native code. Personally, I would say FUCK GNOME! It has never shown any value to me. It is bloated and slow and basically pretty much a "OOOO, KDE doesn't have the RIGHT license" attempt to glom on.

Now that I've pissed off half of the people on this board, and got the other half scratching there heads wondering what I'm on, the fact is that there is a big difference between thinking of using a porting tool and seeing that porting tool as a permanent solution. I don't want to see a permanent conversion of Win32 to Unix. I would much rather see companies eventually put a little effort into making real native ports. I realize there would be a generation or two of complete and utter shit (like Corel Word Perfect Office 2000, and they used to have a native coded suite), but they should eventually be able to move forward with real Unix code. Your article made it sound as if you were saying, "Forget porting altogether. We don't need ports. We need WINE!" That is why WINE is a dangerously stupid idea. Too many people actually believe that it should be, and will be, the final solution to the desktop dilema. I disagree whole-heartedly to that idea, and I still think your article smells of it.

Delusional? Yeah, I probably am. Every once in a while I think that people will start making sense. I guess I should learn better.

--------
Faulty Dreams - Barking at the moon 24/7...

If you think I'm an asshole, it's only because you haven't realized what a fucking idiot I am. - Faulty Dreamer
[ Parent ]

Positive loops (none / 0) (#35)
by hardburn on Sun Mar 25, 2001 at 07:11:05 PM EST

Lets get some obvious facts out of the way: The world runs on Windows. The desktops are Windows, the programs are Windows.

All things being equal, nobody can run binaries built to run on any other operating system. Therefore, if one OS gains a sizeable market share, the others will lose ground at an increasingly exponetial rate. Over the long term, the big OS would eventualy have 100% market share, simply because you can't do anything with any other OS because there are no programs for it. In short, you can't move to another OS because all the programs are on the majority OS, and software makers won't port their programs because there is nobody using those other OSes.

Now lets say that another OS can run the binaries from the majority OS, and can do it reliably (admittedly, WINE is not up to that point yet). People can now actualy choose which OS they want to run. People will then decide which OS to use based on how good each OS really is.

If the little OS is particularly good, then people will start moving over, despite the fact that the majority of programs can "only" run on the other one. When the little OS reaches some critical mass, the software publishers say "Everyone is using our product on this now on an 'emulated' system. Would it not be even better for our customers to get our program to work natively on that system?" And so they go off an make a native version.

Eventualy, the merits of the (no longer) little OS overtake the big one. More and more programs are running natively on the "little one" and there are even a whole bunch of programs that only work on that one. The now-fallen big OS maker could try to make the API emulation that the little one had (only in the opposite direction), but they are way behind. Besides, everyone ditched them because they sucked, remember? The "big bad monopoly" dies. Hard.

I view WINE as a stepping stone to getting native programs on GNU/Linux systems. Sure, it will suck a little bit from having a not-quite-native program running, but thats only a short term problem that will make itself into a long term solution.


----
while($story = K5::Story->new()) { $story->vote(-1) if($story->section() == $POLITICS); }


[ Parent ]
You're assuming to much. (none / 0) (#38)
by Faulty Dreamer on Mon Mar 26, 2001 at 09:00:37 AM EST

I realize this is not going to be taken well by most Linux fans, but you are assuming way too many things.

You are assuming:
1. People care what operating system they use. Aside from a few of us "in the know" types, most people couldn't care less. To most people this is a non-decision.

2. Software makers care what their customers want. *SUPRISE* They Don't. Once they get used to the "easy" way of "porting", they are not going to be easily persuaded to go in a different direction, no matter the benefits to their customers.

Why go on? If the industry comes up with a workable kludge they will cling to it as if it were a life-raft in the middle of the ocean. If WINE gets a solid foothold, you can forget about native apps. I wish I was wrong, but history shows what our industry shoots for: software bloat to promote faster hardware. Given the opportunity to continue this trend, I really doubt anyone will diverge from the common path.

--------
Faulty Dreams - Barking at the moon 24/7...

