| 1 |
-----BEGIN PGP SIGNED MESSAGE----- |
| 2 |
Hash: SHA1 |
| 3 |
|
| 4 |
Caleb Cushing wrote: |
| 5 |
>> Are we to say that we shouldn't allow tools to have support for this. I |
| 6 |
>> think that it is a nature progression that if we are to allow overlays to |
| 7 |
>> extend the portage tree that we should allow overlays to extend other |
| 8 |
>> overlays. |
| 9 |
> |
| 10 |
> I probably shouldn't butt in... |
| 11 |
> |
| 12 |
> first, no I don't want you to merge java-overlay and |
| 13 |
> java-experimental, that's a bad idea (well at least for me) |
| 14 |
|
| 15 |
If you think neither should exist why do you have an opinion about this at all? |
| 16 |
|
| 17 |
> second. I generally think anything beyond a personal overlay is crap. |
| 18 |
> All these overlays like sunrise, java-overlay, and on and on... |
| 19 |
> basically official, overlays that have qa and are pretty stable. are |
| 20 |
> crap. they should be in the tree. an overlay for developers is fine, |
| 21 |
> you know. where you are working on stuff... stuff that someone who |
| 22 |
> wouldn't want to hack on it wouldn't want, because it's too broken. |
| 23 |
|
| 24 |
What makes you think that overlays aren't for developers, aspiring developers |
| 25 |
and interested users where they are working on stuff? It is desirable IMO that |
| 26 |
all such people can easily be given full access to muck around and learn. |
| 27 |
Further, overlays are good places to put ebuilds for software that is more |
| 28 |
experimental than what's expected for ~arch. That includes live ebuilds. In the |
| 29 |
end, overlays have a (far) lower level of guaranteed quality than the main tree, |
| 30 |
for their ebuilds. |
| 31 |
|
| 32 |
> but one of the few good things about gentoo, in relation to other |
| 33 |
> distro's, 1 tree no repos, continues to fall further and further |
| 34 |
> apart. |
| 35 |
|
| 36 |
The only thing I see is that the flow of ebuilds from overlays into the main |
| 37 |
tree, when they and their software have reached sufficient quality, is sometimes |
| 38 |
slow because of lack of developer manpower. For example, Common Lisp is almost |
| 39 |
entirely maintained in the lisp overlay and I have the impression that the |
| 40 |
Haskell team is having similar issues with many of their former developers no |
| 41 |
longer working on Gentoo. |
| 42 |
|
| 43 |
I might even argue that Funtoo is one big overlay. When your own ability to |
| 44 |
contribute directly depends on an overlay, then why are you arguing against |
| 45 |
other people's overlays? |
| 46 |
|
| 47 |
Marijn |
| 48 |
|
| 49 |
|
| 50 |
- -- |
| 51 |
Sarcasm puts the iron in irony, cynicism the steel. |
| 52 |
|
| 53 |
Marijn Schouten (hkBst), Gentoo Lisp project, Gentoo ML |
| 54 |
<http://www.gentoo.org/proj/en/lisp/>, #gentoo-{lisp,ml} on FreeNode |
| 55 |
-----BEGIN PGP SIGNATURE----- |
| 56 |
Version: GnuPG v2.0.10 (GNU/Linux) |
| 57 |
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org |
| 58 |
|
| 59 |
iEYEARECAAYFAkmtDPUACgkQp/VmCx0OL2z6lwCdHqP2shPhD91HU9Ld+f/Ma+K3 |
| 60 |
/6YAnR0cMKEXkqF3ZzA4hkahkPmTQYxR |
| 61 |
=MVhI |
| 62 |
-----END PGP SIGNATURE----- |
Archives