Bill Gates Takes the Stand In WordPerfect Trial 472
Hugh Pickens writes "Remember WorldPerfect? Bill Gates took the witness stand to defend his company against a $1 billion antitrust lawsuit that claims Microsoft duped Novell into thinking he would include WordPerfect in the new Windows system, then backed out because he feared it was too good. Gates testified Monday that Microsoft was racing to put out Windows 95 when he dropped technical features that would no longer support the rival's word processor. He said that in making the decision about the code, he was concerned not about Novell but about one element of the program that could have caused computers to crash. That code, technically known as 'name space extensions,' had to do with the display of folders and files. Novell attorney Jeff Johnson concedes that Microsoft was under no legal obligation to provide advance access to Windows 95 so Novell could prepare a compatible version but contends that Microsoft enticed Novell to work on a version, only to withdraw support months before Windows 95 hit the market. 'We got stabbed in the back.'"
Groklaw has a pretty good article. (Score:5, Informative)
Wrong summary!!! (Score:4, Informative)
"into thinking he would include WordPerfect in the new Windows system" This is WRONG! Novell thought Windows would include some (4) APIs about "name space extensions".
Re:Groklaw has a pretty good article. (Score:5, Informative)
You're comparing Windows 95 to Linux distros of the same era?
Windows 95 was infamous for crashing at least daily, I knew plenty of fairly knowledgeable people who took pride in being able to keep it running for a week. While it was, in theory, capable of multitasking the truth was that very few users would gamble with multitasking under Win9x (except for things like running an IRC client, an MP3 player and a web browser at the same time). Why? Because it crashed fast and hard for seemingly no reason (a single program crashing often brought the whole system down in various fun ways, the common pattern being that a program crashed and you scrambled to save everything you were using in other programs, with alternate filenames of course, just in case the crashing program had corrupted something, within a minute or two the system would bluescreen as you did something like click the Start menu button).
By comparison Linux at the time was rock solid. Yes, both Windows and Linux are more stable than the Linux distros of that time but even Red Hat 4.x and Slackware 3.x were more stable than the average desktop machine is these days.
Re:Groklaw has a pretty good article. (Score:4, Informative)
I dont remember Linux crashing.
I do remember X crashing and Samba Crashing as well as other apps crashing, I dont remember any instances where the Core Kernel Crashed. I had linux as the core of a ISP from 1994-1999 and never had it crash on me outside of apps crashing and consuming memory.
Re:Groklaw has a pretty good article. (Score:4, Informative)
It's worth reading through much more on Groklaw; this article [groklaw.net] explains that they were shell extension namespace APIs which made URL integration possible. It's pretty obvious that if WWW integration is a major new feature relied on throughout your code and Microsoft has promised to implement a large part of it, when they hide those APIs so that partners can't use them it's going to be a big problem.
Re:Novell is killing babies now? (Score:5, Informative)
Re:Groklaw has a pretty good article. (Score:5, Informative)
The sad part is that sometimes I think you do actually believe what you're saying. If you do, you just have no idea of Microsoft's history.
We need to slaughter Novell before they get stronger."
-Former Microsoft VP James Allchin in a 09-9-91 e-mail (as revealed in Caldera v. Microsoft)
"This really isn't that hard. If you're going to kill someone there isn't much reason to get all worked up about it and angry -- you just pull the trigger. Angry discussions before hand are a waste of time. We need to smile at Novell while we pull the trigger." -Former Microsoft VP James Allchin in a 09-9-91 e-mail (as revealed in Caldera v. Microsoft)
"It is Microsoft's corporate practice to pressure other firms to halt software development that either shows the potential to weaken the applications barrier to entry or competes directly with Microsoft's most cherished software products."
-Judge Thomas Penfield Jackson in the Microsoft antitrust trial
"Microsoft has demonstrated that it will use its prodigious market power and immense profits to harm any firm that insists on pursuing initiatives that could intensify competition against one of Microsoft's core products. ... The ultimate result is that some innovations that would truly benefit consumers never occur for the sole reason that they do not coincide with Microsoft's self-interest."
-Judge Thomas Penfield Jackson in the Microsoft antitrust trial
WP had poor support back in the day (Score:5, Informative)
I duno how much Microsoft really had to do with it, but it seemed like WordPerfect really screwed themselves with poor quality software and service.
The original release of WordPerfect 7 ONLY ran on Windows 95 (Not at all on under NT), was late to to release, and was not very stable. They later produced an update of WP 7 that was more stable and ran on NT 3.51/4.0 but the only way to get that was to order a new CD. No downloadable update patches for you!
WP 7 for Windows 3.1 was just a rebadged version of the 16-bit WP 6.1.
Then they pulled the same trick with WordPerfect 8. Initially buggy and updates required obtaining a new CD.
To this day there is still an option to turn off the "enhanced" open/save dialog because it is buggy and crashes under odd environments - especially under Wine.
It also didn't help that at the time it was switching ownership left and right. WordPerfect corp? Novel? Corel? Good way to destroy confidence in a product.
