Finally I get some time to throw my 2 cents into this ... :)
First of all I would like to say it is a shame to see you go, Gareth. You've
done some really great work on this project and I think everybody appreciates
it a lot!! I hope you have a good time relaxing for a while and find a good
job later on - although with your skills I don't think that will be a problem.
Also many thanks to all the other developers you've done great work too.
Especially to Daryll and Alan for putting up with my sometimes stupid
questions. :)
Now ....
>>Certainly, but you have to be "inside" to have any chance of
>>getting this NDA. That is my impression at least. If you aren't
>>already well known you wouldn't know what questions to ask the
>>hardware company and you will therefore get the standard answer
>>about secret and proprietary information which they regretfully
>>cannot disclose to you.
>>(If you are a member of the XFree86 project you will have access
>>to some documentation though, but I can't see any particularly
>>new specifications there.)
>
>If someone is truely dedicated enough to work on XFree86 code,
>they won't have problems getting the information they need, nor
>becoming a registered developer IMHO.
Both are good points. But I think even if you can't get the "inside info"
there is other work that can be done on the DRI that does not require indept
knowledge of the hardware. Such as helping migrate stuff to Mesa 3.5, fixing
minor bugs or improving performance.
That alone would help a lot I would think. If we can take some of those
easier tasks off the hands of the main developers they will have more time to
work on new "cutting edge" stuff while we do some of the grunt work. Then, as
you get better you can move up in the line. But basically, you have to start
somewhere. You can't expect to work on the coolest stuff right away.
>>�The main reason IMHO is that XFree86 is very large and takes many
>>�months to be able to just navigate the source tree, and begin to
>>�understand the Imake build system, etc. �As in kernel
>>�development, one must have a deep interest in overcoming the
>>�initial learning curves to get seriously involved in something
>>�like this. �It certainly isn't a one-nighter thing one can just
>>�pick up.
>
True, that would take a long time. But I don't think you have to understand
the whole of XFree86. You would be working on the DRI which seems to be a
more isolated part of it with less files. If you just make changes to the DRI
driver files or related source files you don't have to learn the rest of the
build system in detail.
>There are also no list of (easy) stuff that needs to be done.
>If such a list existed and pointed out roughly which files you
>would need to look at it might be a nice introduction to
>XFree86-hacking. On the other hand, maintaining such a list
>might be more troublesome for a seasoned XFree86-hacker than doing
>the work oneself.
If you are looking for stuff to start work on why not browse the feature
requests/bugs on Sourceforge. Sometimes there is some easy to fix stuff in
there. Also, just pay attention to the mailing lists and look out for
problems people report. Some of them are easy to fix and maybe even trivial -
but as you solve those you will get a feeling for the code and can move on to
greater things.
>If someone is truely anxious to do XFree86 work, one MAJOR area
>that needs doing, is porting the remaining 3.3.6 servers to the
>4.x driver model. �The DESIGN doc, and a few others included in
>the sources, coupled with perhaps some hardware specs, and some
>sample hardware is a good start, along with the free time to do
>it. �I would like to hack on some of that myself but higher
>priority issues prevail.
This is another good example for tasks that can be done which do not require
insider info.
Also, some people have said that there is no documentation. This is also
something that people can write themselves. If there is interest I can put a
little "knowledge archive" on the website. So, as you start working with the
source you can document your steps and experiences. Then, as other people
start working they can look through what you have written. This could grow
into a collection of lots of little articles that could eventually be
combined into a larger document.
I am not saying you should write a novel here. Just write one or two pages
whenever you figure something out that you feel would be valuable to other
people.
Ok, that's all I have to say. Cheers...
- Frank
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/dri-devel
