On Wed, 7 Mar 2001, Christoph Egger wrote:
>
> Hi list!
>
>
> Eric said in a former mail, he is working on an unknown extension.
>
> Peter said in a former mail, he wrote a new (unknown) display-target
> for an unknown extension.
>
>
> Folk: I won't caim, but it might be that everyone is working on
> something, that might be worthful for GGI, _why_ is doing the work
> NOT NOTICED on the GGI-list?
Define 'noticed'. I'm sure that lots of people are reading the
postings. If you mean 'why don't more people step up and contribute',
that can be explained by two simple facts: graphics systems coding is
difficult in general, and GGI coding is difficult in particular. You
gotta get used to the reality of the situation, which is that it takes a
LOT more development before GGI-related code is at a state where many
folks will be willing to pick it up and hack on it than it is to, say,
code up a perl script or GNOME C++ apps-level code.
If you want to pick up interest in your code, the best way to
promote this is to make sure that your released code is "easy to hack".
This means:
* The code builds and runs. If this isn't there, many people will look no
further.
* Demo code is provided, with lots of comments. This provides a good
basis for extending the API featureset, since people can cut and paste
the demo code and extend it to test their new features.
* The library code is also well documented
* All allocated resources are freed (hi Marcus! |->).
* The current state of the code, especially known bugs, is well
documented. There's nothing like cutting and pasting the same buggy
code a few dozen times, only to find that you have tripled your
debugging time....
* You (the author) are willing to spend time discussing the code, design,
and bugs. Even if you are not willing to code, always be willing to
debug.
> If it is so, then there is no wonder for
> me, when everyone in the world believes, GGI is going to be dead...
We still need to get our house in order before we start shouting
'GGI really works this time! We really mean it, honest!' to the world
again. I say we junker down and code, fix bugs, chat on the list and work
to clean up and enhance the extensions we have. I'm hopefully going to be
able to do some more GGIMesa work here soon, and when PhoeniX is released
(soon?) Steffen and I will hopefully have a chance to make KGI work
properly under 2.4 kernels.
> To figure that out, I want to make a survey:
>
> 1. a) Is there any work you did, which you haven't announced here?
Sort of. I have some code in my local degas/ tree here at work
which allows to run the DirectX target on Linux using WINE. It's a pretty
hideous mess at the moment, but I'll try to clean it up such that it is
suitable for committing soon. The DirectX target will need to be upgraded
to conform to DX7 for this, though. Any objections, John F.? I think
that requiring users to upgrade to DX7 (not 8) is not unreasonable. 6.1
is pretty outdated these days....
> 1. b) If yes, what have you done?
See above.
> 2. What work are you _doing_ now?
Minor GGIMesa cleanups.
> 3. What do you plan to work on in future?
LibGGI3D, LibXMI, GGIMesa and KGI.
Jon
---
'Cloning and the reprogramming of DNA is the first serious step in
becoming one with God.'
- Scientist G. Richard Seed