If you think I'm an asshole, it's only because you haven't realized what a fucking idiot I am. - Faulty Dreamer
[ Parent ]

Is Wine always the best solution? (4.60 / 5) (#6)
by Anonymous 242 on Fri Mar 23, 2001 at 12:13:05 PM EST

Can it really be argued that DB/2 or Oracle on Linux would have been satisfactory in running the Windows versions under Wine?

I'm not among those that think that Win-OS/2 hindered Big Blue's attempt to unseat Windows from being King of the desktop. I place the blame for that squarely on Big Blue's rather bizarre marketing and lackluster education programs combined with the reluctance of PC manufacturers to preload an OS written by a competitor.

That said, I really have to wonder if Wine is very much of a big deal. Lack of applications are a very large reason many people don't consider switching to Linux. I think the state of the X (and KDE and Gnome) UI and lack of preloads are even bigger reasons. Until the a very large chunk of PCs come with Linux preloaded and until the X UI becomes more accomodating to the casual user, it doesn't really matter how well Wine runs Windows programs because few people will care.

The other thing that native ports will buy that Wine won't is native compound documents. This is a potentially huge deal. Unless the Wine folks figure out a way to use native bonobo or kpart pieces, Wine apps will be second class citizens. Windows strength is that it offers a way to integrate desktop apps. Unless Wine can integrate Windoes apps at the component level, it is not (in my mind) a viable solution.

Lastly, porting is a good exercise for most companies to go through. If one splits the UI from the back end properly, porting is relatively easy. Porting is only a major chore for poorly written apps. Kludges seldom port very well. Such a design will help companies in ways other than porting from one UI to another. For example, many desktop apps could make their products into one type of server app or the other once they can split the UI from the back end.

WINE without cheese (5.00 / 1) (#7)
by spaceghoti on Fri Mar 23, 2001 at 12:20:47 PM EST

While I'm not holding my breath for a reasonably accurate and compatible version of WINE to appear on the digital horizon, I would go to WINE over Windows in a heartbeat if I thought it would work. I got started in Unix and prefer it over everything else I've tried (X-windows rock my world), but the applications that come with Unix aren't as choice. The job market I've broken into prefers Windows (mostly NT) so my expertise has focused there.

If I could combine the best of both worlds with WINE, I'd do it. Unless Microsoft gets beaten into releasing their source code or someone cracks them to make WINE work properly, I'll wait.



"Humor. It is a difficult concept. It is not logical." -Saavik, ST: Wrath of Khan

Personally... (1.75 / 4) (#9)
by wiredog on Fri Mar 23, 2001 at 12:53:50 PM EST

I like something stronger than wine.

The idea of a global village is wrong, it's more like a gazillion pub bars.
Phage

Port vs. WINE (none / 0) (#14)
by Farq Q. Fenderson on Fri Mar 23, 2001 at 01:32:44 PM EST

While WINE is a Good Thing, I don't use it. Many people don't use it. This means that those who do not make use of WINE cannot use the program.

In addition, and more importantly, a port of the program generally involves retargetting of features, which means that the ported copy integrates with the new target OS much more nicely.

For example, text files produced by a program operating under WINE will likely produce <CR><LF> for newlines, instead of the <LF> used on Unix systems.

This is just a trivial difference, since there's lots of DOS-to-Unix conversion programs. The point is, however, that there's another step in system integration. Some other incongruent features, I expect, would be much more of a hassle to deal with.

Finally, most systems have features that other systems don't, and it's possible that the use of those features could be advantageous to the program's design.

In short, a WINE program will not adhere to the conventions of the OS that WINE is running on.


farq will not be coming back
Does WINE really work? (4.33 / 3) (#18)
by SIGFPE on Fri Mar 23, 2001 at 01:59:45 PM EST

Now I'm going to be accused of trolling but I don't care. Wine doesn't work well enough to be useful. People who claim it does are lying (and I don't mean mistaken - they are either deluded or deliberately lying IMHO). Every time someone has recommended it to me over the last two or three years I have tried it. I can easily produce impressive looking screenshots (the main form of evidence people present in favour of WINE) but it's not stable. Do anything useful and it breaks. I and colleagues at work have tried on many different machines with poor results every time. (Hey, maybe I'm crap at installing Wine, but given my many years experience of working with Linux and Windows I think that's a comment about the Wine documentation rather than me.) Hmmm...I did manage to get one non-trivial program to run for extended periods (30 minutes or so) without crashing - Starcraft. Starcraft ran pretty damned fast too. Don't misunderstand me: I'm really impressed with WINE and the authors deserve much kudos - but I just don't think it's actually usable in a real workplace.

