| 1 |
On Wed, Apr 11, 2001 at 07:50:15AM +0200, Achim Gottinger wrote: |
| 2 |
|
| 3 |
> I think it is generally a good idea to add configure to the patch, because it |
| 4 |
> is a bit confusing that the configure in work exists but does not work. So |
| 5 |
> eigther it must be removed if configure.in gets modified or added to the |
| 6 |
> patch. |
| 7 |
|
| 8 |
I agree with your suggestion. Sorry, I should have made that more clear in the |
| 9 |
previous email. |
| 10 |
|
| 11 |
> > Though I'll agree, it would be nice to have a Perl-free system (drobbins hides) |
| 12 |
> |
| 13 |
> Unfortunately xfree requires perl, so it whould be a xfree free system |
| 14 |
> too. |
| 15 |
|
| 16 |
> BTW: It whould be nice to have a python free runtime system too (nothing |
| 17 |
> personal this time, so no need to run or hide :-)). This is impossible at |
| 18 |
> the moment because of the env-update thing. It might be required on embedded |
| 19 |
> platforms where disc space is rare (spython ~5MB). I suggest recode |
| 20 |
> env-update in bash or c. |
| 21 |
|
| 22 |
Well, pkgmerge, ebuild and emerge are in python right now too. I don't have a problem |
| 23 |
if people want to start converting the more stable code to C++. A good start would be |
| 24 |
env-update and the .tbz2 functions (/usr/lib/python2.0/xpak.py). Since we're moving to |
| 25 |
an object framework, I think that C++ would be a better fit than C. |
| 26 |
|
| 27 |
If anyone is interested in doing this, please contact me first. Then I'll carefully |
| 28 |
coordinate my work in python with yours. |
| 29 |
|
| 30 |
Best Regards, |
| 31 |
|
| 32 |
-- |
| 33 |
Daniel Robbins <drobbins@g.o> |
| 34 |
President/CEO http://www.gentoo.org |
| 35 |
Gentoo Technologies, Inc. |
Archives