Re: Last few pieces of GTK+ documentation needed

[David Benson <daveb idealab com>]

On Fri, 22 Oct 1999, Torsten Landschoff wrote:

> Hi *,=20
> 
> Hmm, a bit quiet on this list, isn't it? Or do I have a problem with my=20
> mail system?

Yes, I'm seeing =20 at the end of every line of your post. :)
But the list itself has just low traffic, I think.

> But I have some issues with writing those. Looking at the available=20
> documentation there are often notes like "this is obsolete" or "deprecated".
> As I am not a core GTK+ developer (I would like to work on it but I am=20
> quite envolved in my Debian work already...) it is not clear to me from=20
> where I should derive such information.=20
> 
> I looked at the source and found that there isn't even a comment "this is
> obsolete" or something. With CVS it should be really easy to get such=20
> a simple change into the repository - I can't understand why this is not=20
> done.

First off, widgets are pretty hard.  So be aware you're
tackling a tough issue.  Be prepared to spend a _lot_ of time.
It took me forever to write signals and widgets seems a lot harder.

It'll be a great learning experience though.

I see some references to "deprecated" in the .h files.
Note that usually the private functions are marked "Non-public" in
the .h file.

Note that they do a _lot_ of pruning-- there just aren't
many deprecated functions, really.

I'm sure people on this, and the dev list, will
be happy to answer questions.  (I want to thank
Tim Janik for being extremely helpful in
proofing the signal docs.)

The real trick is that you have to get the whole
thing into your head, chew on it, and spit it out
as cleanly and cogently as possible.
I want to emphasize getting beyond a description
of each function and primarily instead writing a nice summary.
If you have that, the function descriptions will
follow naturally.  Otherwise, they won't be that helpful,
and they'll prevent others from trying.

- Dave Benson