Unless it's improved dramatically over the last few weeks that is.

VMware, on the other hand, is truly awesome and usable, if a little slow on an underpowered machine. (I don't work for them BTW)
SIGFPE

No, it doesn't--the point is that it could. (none / 0) (#26)
by Moss Collum on Fri Mar 23, 2001 at 07:46:12 PM EST

No, WINE doesn't work well enough to be useful. That's why this article suggested that we should work on it until it is useful.

(Mind you, I do use WINE for some things--it even runs Macromedia Flash--but it's incomplete enough that every program I've run under it has its share of annoying little issues.)

This is a .sig.
Now there are two of them.
There are two _____.


[ Parent ]
I should have finished my point... (none / 0) (#29)
by SIGFPE on Fri Mar 23, 2001 at 10:10:32 PM EST

I don't think it's a slur on the developers who have done a fantastic job - they have to follow a very large moving target. I can't imagine anything other than a large consortium or corporation actually succeeding. I think it would take a few lifetimes just to read the Win32 documentation - let alone reimplement it!
SIGFPE
[ Parent ]
You're right (none / 0) (#30)
by DeadBaby on Sat Mar 24, 2001 at 01:27:19 AM EST

I've never gotten it to work either. The people who suggest <insert app here> will probably work fine under WINE are lying on purpose to lure people intro trying Linux.
"Our planet is a lonely speck in the great enveloping cosmic dark. In our obscurity -- in all this vastness -- there is no hint that help will come from elsewhere to save us from ourselves. It is up to us." - Carl Sagan
[ Parent ]
VMWare costs $420 (none / 0) (#43)
by pin0cchio on Mon Mar 26, 2001 at 11:39:09 AM EST

VMware, on the other hand, is truly awesome and usable, if a little slow on an underpowered machine.

Except VMWare for Linux costs $100 and Windows costs $320. $420 is four months' disposable income for many of us.


lj65
[ Parent ]
If... (none / 0) (#19)
by Zeram on Fri Mar 23, 2001 at 02:14:13 PM EST

Linux could get down like *BSD then maybe you'd have an arguement. But there is always going to be something that doesn't work with Wine for what ever reason. So get over it, native code is better, ferw places to look to find the cause of a problem, and generally better stability.
<----^---->
Like Anime? In the Philly metro area? Welcome to the machine...
What would be the point then? (3.00 / 2) (#21)
by Anonymous 6522 on Fri Mar 23, 2001 at 02:39:39 PM EST

If everyone who uses Linux is just running everything in WINE, why use it is the first place? What would be the point? And what's stopping MS from changing their APIs so that what runs in WINE won't run on Windows and vice versa?

APIs (5.00 / 1) (#28)
by Zer0 on Fri Mar 23, 2001 at 09:12:05 PM EST

And what's stopping MS from changing their APIs so that what runs in WINE won't run on Windows and vice versa?

The fact that changing the APIs would break all the existing programs using them :). What Microsoft will continue to do is add new APIs, which is a good thing. More APIs, new controls etc, means there is less work for me.

[ Parent ]

Just wondering (3.00 / 1) (#24)
by FyreFiend on Fri Mar 23, 2001 at 03:56:52 PM EST

(disclamer: I'n not a programmer so I probably don't know what the hell I'm talking about).

From what I understand of Wine it tries to translate the the system calls between the windows program and the underlying OS. So, instead of trying to get companies to port their programs or have their programmers work on wine why not try to get the companies to let the wine developers pick their programmer's brains a bit.
Something along the lines of, "When your program does $foo it crashes wine. Can you give us some insite as to what your program is doing there so we can get wine work with it?"

Just a thought.

