Yes, this is basically it. Mac users whined mightily about getting a Windows port, so MS forked their codebase and now the Mac is always behind feature parity with the Windows version.
(IMO the UI in Word 6 was the least of its problems and probably could have been fixed without losing the cross-platformness.)
However a MSDN seat license (at $5000 or whatever) is certainly not a cost-effective way to buy MS Office. I believe tho that you can downgrade Access 2007 to 95 with the corp license, so its useful to have the media.
Why they didn't write one portable VBA engine for Windows and Mac I don't know. It sounds like it was portable at one time, long ago, back in the days of 8MB 030 Macs. It used some sort of pcode compiler and apparently they couldn't port that to modern systems. Win64 apparently has the same issue.
My guess is that they're cooking something which VBA to be hosted on the.NET runtime, and this will be used for both Win and Mac products.
USENET's killer feature is that the data is separated from the program. I can read USENET with multiple clients, as long as they adhere to the same protocol. True, but Usenet is so technically and organizationally archaic that the vast majority of people chose to ignore that killer feature and conduct their discussions on the web instead.
Had Usenet "evolved" it still might be an vigorous discussion network and not a refuge for old timers, kooks, and trolls.
While you wouldn't be able to trademark "Open", there is an organization called "The Open Group" and they'll gladly publish your open standards if you pay them.
In ~2002 I consulted at a place where they were implemeting 2-digit year fields in a new system. "Now that Y2K is over we don't have to worry about it anymore..."
The OSI did not invent the term "Open Source". Yup. The term is derived from "Open Source Intelligence", which is (simply put) stuff you can read in a newspaper. There never was any implication that you could freely modify the 'source'.
The reality is that OSI's extended definition of this "open source" caused a lot of confusion. How many stories has Slashdot has run about "available source"/"not really open source" software from Microsoft, Apple, Sun, etc etc. If OSI had clear branding that wasn't overriding an existing term this wouldn't be half the problem it is.
I guess I don't understand your POV. Just because you had to sit on it for work doesn't mean that's the right way to proceed. All these things have vendor strings and PCI IDs, Windows should be smart enough to ignore irrelevant drivers. Linux boots on all this stuff with a single kernel after all.
HP should NOT be using the same image for their Intel and AMD-based systems. Why not? Shouldn't Windows be flexible enough to use a single system image for commonly available hardware?
First, this configuration obviously worked fine for SP2. Second, Microsoft controls the driver certification process, so they should be able to ensure that Intel drivers aren't loading on an AMD system. This is a pretty minor fuckup, but it's firmly in MS's lap.
I dunno, if you could buy a new engine for a hundred bucks, they would just reinstall it rather than rebuilding it. Any repair that takes longer than a couple hours is uneconomic when a decent Dell is $500.
Also I have a feeling that these people who claim to "know computers" are actually mouse jockeys who have not RTFMed (resource kits) or have any real idea how windows works under the hood. The typical tech seems to know nothing beyond run anti-spyware and defrag.
Weird part is that the author has written numerous forum posts describing his issues with the.NET API (particularly the collections). But rather than summarizing the technical nitty-gritty for his editorials, instead he chose to devote space to how dumb VB developers are and then leapt right to his conclusions.
I wasn't implying that you couldn't have a successful site with RoR (I have one). Just that compared to about anything else, it's slow from top to bottom.
The Ruby Rails App Server finger pointing misses the point IMO.
But the Open Group is responsible for licensing the name "Unix" and setting the compatibility standards, not SCO. Basically SCO and Open Group were shell companies so that AT&T, IBM, HP, SUN, Novell and other Unix licensers would not get in trouble for anti-trust violations by holding the other companies hostage for technology they all shared. Very insightful description.
(Novell in particular always had a very peculiar relationship with UNIX because their main revenue source was/is NetWare. SCO was made the System V frontman largely for PR purposes.)
The biggest problem with Rails is that everyone involved is a huge zealot, and it's difficult to find any sort of objective assessment of where the faults are. To discover any of this stuff requires wading hip-deep in various blogs and mailing lists.
The whole "nobody told me this" is a terrible situation to find yourself in when you're in production.
open smopen. Intel and AMD are joined at the hip with patent cross-licensing and always have been. VIA licensed their stuff too (through IBM I believe).
Better argument might be that the original i386 patents have now expired, so if you're careful and have good lawyers you might be able to build a work-alike.
The summary asserts that Canonical is a "traditional for-profit company," but the Wikipedia entry you point to paints a picture of a company that is not traditional. Wikipedia is written by fanboys, news at 11.
Canonical is more in the reputation-building phase, so there's not much else to say other than how they are building their reputation. Wait for the IPO prospectus to see their revenue plans I guess.
Excel 2003 at least acts like a weird bastardization of MDI and SDI. I don't the the GP is correct that they "half-assed" it but they definitely did something funky to try to make everyone happy.
