Jonathan Anderson's Web page

Why is Microsoft so loathed?

I asked lots of different people and this is what they said

I ASKED A UNIX SUPPORTER WHO CURRENTLY RUNS A LAN AND INTENDS TO GO INTO BUSINESS SETTING UP AND MAINTAING NETWORKS. HIS EXPERIENCE OF NT WAS 'SECOND-HAND' FROM NT ADMINISTRATORS.

Has Linux fallen by the wayside compared to NT?

Fallen by the wayside?
Far from it! It's estimated to be on 5-10M boxes now, way ahead of WNT (and continually widening the gap). Alive and well, believe me :)

What about predictions that NT will eventually take over? Surely people want to use the same software everywhere so they will choose MS?

Even without any bias, I can see that MS are screwing themselves into a big hole with featurism and continually changing proprietary data formats. They've managed to screw themselves right out of Java which is not going to go away, pointy-headed top executives are starting to question the extreme price involved in maintaining NT based networks, their various desktop products are getting so unstable that even Windows fanatics are starting to look elsewhere, the list goes on.
Also, the US justice department is now after them again for illegal marketing practise. Fining them $1M a day over MSIE was just the first step, they're "investigating" a bunch of other stuff as well.
MSIE itself isn't going to last long. It's based on a broken contract with Sun for Java. There'll be a cease-and-desist on it based on that soon enough.
Anyway, if I were an MS top exec, I'd be looking for a way out, right now.

But surely if people can use Visual Basic etc. instead of having to employ programmers for 30 grand a year to write in Java, companies could make savings, and NT would be the place to go to?

The difference in the programme results will be plain, however. I've seen my share of "VB Internal Error"s via http. I wonder why... :)
The differences between NT and unix are based on system function. NT still crashes. Not as often as W95, but often enough to cost companies using it with lots of downtime. The performance compared to any decent UNIX is low when there are more than a dozen users - it doesn't scale at all gracefully.
Proprietary "standards" like SMB, MS mail, and so on are incompatible with the rest of the world (i.e. UNIX).^[1] Administration of NT, contrary to advertisement, is said to be an utter nightmare. It'll function with the defaults but as soon as anything goes wrong it's a mess and a half and hard to recover.
Anyone who thinks they don't need a sysadmin with NT will be nastily surprised the minute anything goes wrong - and it will.
Support from MS is expensive and often required, yet it usually doesn't provide results as the guy on the 'phone is just going through the same documentation you already have.
And so on.
For a higher initial price but low running cost and high reliability a company can install AIX, Solaris, or a commercial BSD. For a near-nothing price and running cost, Linux or a free BSD. If you have a competent sysadmin (which you'll need in any case), all of those do just fine whereas he'll be in an unending war with NT.

But NT must surely get better. And when it does, people will start to look to it. Perhaps in 5 years' time UNIX may be no more...?

It depends upon NT getting better.
By all reports, 4.0 is less stable and more bloated than 3.51. Who's to say 5 won't be worse yet?
I find it interesting, by the way, that while MS advocates once praised NT up and down and proclaimed that unix was dead, they're now saying "Wait 'till it's fixed, UNIX will be dead..."
I don't think the commercial UNIX people have all that much to worry about from MS. They have more to worry about from Linux, as it's caught up to and is passing a few of them in function and stability. All continue to be way ahead of NT.
I like AIX lots, by the way. It's an impressive system. It just needs a few GNU utils to become very nice from a user standpoint.
^[1]Darren Edmunson writes: Samba is well documented and there are free unix implementations, not exactly proprietary. SMB is proprietary in the sense that its Microsofts implementation of a subset of the open Samba standard.
...but the point being made was that it limits things - it doesn't - I can print to any of our NT boxes' printers from my BSD box using a simple command line, I can send winpops to people from here....

QUOTE FROM AN INTEL SOFTWARE ENGINEER RECALLING DEALING WITH MICROSOFT

"We'd go to the application guys and say: 'We studied Excel, and we found this loop that if you change it you can speed up the whole thing by a factor of eight.'
"They'd say: 'We don't care. We're going to add this new feature. People buy our applications because of new features.'"

I ASKED A SOFTWARE ENGINEER WHO WORKS FOR ACORN COMPUTERS ABOUT MICROSOFT AND THE WORLD WIDE WEB

Actually - what do you think of MS's approach to the Web, and do you think it is fair of Sun to sue them over Java?

MS are killing the Web, but I'm past caring (the bits I'm interested in remain largely unaffected; the mainstream Web these days is utter cack with or without MS's 'enhancements'!).
Sun are fair to sue on grounds of breach of contract; MS are claiming Java compatibility, saying they pass more of the validation suite tests than Sun itself, and yet our (Acorn's) own Java implementation and that of Navigator 4 seem to run far more applets (in a test of about 14, MSIE 4 failed on 6, others worked with them all). We're using Sun's code. MS are using their own. MS's doesn't work properly (as usual).
However, who is or is not right is never the issue in the American legal system these days. Microsoft have more money; that's the end of it.
Microsoft will of course eventually topple under their own weight but it'll be a few years yet before we see it, and it annoys me to think how long technology might be held back by this one company.

Do you really think anyone could possibly come in under any reasonable cirumstances and snatch away IE's grip on the market, once it has taken over?

No, not until Microsoft disappear (or downsize).

Do you think a takeover of the browser market is inevitable or could it in fact be stopped in its tracks?

It is inevitable in the mainstream (though I think it's very unlikely that we'll see 100% coverage - well, we won't, since far too many people won't use MS stuff on principle). I'm more worried about Microsoft's incursions into news (e.g. MSNBC) and TV (cable operations) - this gets more Orwellian. They can (and do) shaft computers as far as they like; I'm frankly past caring. If people are stupid enough to pay out good money for the abominations Microsoft ship these days, then thats their problem. Personally, if I'm left with a Microsoft-or-nothing choice and the Microsoft solution doesn't do what I want fast enough or reliably enough, I'll go and sell double glazing or something!
There *will* be a backlash eventually, but people have been predicting it will happen tomorrow for several years. I've no serious thoughts on timescales here. The sooner the better, though. Computing is stuck in a very deep rut right now. With entry level machines running at (conservatively!) 200+ MIPS [Incidentally - a Pentium 166 MMX runs at 333 MIPS], we should be able to do a vast amount more than you can achieve under Windows 95 / NT 4.

Would such a takeover be a good or bad thing?

Oh, come on. Any monopoly is unquestionably a bad thing.