WordPerfect killed itself (Score:5, Informative)
One of the three members of the trio who ran WordPerfect corporation, Pete Petersen, wrote a detailed book about the WordPerfect saga called Almost Perfect [wordplace.com]. Go read it now, it's a fascinating tale of a once-great company so busy shooting itself in the foot that it hasn't noticed that it's going down the tubes. WordPerfect Corp was doing such a good job of committing suicide that it really didn't need any help from Microsoft, or anyone else for that matter.
Re:Groklaw has a pretty good article. (Score:5, Informative)
Wasn't there a Windows95 bug that would 100% crash the OS after 46 days? And it took years to find this bug because usually the OS would crash much much earlier...
49.7 days. Affected Windows 95 and 98. http://news.cnet.com/Windows-may-crash-after-49.7-days/2100-1040_3-222391.html [cnet.com]
Re:Groklaw has a pretty good article. (Score:5, Informative)
Wow. Just wow.
"very few users would gamble with multitasking under Win9x (except for things like running an IRC client, an MP3 player and a web browser at the same time). "
Sounds like actual, genuine multitasking to me. And hitting one of the soft spots, the TCP stack, pretty hard. Browsers of that era weren't much to write home about and were by themselves crash-worthy. MP3 players then were pus. mIRC was tolerable.
Just as a note, I weas running W4W 3.11, dialing into a local ISP and hitting my AOL account via TCP/IP pretty much every night. It would crash every 2-3 hours. Trumpet Winsock was all there was. Then I bought the Win95 upgrade (and a full version for a second machine running Slackware 0.9 at the time). the full version was entirely normal, but the upgrade ran without rebooting until the first patch came out, something about DUN I think. I know of no other machine that did that, not even any of my others. Scary. I was pained to reboot it, and it never ran more than a week after that. One theory was that some modules from the upgrade stayed in memory and I was running a transitional W4W driver somewhere, but that's insane.
Gates' claim that they wanted to clean up 95 and that meant leaving out file naming stuff that WP relied upon, though, is disingenuous and a lie. The same APIs were used heavily by Novell for their NetWare client in W4W, and that was a target - MS was dedicated to crushing the NetWare client. Novell kept coming back, but finally succumbed. And discussion about Gates' requirement to make APIs available to all is a lie also - it may have been a legal requirement, but it was ignored, and the Word team took full advantage of their insider access to Windows APIs. Isn't this settled fact, and one of the foundations of the now dying Justice consent decree/antitrust judgement? Really? We still discuss this?
Re:Groklaw has a pretty good article. (Score:1, Informative)
Boo fucking Hoo.
Let me put this in a way most people these days seem to understand.
"I want to fucking kill andrioid, I will make sure its dead".
-- Steve Jobs
Oh, suddenly thats ok, but microsoft as a company discussing destroying another company isn't?
Re:Groklaw has a pretty good article. (Score:5, Informative)
OS/2 was preemptive, but the GUI had a single input queue, so a misbehaving application could grind the OS to a halt. That was my biggest frustration with the OS at the time. Ironically it could multitask Windows GUI and DOS Command Line applications better than Windows or DOS could, and also better than it could multitask OS/2 apps. If you made your application multithreaded and had one thread just for processing the input queue, you'd never have that problem in OS/2. Not even IBM ever actually did this.
To this day Windows suffers from the application processing the window frame controls. An application that takes a few seconds processing will sieze up and become unmovable. Eventually windows will detect this and allow you to perform some operations on the window. This is why X11 in Linux has always felt more responsive to me. Even if the application locks up, it's not responsible for handling its window frame controls, so you can still move, shade or minimize the application with no delay.
Re:Groklaw has a pretty good article. (Score:5, Informative)
No, it's not about any chip. Win95/98 had a 32-bit time counter incremented 100 times a second. The bug finally got patched one day, so this means your box runs a fixed version.
Re:Groklaw has a pretty good article. (Score:4, Informative)
That's because the window borders and title bar are handled in user-space libraries that run in the context of the client. If the client stops pumping messages, there's no way for these parts of the window to respond. There's no separate message pump or thread for them. It's stupid, but I guess it makes it easier for apps to customize the decorations by simply handling the non-client messages instead of passing them on to DefWindowProc.
I thought crashing was a feature (Score:4, Informative)
Re:Groklaw has a pretty good article. (Score:2, Informative)
It's difficult to explain, and unless you have at least an MS in CS and several years hands on experience in OS internals it'll probably go right over your head. But anyway, here goes: it was pile of utter cunting shite.
A bug in the timer chip? (Score:3, Informative)
Is this billg talking, cause he's also able to selectively distort the historical record when it suits him too.
" Symptoms: After 49.7 days of continuous operation, your Windows-based computer may stop responding (hang).
Cause: There is a problem with the timing algorithm in the Vtdapi.vxd file.
Resolution: For Windows 95, For Windows 98" link [microsoft.com]
The windows client uses the API call GetTickCount() [vbforums.com] to get the current uptime. The value returned by that function can't be larger than 49.7 because elapsed time is stored as a DWORD value.
"Air Traffic radio control system crashes every 49.7 [theregister.co.uk] days"