Slashdot Mirror
Some % of their Windows corporate sales depend on there being a mostly-compatible Mac version.
This was probably a bigger deal 15 years ago, but it was always an advantage that MS had over their competitors.
Yes, this is basically it. Mac users whined mightily about getting a Windows port, so MS forked their codebase and now the Mac is always behind feature parity with the Windows version.
(IMO the UI in Word 6 was the least of its problems and probably could have been fixed without losing the cross-platformness.)
However a MSDN seat license (at $5000 or whatever) is certainly not a cost-effective way to buy MS Office. I believe tho that you can downgrade Access 2007 to 95 with the corp license, so its useful to have the media.
My guess is that they're cooking something which VBA to be hosted on the
Had Usenet "evolved" it still might be an vigorous discussion network and not a refuge for old timers, kooks, and trolls.
While you wouldn't be able to trademark "Open", there is an organization called "The Open Group" and they'll gladly publish your open standards if you pay them.
Did I say anything about trademarks? I mearly stated that you guys copped an existing term from the intelligence community.
And you have been very inaccurate about the trademark status of "open source" in the past, so you're hardly one to talk.
In ~2002 I consulted at a place where they were implemeting 2-digit year fields in a new system. "Now that Y2K is over we don't have to worry about it anymore..."
The reality is that OSI's extended definition of this "open source" caused a lot of confusion. How many stories has Slashdot has run about "available source"/"not really open source" software from Microsoft, Apple, Sun, etc etc. If OSI had clear branding that wasn't overriding an existing term this wouldn't be half the problem it is.
I guess I don't understand your POV. Just because you had to sit on it for work doesn't mean that's the right way to proceed. All these things have vendor strings and PCI IDs, Windows should be smart enough to ignore irrelevant drivers. Linux boots on all this stuff with a single kernel after all.
First, this configuration obviously worked fine for SP2. Second, Microsoft controls the driver certification process, so they should be able to ensure that Intel drivers aren't loading on an AMD system. This is a pretty minor fuckup, but it's firmly in MS's lap.
Then why did you bring up the cost? Windows has its problems, but face it, Linux ain't cheap to support either.
haha. Who exactly are you going to hire to repair your linux system? A $75/hour enterprise unix admin?
I dunno, if you could buy a new engine for a hundred bucks, they would just reinstall it rather than rebuilding it. Any repair that takes longer than a couple hours is uneconomic when a decent Dell is $500.
Also I have a feeling that these people who claim to "know computers" are actually mouse jockeys who have not RTFMed (resource kits) or have any real idea how windows works under the hood. The typical tech seems to know nothing beyond run anti-spyware and defrag.
The cargo area in a CRX is really tiny, perhaps the body wouldn't fit.
(If the argument is that the state didn't prove premeditation, I guess I could see that. His clean-up job was pretty sloppy.)
There seems to be a ton of windows code that looks kinda like this: "$WINDIR\System32\..."
Like they got it half right and then just went ahead and hardcoded the other half.
Weird part is that the author has written numerous forum posts describing his issues with the .NET API (particularly the collections). But rather than summarizing the technical nitty-gritty for his editorials, instead he chose to devote space to how dumb VB developers are and then leapt right to his conclusions.
If you're going to nitpick, it's COM+ (with IIS) that manages the connection pool. Inproc ADO doesn't have it.
I wasn't implying that you couldn't have a successful site with RoR (I have one). Just that compared to about anything else, it's slow from top to bottom.
The Ruby Rails App Server finger pointing misses the point IMO.
(Novell in particular always had a very peculiar relationship with UNIX because their main revenue source was/is NetWare. SCO was made the System V frontman largely for PR purposes.)
Even ASP has db connection pooling :/
The biggest problem with Rails is that everyone involved is a huge zealot, and it's difficult to find any sort of objective assessment of where the faults are. To discover any of this stuff requires wading hip-deep in various blogs and mailing lists.
The whole "nobody told me this" is a terrible situation to find yourself in when you're in production.
RoR has scaling problems on all levels. Ruby is slow. Rails is (or can be) slow. The application servers have issues. Everything is bloated.
Its not like you can point at one element and say "there's the problem", its really a top-to-bottom issue.
open smopen. Intel and AMD are joined at the hip with patent cross-licensing and always have been. VIA licensed their stuff too (through IBM I believe).
Better argument might be that the original i386 patents have now expired, so if you're careful and have good lawyers you might be able to build a work-alike.
Canonical is more in the reputation-building phase, so there's not much else to say other than how they are building their reputation. Wait for the IPO prospectus to see their revenue plans I guess.
Excel 2003 at least acts like a weird bastardization of MDI and SDI. I don't the the GP is correct that they "half-assed" it but they definitely did something funky to try to make everyone happy.