--
Only kings, presidents, editors, and people with tapeworms have the right to use the editorial "we".
-- Mark Twain


Yeah, but ... (4.00 / 1) (#27)
by lucas on Fri Mar 23, 2001 at 08:20:04 PM EST

WINE is nice, but it's not an entire solution that I would feel confident recommending... particularly because it is based on an API that Microsoft controls.. and Microsoft is not exactly friendly on attempts to infringe on its proprietary technologies.

Quite honestly, some of the apps written for MS-Windows really suck anyway. I would rather use a GNOME/KDE app than the Windows equivalent... particularly as MS-Windows downloadable apps are often nagware and/or shareware that is crippled.

I'd also say it would be good to let new, independent programmers create new solutions for the platform instead of importing Windows juggernauts.

Win32 is an ECMA standard (none / 0) (#42)
by pin0cchio on Mon Mar 26, 2001 at 11:33:38 AM EST

particularly because it is based on an API that Microsoft controls

ECMA, the people who standardized JavaScript, have also standardized Win32 as ECMA-234.

and Microsoft is not exactly friendly on attempts to infringe on its proprietary technologies

So why did Microsoft submit Win32 to ECMA for standardization?


lj65
[ Parent ]
Win3.1 is an EMCA standard (none / 0) (#44)
by kallisti on Mon Mar 26, 2001 at 02:30:06 PM EST

The link you gave points to documents defining the Windows 3.1 interface. The date given is December 1995, so I would guess this is an example of making "standard" something that is no longer of any use, as Microsoft ditched 16bit Windows rather rapidly.

[ Parent ]
I object ... (4.00 / 1) (#31)
by strepsil on Sat Mar 24, 2001 at 06:16:59 AM EST

(just look at Mandrake 7.2, which I shamefully admit to installing once)

Speaking as a very happy Mandrake user - "Hey!"

What's wrong with it? Just because they tried to make the installation and management easier for people doesn't mean it's bad. I love it. Why the hell shouldn't they do that?

I'm not clueless - I admin a bunch of IRIX, NetBSD, Tru64 and Redhat servers for a living. I do not want to be a sysadmin when I get home. I just want to hit the Unreal Tournament icon and kill some people. Mandrake gives me that ability easier than any other Unix I've ever seen.

Hmm. Now - am I off-topic or not? It was mentioned in the article, right? :)



Mandrake actually works? (none / 0) (#32)
by lavaforge on Sat Mar 24, 2001 at 07:26:02 AM EST

What's wrong with it? Just because they tried to make the installation and management easier for people doesn't mean it's bad. I love it. Why the hell shouldn't they do that?
Mine may be an isolated case, but in my experience few of Mandrake's much vaunted configuration utilities don't actually work, and I end up stuck in VI fixing things.
"In theory, there is no difference between theory and practice. But, in practice, there is." -- Jan L.A. van de Snepscheut
[ Parent ]
I have too many problems with it (none / 0) (#33)
by hardburn on Sat Mar 24, 2001 at 12:44:09 PM EST

I'm a Debian user, and thats what I use on almost all my boxes. I wanted to try Mandrake once because I wanted to know if it was good enough that a Windows user would be able to use it with minimal learning curve. I think it is nearly there, and that is nice. However, in making it like Windows, they made it really like Windows, crashes and all. Corel once had the distinction of "flakiest distro ever". Now that they're gone, Mandrake holds that title.

Within a few weeks, Mandrake crashed so hard that X stoped working. Instead of bothering to repair it, I just installed Debian.


----
while($story = K5::Story->new()) { $story->vote(-1) if($story->section() == $POLITICS); }


[ Parent ]
WINE == Bad Thing(TM) (4.00 / 1) (#34)
by der on Sat Mar 24, 2001 at 01:49:43 PM EST

Before you get your panties in a knot, my standard disclaimer: I am not a troll, or a 'zealot' and don't need replies and flames telling me that I am. I'll respect your opinions, these are mine, take 'em or leave 'em.

