Richard,
> I agree with this. IMHO, Language wise, Microsoft has put all of its eggs
in
> the .NET basket. They really want this to succeed.
>
> Of Bill's examples above, I can only concur that PenWindows was
> a malicious diversion. (Read Microserfs on how this came about.
> Very interesting reading.)
I'll check it out. From having connected to the 2.0 Pen API, I have a
strong suspicion that it was cancelled during development, and they probably
back-tracked to a previous stable beta version. There are many confusions
between HANDLE and various ways of saying DWORD that give the appearance
that they were starting to clean it up and then quit.
BTW, all they'd have to do to largely un-scum themselves on this is release
some source code for the ink handling functions, and put the electronic copy
of the book that's now out of print on an FTP server. The fact that MS WILL
NOT DO IT should tell us something.
I'll grant that MDI is a weak example (hence the "lesser extent" phrase),
but, it was still something they were telling us to do, and at a time when
they were creating the first version of MS Money.
We can't blame MS for the Visual J++ disaster, because that ended due to
legal action by Sun. But, let's talk about Sun. I have no inside
knowledge, but, it struck me at the time that MS might have conjured C# as a
way to get around the rulings against them. I wouldn't put it past Sun to
wait until MS is almost ready to strike and then hit them with yet another
law suit. However, I also think that the way the chads fell in Florida will
make them less likely to try it (sorry to inject politics, but, it's
relevant).
> The rest has just naturally evolved in some 'better' replacement,
> which should be followed by languages and end user software.
DirectX RM is more complicated - it's fall caused OA to cancel a project.
They were kind enough to make the code available to us after cancelling it,
so I'm saying they did anything wrong. It got too hot to handle, and they
made a wise decision - it wasn't a fun decision, but, it was the right one.
However, it is very easy to believe that .NET will be just as volatile.
What then? Ok, it would be the guts of the VM vs. an add-on with poor
sales, but, what happens if MS decides to drop support for a critical
feature?
What's the word on resumable exceptions for a .NET app? I heard stuff like
"MS said they wanted to work with us on that" - sounds like the typical
pre-sale rose-colored-glasses stuff. Does it work? I would have a tough
time getting by w/o resumable exceptions. I'm down to one use of them, but,
it's critical.
One of my apps will use MS Agent if it's present. I chose that approach
because it allowed me to avoid the more invasive requirements of getting a
redistribution license. There are two troublesome conditions that MS
typcially includes: (1) the redistribution license is specific to one
version, so it becomes a repeating headache; (2) one has to agree to make
reasonable efforts to adopt a new version when it is released. The second
one is the show stopper. First, I don't have time to poll for when I need
to upgrade, and more importantly, I might not _want_ to upgrade. I can't
have a pre-condition to a license agreement being one of the factors in a
go/no-go decision.
So, just link to the MS Agent site, right? Maybe. I tried that, allowing
users who wanted agents to go through the same process I did to get them.
Problem: MS updated the site, and now we get something like
Your search using for "download agents" returned 0 results.
Makes me glad the app will start without MS Agent. Note that I haven't
worked very hard at finding how to download this stuff because it's not
critical, but, what happens when it is critical, time's tight, and the site
just happened to change last week?
Have a good one,
Bill
--
Wilhelm K. Schwab, Ph.D.
[hidden email]
Locked
