Slashdot Mirror
On Microsoft Porting to Linux/Unix
skubalon writes "Mainsoft confirmed today that they are indeed porting Microsoft's apps to Linux. The story was first reported in Paul Thurrott's WinInfo yesterday. Internet Explorer and Windows Media Player will be among the first apps to be ported." On the other hand we have this submission:rendell writes "According to this story on Beta News, Microsoft is denying the rumors that they were in the process of porting some of their software to the Linux platform. Especially focusing on the rumor that the main project was Office." So - who knows? My analysis: Microsoft is correct - /they/ are not doing the porting. Mainsoft however, appears to have the powers to execute the porting.
I found this unfinished press release poking blindly around the Mainsoft press release directory. Enjoy!
/etc directory and spread around the file system.
Mainsoft announces major breakthrough in porting Windows functionality to Linux
San Jose, Calif - April 1, 2001 - Mainsoft, the leader in common code-base cross-platform solutions for the enterprise, today announced a major breakthrough in porting Microsoft Windows functionality to the Linux Operating System.
The first breakthrough is a patch to the linux kernel, kernmem.sys, allowing any linux machine to display important debugging information any time an application needs to. This is an important first step to porting other Windows applications, as it allows any user to immediately see when an application has requested the system to be rebooted.
"Windows has had this functionality for a long time." said Yaacov Cohen, president of Mainsoft. "In the Windows world, this feature is known as the Blue Screen of Debugging". He continued, "We feel this has been a missing feature of Linux since its beginning, and will allow normal users of Windows to feel comfortable that Linux now behaves like a real Operating System."
The second breakthrough is the porting of the registry to Linux. The registry will obviate the need for hundreds of configuration files in the
"With a working registry in Linux, we can replace all those antiquated file and user permissions and SUID bits that clutter up the Linux system." said Miguel De Icaza, Linux pundit "Now every Linux system will be as secure as a Windows machine, allowing consumers to feel safer about automatic registration and other new Microsoft technologies."
The largest breakthrough is the porting of the Microsoft Entertainment Pack to Linux. Containing the most widely used applications on computers today, the Entertainment Pack will bring Solitaire, Minesweeper, and Yambo to Linux.
"These productivity applications account for more than 50% of all CPU time used on Windows machines" said Rob Malda, Chief Productivity Application Tester at slashdot.com. "With these applications now available on all Microsoft Linux compatible distributions, productivity will soar".
Future enhancements to Linux will include the return to the Single User - Single Machine philosophy which fueled the explosive growth of the PC market in the 1980's. Removing the ability to have more than one user logged into a machine at any single time makes more efficient use of the resources of that machine, and simplifies licensing of future applications under the M$GPL.
the AC
Hemos is like...sci-fi fans;he thinks technology is cool, but he hasn't bothered to understand the science it's based on
Frankly, I wouldn't be surprised if MS is porting Office to Linux. Will they release it when it is done, probably not. Why? Simple, they don't need to yet. MS still has a huge portion of the office market and until they think that they are actually being challenged on that front they don't need to worry about releasing their office product over to other systems. It is still good to have an ACE up the sleeve though. If they are porting Office over right now, when something like StarOffice, or if WordPerfect makes a miracle comeback and starts to eat up their market AND they can trace the loss back to Linux, then they release the product and try to kill the competing office product. Or if the DOJ gets their way and the Surpreme Court (SC) hears the case, AND if the SC decides that they do need to be broken up then the non-OS portion of MS can release the product and make more money on OSes that compete with Windows.
It is smart for them to port. It is smarter of them to allow someone else to port for them. And it is even smarter to hold back that which could potentially hurt them until it can help them.
Disclamer - Opinion of Person
Regardless of who is doing the porting, Microsoft apps for Linux sounds a little shady. Think about it: MS has got a pretty good stranglehold on the desktop market, and one of thier primary up-and-coming competitors is Linux. Now, does it sound like a terribly astute business decision to port your applications (which are what lock users to your OS) to your competitors' platform?
The only way I can see this making any sense is if MS has resigned themselved to being split into MS/OS and MS/Applications as per the initial DOJ v. MS ruling.
What are we missing? What am I missing?
--
We may not imagine how our lives could be more frustrating and complex—but Congress can. – Cullen Hightower
Mainsoft makes a product called MainWin, which is an implementation of the Win32 API. They aren't actually modifying Microsoft product code to port it to Linux, they are just creating a compatible API layer and working to make it compatible enough that major Microsoft apps will run on Linux. It's no different than WINe.