Now, with that out of the way.. :)

I use GNU/Linux because it's the only real feasible Free Software OS out there ATM (AFAIK anyway), not because I think it's 'neato' (although UNIX rocks you ;) ) or because M$ sux0rs (although they do). IMO, getting more proprietary software ported over to GNU/Linux (or the BSD's or whatever Free OS turns your crank) is just the process of completely degrading the whole 'point' of GNU/Linux in the first place (Freedom - Once again, IMO).

If freedom isn't the 'point' behind our oh-so-cherished OS, why not just use a proprietary UNIX? Especially when 'Linux' was immature and 'useless' (to users..) you would expect people to use a superior proprietary UNIX if they cared only about 'quality'.

So, in summary, proprietary software being ported to Free Software OSes is a Bad Thing(TM), IMO. :)

<Linda Richman>I've given you yet anothah tawpic, discuss!</Linda Richman>

Only feasible Free system? What about BSD? (none / 0) (#40)
by pin0cchio on Mon Mar 26, 2001 at 11:25:34 AM EST

I use GNU/Linux because it's the only real feasible Free Software OS out there

What about FreeBSD or NetBSD? "They don't have any apps." Most Linux apps recompile on *BSD with few to no changes. "They don't have any closed-source apps." Why would you want one in the first place? Even then, FreeBSD has a compatibility layer for Linux86 apps. "They don't have any native closed-source games." If you want a game machine, go buy a $100 Dreamcast console and connect it to your television or tuner card, or just install FreeDOS.


lj65
[ Parent ]
Bad wording on my part. (none / 0) (#46)
by der on Mon Mar 26, 2001 at 04:44:20 PM EST

I didn't really mean to discredit the (free) BSDs, as they are obviously just as valid an OS as GNU/Linux. I was trying to 'discredit' this whole "Linux is nifty" thing going on right now, with people using 'Linux' and wanting proprietary software ported etc. It was meant to supplement my "why not just use a proprietary unix" point.
Just bad wording on my part. :)

[ Parent ]
I hate *most* proprietary software (none / 0) (#48)
by hardburn on Wed Mar 28, 2001 at 03:16:25 PM EST

I don't allow any proprietary software on my GNU/Linux boxes. I make one exception: Games. Free Software games are perfectly fine, but I have no particular problems running non-Free games. Thats why I like WINE and TransGaming; not because I want proprietary Windows programs on GNU/Linux, but because I want to play Windows games on GNU/Linux *now*.


----
while($story = K5::Story->new()) { $story->vote(-1) if($story->section() == $POLITICS); }


[ Parent ]
You discount the amount of free software for Win32 (none / 0) (#41)
by pin0cchio on Mon Mar 26, 2001 at 11:29:32 AM EST

IMO, getting more proprietary software ported over to GNU/Linux (or the BSD's or whatever Free OS turns your crank) is just the process of completely degrading the whole 'point' of GNU/Linux in the first place (Freedom - Once again, IMO).

What about all the Free software written in Visual Basic? What about all the Free software written in C for straight Win32 because there aren't any free(beer) X11 servers for Windows 9x?


lj65
[ Parent ]
... What about it? (none / 0) (#45)
by der on Mon Mar 26, 2001 at 04:38:41 PM EST

Subject says it all :)

I don't see your point.

[ Parent ]
Windows software is "different" (none / 0) (#36)
by MoxFulder on Sun Mar 25, 2001 at 08:50:54 PM EST

Although I think that WINE is kind of a cool "toy", I don't think I would ever want to seriously use Windows programs under Linux. It's not that I am a free software purist, and I'm not worried about stability or performance issues, it's simply that Windows programs don't fit in with the Linux "style" of computing (at least as I perceive it!).

Windows programs tend to be monolithic rather than built out of small components. They usually cannot interoperate with other programs very well, and forget about putting them into a script.

I think this is why I don't like Windows programs (and they especially grate on my nerves when I have well-made native Linux apps onscreen at the same time). The few things I do like about Windows are the few components that work well with lots of different programs, like the international character set support.

