MS Vista takes another hit

From a trade rag. This could be interesting:

-- Microsoft Exec: 'We Really Botched' Vista Campaign

Microsoft's efforts to fend off a class action lawsuit over a
controversial Windows Vista marketing campaign took a hit last week when executives' e-mails were presented in court.
Redirect to the target website

 

I'm sure in a few years the entire Windows world will be running on Vista, save a few old derelict machines. That seems to be Microsoft's pattern, but this time is worse from my perspective, and my perspective is pretty long. I've been designing hardware and writing software from before the days of Microsoft (back when dinosaurs ruled the earth; come to think of it, dinosaurs still rule the earth, or at least the free market). I worked on the DEC PDP-8, PDP-10, PDP-11, PDP-14, IBM 1800, Harris something or another, and VAX (my VAX experience may have come after the IBM PC hit the market). Those were the mini-computers. I also worked on the Intel 8008 (missed the 4004 by a smidgen), Intel 8080, Intel 8085, Z80, some assorted Motorola processors, and many of the Microchip micro-controller family. Microchip doesn't really count, since it came later and is all embedded, but that's what I work with now.

I remember when MS-DOS hit the market. We all thought it was a joke. It didn't come close to any of DEC's operating systems. Here I was, writing system code for a real-time, multi-user, multi-tasking OS when out comes this piddling little piece-of-s**t OS. I thought it was toy, ripped off from another company, and not a very good copy at that. I gave it no chance for success. Boy was I wrong.

Move forward two companies and a few years, and I found myself heading a development engineering group in an electronics company where we were all using IBM PCs running MS-DOS. Windows was just arriving on the scene. We were doing a joint development project with Siemens in Berlin. At that time, MS Windows and GEM Desktop were fighting over who would win in the PC market. My colleagues at Siemens were backing GEM, and I backed Microsoft. It was kind of on party lines, since GEM was doing better in Europe and Microsoft was doing better in the U.S. My contention was that OS/2 was the operating system of the future, and MS Windows was the logical path to OS/2. We went with Windows and jumped on the bandwagon at Windows V1.03. What a cluster-f***. We couldn't even think of delivering a product until Windows 2.0. In Windows 1.x, the various library modules wouldn't work together. Even Microsoft told us to wait.

Windows 2.0 was an improvement. It worked reasonably well and offered some but not all of the GUI features that MAC users had already enjoyed for a few years. The user interface was pretty good, but you still had to dive into the command prompt to do any real setup and configuration.

With Windows 3.0 Microsoft pulled ahead. Apple had been resting on it's laurels for a while, and Microsoft used the time to close the technology gap and buy it's way into the marked. The Windows train had left the station and there was no stopping it. Windows 3.0 was still a disaster under the hood, but most users didn't care about that. It was a bunch of DOS code cobbled together with bailing wire and bandaids.

Meanwhile, OS/2 was quietly dieing on the vine. It was supposed to replace Windows, but Microsoft was developing it with IBM, and Microsoft never likes to share anything with anyone. As soon as they had a chance, Microsoft dropped OS/2 and rolled their own with Windows NT. OS/2 was a nice OS and could have been good if allowed to fully develop. I met Bill Gates and Gordon Letwin in NYC during an OS/2 beta developer's conference. Gordon was great. There were all of these marketing suite type guys running around, and here was Gordon with his big, bushy beard and Hawaiian print shirt. You had to love him. Besides that, he used to work for Heathkit.

Windows 95 and 98 were in-between steps which removed some of the ugly code from the inside of Windows, and introduced a few gee-whiz pieces of dancing bologna user interface multi-media crapolla. They were better, but still not robust under the surface. Windows NT was developed at the same time, but took a different path. While 95 and 98 were holding the seething masses at bay, the NT guys were trying to make a good OS. They started from scratch and built a reasonably good OS, but it lacked all the fancy end-user features of Windows 95 and Windows 98. Windows NT was used mostly for servers and commercial desktop machines.

The two Windows camps finally re-converged with Windows XP, which took the insides of NT and put fancy end-user features on top. Of course hard core server and commercial versions still exist, and go by names such as Windows 2000 Server.

The transition to XP was hard on developers and end users. Much like Vista, it was a large jump, and there were a lot of compatibility problems and resource issues.

The transition to Vista may only appear worse than some of the others, or perhaps it is worse. I'm not sure, but to me it appears this has been caused by two issues:

1) Security: Microsoft has never done a stellar job with security. If they had, we wouldn't have all of the problems we have today with malware, viruses, and trojan horse software. Microsoft has always favored features over function. Rather than starting with a strong base and slowly adding features, they have cobbled furiously and produced a house of cards. They have tried to fix this with Vista. Whether the attempt is successful, only time will tell. What it has done is change the rules in the middle of the game. Developers are used to playing fast and loose with Windows, and much of the existing code does just that. Microsoft has locked the doors, and a lot of that code has been left out in the cold. It may turn out to be a good thing, but the transition is bumpy at best.

2 Timing: This really isn't a separate issue, but is part and parcel of the security and compatibility issues. It takes time to support a new OS. Applications have to be tested and sometimes re-written, changes are made and tested, and patches are issued. Developers can't start this effort until they have a working version of the OS for testing. Microsoft wants to cut this short for Vista. Vista and XP can co-exist for a few years, but Microsoft wants to get out of the business of supporting XP and start earning a return on the Vista development. You can hardly blame them for that, but it is folly to force this before the market is ready. Microsoft's headlong rush into Vista is generating a user revolt. I can't think of a better approach to driving customers over to the MAC camp. This campaign is penny wise and pound foolish.

Tom

 

I have some shrink-wrapped packages of DOS 6.2. I could send you one to make up for the trouble you've experienced. <smirk>

Tom

 

Vtie, all good points. I agree with you about sloppy programming. The Windows world has been full of it, since many of the applications were written by amateur programmers. Even a lot of the professionals are fairly sloppy. I think this largely came from the type of person who wrote programs for PCs. Many of them started as hobbyists with little formal training, working largely without peer review.

I too am happy to see Microsoft bite the bullet and do something about security. It's too bad it's taken this long, but as you point out, they are dealing with a lot of third party baggage. Every time they change anything it generates huge amounts of testing and work for everyone.

DEC had a similar problem with the PDP family of computers. They were so popular and there were so many varieties that it became difficult to support them. The switch to the VAX allowed DEC to stop supporting the PDP, as well as introduce a new, more modern architecture.

Device drivers, as you point out, are a different sort of animal. You have to expect them to be pretty well linked to the OS. I've written a number over the years, and they weren't always pretty. If the device drivers supply a good API, then it's not so bad for everyone else. Only the driver guy has work to do when the OS changes.

Thanks for your good comments,
Tom