- "It's just a matter of opinion!" - PRIMUS
This is a major contract for them, but I have my doubts on their ability to produce. First, it won't be done in the next few years. I mean, you don't just port apps as complex as what Microsoft puts out to another *platform* in a few weeks. I also have to wonder about the quality once it will be done - if they "do it right", a proper port could take 3-4 years.. which means whatever they're porting now could very well be out of date. They'll need continual contact with MS to keep their code up to date.. that's alot of extra work for both of them.. adding a huge cost.
Finally, what if it fails - what if a bug-ridden copy of MS Office ships under the linux banner? MS people will *OF COURSE* run to get the first copy, first release, of it. And they'll be very disappointed with what they see. They won't see this as a failure of Microsoft, who has provided a perfectly working copy under Windows, but as a failure of Linux for not supporting their favorite app "the way they wanted it to".
Assume for a moment that someone with the Win32 and Office source code is porting the office apps to Linux (and pretty much only Linux86, maybe LinuxAlpha if you're lucky, but you get the same horrible 32 bit kludging you do in NTAlpha). Assume it's Mainsoft. They're still doing it with Microsoft's permission, and M$, unlike IBM of old, is a company with a vision. The Office group does not crap on the OS group.
If you or I were given all this source code and told "port Office to Linux" we would run away screaming in horror and hide under our beds for a year. Perhaps if we had enough intestinal fortitude we'd identify the APIs that Office uses that aren't supported by Wine and implement those. Licencing would be a bitch, of course.
This isn't what Mainsoft are doing. They're porting Office without benefit of Wine. Remember that Microsoft wants you to buy Office, but most of all they want you to buy Windows and Office.
My predictions:
Office for Linux will be slower, uglier and less stable than the Windows version;
when Office for Linux crashes it will take down the OS. Don't say it can't be done - remember that the smart guys at Microsoft figured out an unknown weakness in the Linux TCP/IP stack for the Mindcraft tests;
Microsoft will blame it on Linux and offer "competitive upgrades" to W2K.
You just see if I'm right!
--
--
E_NOSIG
This may not be so bad. Personally, I like IE over Netscape. It's stable (on my system, at least) and it's more polished. I prefer to use IE when I'm in Windows over Netscape. I wonder how they can make an IE port to *nix stable though. What about all of the secret API code that's used in Windows? Once it's ported, does that mean the *nix version will have less functionality?
And Media Player is equally good news. There are many times I've had to reboot into Windows if I needed to see a particular video.
No flames please, I'm just being honest.
This will be interesting. The thing is, while there are still better apps, Microsoft's non-Windows offerings are actually respectable. Still not the best out there, mind you, but still leaps and bounds better than their Windows counterparts.
Take Internet Explorer, for example (before I begin, let me state for the record that I'm a Mozilla guy myself). We all know about the latest round of Embracing And Extending done in IE5.5/Windows. However, currently IE5/Mac is quite nice. It's very close to full CSS1 compliance (it's not quite there, as they claim, but the only non-compliance I've seen yet is thay they don't support all the border types around boxes). The standards-compliance is second only to Mozilla's. It's fairly stable (it crashes me more than Netscape does, but not as much as the latest Mozilla builds; then again Mozilla is still in-development software so that's forgivable).
Even Office is actually usable on MacOS (and I use Office/Windows quite often, so I'm familiar with the usability nightmares on that platform), though I still prefer AppleWorks. OE's not bad, though I prefer PowerMail, and so on and so forth.
Why is this the case? You'd think that if the lowly porting teams could make decent software, the mammoth WIndows development teams could do even better. My personal view is that the ports are better because they don't have an OS monopoly to leverage, so they know they have to make software that would actually stand a chance in a competition based on merit. Contrast this with the Windows development team, which we already know makes a poor product (I work with developing Windows every day, and there are parts of that OS that I could have done designed better back in high school, and implemented not long afterward).
So the ports to Linux just might work out. I'm willing to withhold judgement until I see the stuff, anyway. I doubt that Microsoft is trying to sabotage Linux; they're merely hedging their bets. Office is their major moneymaker, and if Windows dies (and I doubt it's got ten years left, not after the DOJ rulling) they need to have Office on whichever platform wins the ensuing chaos. As MacOS and Linux are the two biggest contenders, they need Office on both, just in case.
----------
why Office 98 for the mac is one of the best programs that Microsoft makes!
Having used MainWin to port an relatively large (300,000 lines) Windows application to UNIX, MainWin's Win32 API implementation at the time was pretty damn lacking. Actually, I found it to be a bit of a pain in the ass.
... else ... endif preprocessor mess throughout "lower-level" parts of the code. Yeah, MainWin made the task a lot easier, but then again, our code was really pretty high level. We didn't really use anything that I would consider low-level, but we still had to kludge things up. Microsoft's code is pretty much known to use undocumented APIs and other such mess---that's why MainWin needs the Windows source code to even do the port. If we had a bit of difficulty porting our Windows program, I can't even imagine the nightmare of porting a Microsoft application.