"If good things lasted forever, would we realize how special they are?"
--Calvin and Hobbes


OS/2 Syndrome (4.50 / 2) (#37)
by yannick on Mon Mar 26, 2001 at 03:19:40 AM EST

Putting too much emphasis on systems like WINE will lead to the same problem IBM faced with OS/2. Waaaay back, when OS/2 almost had a chance against Windows, IBM realised that one of the best ways to lure people over was to provide support for Win32 applications under OS/2. They argued that OS/2 users would be able to run native applications as well as Win32 programs, so they'd get the best of both worlds.

This ended-up backfiring, as software developers began to write software exclusively for Win32 systems, knowing that OS/2 users would also be able to run their stuff.

In the end, OS/2 died a slow, painful death. But the lesson the Linux community should take away is clear: put too much emphasis on being able to run Windows software under Linux, and people will start to wonder why the shouldn't just run Windows. Developers will stop writing native Linux apps (which are preferable to programs running under emulation), and Linux will lack the software base that it has worked so hard to establish.

Cheers,
Yannick
------
"Myself when young did eagerly frequent / Doctor and Saint, and heard great Argument / About it and about: but evermore / Came out by the same Door as in I went." -- Omar Khayyam

Reverse is true, too (none / 0) (#39)
by hardburn on Mon Mar 26, 2001 at 10:33:17 AM EST

OTOH, OS/2 and NT actualy have the same core. It was orgianly to be a joint project, but a split between Microsoft and IBM caused a split. Microsoft took what they had and made NT, IBM took what they had and continued to call it OS/2. NT is actualy compatible with all the old OS/2 programs.


----
while($story = K5::Story->new()) { $story->vote(-1) if($story->section() == $POLITICS); }


[ Parent ]
The Win-OS/2 Syndrome is a myth (none / 0) (#47)
by Anonymous 242 on Tue Mar 27, 2001 at 04:10:54 PM EST

The reasons OS/2 failed:
  1. Lack of education. IBM either could not or would not educate people to understand that Windows was a replaceable element on the IBM PC and compatible platform. In the minds of most people (including CIOs) a Windows PC and an IBM compatible PC were synonymous terms.
  2. Brain dead marketing. IBM's big chance at cracking the market was Winter 1994/Spring 1995 when Windows 95 was late, later and still later. Instead of campaign that made people want to use OS/2 billboard ads with slogans like "OS/2 obliterates my software" were being run.
  3. The preload situation. Honestly, if you were Compaq, Dell or Gateway would you preload an OS supplied by your competitor in the razor thin PC hardware market? Preloads are really the heart of the game. How many people would be continuing to argue Windows was easy to use if one had to install it one's self?
  4. The PowerPC gamble. The resources that ought to have gone into OS/2 v4 went into OS/2 PowerPC. When the PowerPC chip failed to take off as an x86 replacement, it took any existing momentum out of OS/2. IBM effectively put all of its OS/2 eggs in one basket and lost them all.
  5. The abandonment of the SOHO user with v4. IBM arguably still had a window of opportunity and fudged it by ignoring (1) superficial issues that are important to the end user such as decent anti-aliased fonts and (2) underlying core design issues such as the synchronous input queue.
  6. Hardware support. IBM failed to realize the importance of consumer hardware support until very late in the game.
  7. Internet support. OS/2 v3 had very problematic PPP support out of the box. If v3 had had decent PPP and if IBM had been able to convince Netscape to port Navigator to OS/2 much, much sooner, Windows might very well have been trampled. A hypothetical OS/2 with good PPP support bundled with Netscape Navigator would have at minimum kicked the heels out from under Windows 3.1. It also might very well have unseated Windows 95. And if IBM had gone one further and convinced AOL to jump on board, OS/2 might very well have been unstoppable.
There were plenty of great programs available for OS/2 in the hey day of v3. The problem is the ever decreasing market left a good deal of the ISVs high and dry. It is hard to build a compelling business case to target a market either decreasing or staying static when another market is growing almost exponentially. A computer only needs one copy of a program. Should an ISV target OS/2 that has few new users or Windows that has a market that grows with each sale of a new PC?

There quite a bit that the Linux community can learn from IBM's OS/2 debacle. Very little of it has anything to do with issues surrounding WINE.

[ Parent ]

Why port? | 48 comments (45 topical, 3 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!