93 lines
4.0 KiB
Plaintext
93 lines
4.0 KiB
Plaintext
From: tismer at appliedbiometrics.com (Christian Tismer)
|
|
Date: Sat, 3 Apr 1999 11:40:24 GMT
|
|
Subject: WinDLL CallDLL
|
|
References: <370327E3.66047D53@easystreet.com>
|
|
<3703CB0C.83AB035E@ingr.com> <370402DE.E6AD7C37@easystreet.com>
|
|
Message-ID: <3705FE28.372A3D7A@appliedbiometrics.com>
|
|
Content-Length: 3696
|
|
X-UID: 105
|
|
|
|
|
|
Al Christians wrote:
|
|
>
|
|
> I'm still getting some success running the DLLDemo.Py, but none
|
|
> with any of the others. I get addressing exceptions from them.
|
|
|
|
Well, I have never tested with NT SP4, so we can't be sure
|
|
which piece of software is going nuts here.
|
|
I will send you exactly the .pyd files which I use,
|
|
and we will try some basic steps together. If that doesn't
|
|
work, I would believe you have a problem with you system.
|
|
|
|
[other path fiddling stuff which does not address the problem]
|
|
|
|
> I've done that, too, and made the change to DynWin\__init__.py given
|
|
> by Christian Tismer.
|
|
|
|
Whatever you did, there is no reason for the kind of crashes
|
|
you get. Dynwin should not work at all since it cannot import
|
|
a module, or it should work just fine.
|
|
|
|
> I've deleted my \Python directory and done several reinstalls of
|
|
> Py152b2.exe. No success. Do I need to install from the source.
|
|
> I've also tried re-building CallDLL.Pyd using MSVC++ v5. That
|
|
> seemed to succeed. But I get the same run-time results.
|
|
|
|
No, you need to use another approach to track this down.
|
|
There is a different problem which you will not solve
|
|
unless you do smaller steps. I never had the problems
|
|
you have, just different ones. We need to step by steps
|
|
check out if you can use calldll and npstruct as necessary,
|
|
then check the modules which are built upon it.
|
|
I will attach some examples which you should try.
|
|
|
|
> I don't complain about something given gratis, like all this fine
|
|
> python software, but there does seem to be some potential for improving
|
|
> the packaging or beefing up the documentation a little here and there.
|
|
|
|
Sorry, you are right in the general case but not in this one.
|
|
Sam has done some great work. He has a very special minimalistic
|
|
approach, and his number of users is relatively small compared
|
|
to, say PythonWin. He never claimed that his stuff is ready,
|
|
but it is all alpha stuff where he is providing a preview
|
|
for the curious. Using DynWin is not using a ready-made package
|
|
which has a couple of maintainers, but you are on your own, or
|
|
better: You are jumping into the boat to help Sam with improving it.
|
|
|
|
> I can offer some time to try to help do that for some of these packages
|
|
> that run on NT. If any developers of python code that is licensed as
|
|
> some variety of free or open source want to take advantage of the point
|
|
> of view of a typically brain-dead potential user for a few hours and
|
|
> collaborate on recording what we have to learn to do to get this stuff
|
|
> installed, they could contact me. I suppose, however, that in most
|
|
> cases it is pure lack of time that prevents production of more
|
|
> documentation, and that it would be easier for the developer to write
|
|
> it directly than to try to collaborate.
|
|
|
|
This case is even worse since at the time Sam was working
|
|
on DynWin and LumberJack, having reached his "critical mass",
|
|
he was hired away
|
|
"""Ack! I have taken a full-time job, so work on lumberjack will
|
|
halt for a couple of months. Stay Tuned..."""
|
|
|
|
This was from October 17, see
|
|
http://www.nightmare.com/~rushing/lumberjack/index.html
|
|
|
|
That means he might be in the "gosh I simply got no time" state
|
|
which I know very well. Until that becomes better, people
|
|
like Gordon McMillan and me might be able to help.
|
|
|
|
cheers - chris
|
|
|
|
--
|
|
Christian Tismer :^) <mailto:tismer at appliedbiometrics.com>
|
|
Applied Biometrics GmbH : Have a break! Take a ride on Python's
|
|
Kaiserin-Augusta-Allee 101 : *Starship* http://starship.python.net
|
|
10553 Berlin : PGP key -> http://wwwkeys.pgp.net
|
|
PGP Fingerprint E182 71C7 1A9D 66E9 9D15 D3CC D4D7 93E2 1FAE F6DF
|
|
we're tired of banana software - shipped green, ripens at home
|
|
|
|
|
|
|
|
|