It's been two years since I've used it (I left that company) but some things that stick out in my mind were that the resource files were somewhat different under MainWin. (We wrote a pretty crazy Perl script that modified the Windows resource file and covered up those differences.
I also remember lots of ifdef UNIX
But in the end, you really should have separated your user interface from the rest of your code. Then, doing a port is just a matter of hiring some people to make a new GUI for each platform. That's (usually) not so difficult. Motif (Gtk+, Qt, whatever) code for UNIX, Win32 for Windows, MacOS for Mac. Of course, Microsoft certainly didn't consider cross-platform code when they wrote (or bought) the Office products.
Given that Microsoft will eventually start porting apps, the question becomes: does this really benefit the *x community?
Not particularly. The apps will be large and unwieldy, and X doesn't fare well running large and unwieldy applications. See: Netscape 4.x
Sure, Linux users will be able to open up Word documents and Excel spreadsheets.
Which means the situation will remain the same. See: StarOffice, Gnumeric, KOffice, AbiWord(?)
Sure, Solaris users will be able to use Windows Media Player.
This is probably the only major impact, in terms of what isn't already available for Unix. Lots of websites are using Windows Media exclusively for their streaming, which is unfortunate for anybody that doesn't have Win9x. So getting more exposure for the player would be good. Maybe it'll be better than RealPlayer. Eh.
But The OS is where the money is.
Not even slightly. Have you gone out to a software store recently and compared the price of Win98 to the price of the various distros of MS Office? MSOffice has, for the last several years, been the single most profitable application for MS as far as end-users is concerned.
Why would Microsoft port enough applications for Linux to become viable as a end-user desktop?
This will happen with or without them, and they should be smart enough to realize that by now. Maybe they decided to give in and get their foot in the door before Linux becomes a viable end-user desktop without them.
I reason that they wouldn't. Hell, they might even keep the releases on *x one step (in features and bug fixes) behind the Windows releases.
They'll probably do that anyway. Most commercial vendors that maintain Linux ports do. See: Corel WordPerfect, Borland whatever-that-program-was, etc.
Is it just me? Microsoft in no place seems to be denying the port of Internet Explorer and their Media Player. What they're saying is that office will not be ported to Linux.
So while the rumors were in part wrong, we have on one side Mainsoft saying that they are not porting office but IE and Media player and on the other side MS only saying that they are not porting Office...
Microsoft did not deny the possibility that some of its apps would be ported, they only said "WE are not doing it" and "not Office".
"If liberty means anything at all, it means the right to tell people what they do not want to hear"
This looks interesting.
With this demonstration application you'll see, first-hand the native Linux performance with full Windows functionality available only by using MainWin. We want to take this opportunity to thank Microsoft Corporation for providing the ORIGINAL source code of FreeCell for Windows game.
New Amiga-based PDA announced.
Microsoft Funds "KDE Foundation".
Linus Torvalds admits: "I run Win2K".
Pat Buchanan to nix Copyright Law: Geeks vote for Pat!
UI Expert proves: Vi is better than Emacs.
Finally free: the Linux Kernel is BSD'ed
etc, etc, et-freaking-cetera...
engineers never lie; we just approximate the truth.
Look at Office 2001 for the mac, you can see what I mean. Mircosoft knows Mac users dont' like Microsoft. So they are downplaying everything that says Mircosoft. It's totally carbonized, etc.
I'd expect the same with a linux port. I'd assume the first version would be just a MainWin "port". But then a linux ap division. Which I hope is as talented as their MacOS divsion. Thier 2nd offering of Microsoft products would use all gnome services, bonobo, corba, gtk. etc.
And for all you doubting thomas's. Interenet Exploer 5 for MacOS is the most standards compiant browser on the market. I'd execpt no less form Microsoft on a linux port.
Regardless of who is doing the porting, Microsoft apps for Linux sounds a little shady. Think about it: MS has got a pretty good stranglehold on the desktop market, and one of thier primary up-and-coming competitors is Linux.
.NET all this would become clearer to you.
.NET family exist on every platform can only be a plus.
If you had focussed on any of the articles on MSFT's
In MSFT's vision of the future, all apps are hosted on the server and rented by clients. To do this clients will need browsers, audio and video players, libraries, etc that can view MSFT proprietary content. Since most of these hosted apps will use client side scripting and advanced DHTML/XML techniques making sure that browsers that can access all the
Who cares if a few of the desktops run Linux? MSFT's major money makers have always been the Office line of products. If browsers are provided so that Linux users can now buy office licensees then this can only improve their bottom line.
The Queue Principle
I'm sure they could get the real specs to the WINE crew and get it finished in a short time.