| 1 |
Greg KH wrote: |
| 2 |
> On Wed, Jan 17, 2007 at 01:53:12AM +0100, Markus Ullmann wrote: |
| 3 |
>> I've talked to a friend of mine recently. He's a FreeBSD dev and he said |
| 4 |
>> they tried git for their ports tree (which is basically the same what |
| 5 |
>> we're talking about) and it was more or less a big pain for multiple |
| 6 |
>> reasons. |
| 7 |
>> He said he'd personally take svn after that experience. |
| 8 |
> |
| 9 |
> What was the reasons he cited? |
| 10 |
|
| 11 |
Given that ports is pretty similar to our gentoo-x86, I'd guess about |
| 12 |
the same ones mentioned at |
| 13 |
http://dev.gentoo.org/~antarus/projects/gleps/glep-0666.txt -- I quote |
| 14 |
from there: |
| 15 |
|
| 16 |
I think migration for many would be frustrating and detailed guides |
| 17 |
for doing things in GIT would be the norm for quite some time. GIT |
| 18 |
also has some other issues: |
| 19 |
|
| 20 |
1. Git currently requires you to check out the whole repository. |
| 21 |
This includes *all of the history*. |
| 22 |
|
| 23 |
2. Git cannot update portions of the repository, it can only update |
| 24 |
the entire thing. |
| 25 |
|
| 26 |
3. Due to git's choice of packing format (which does save a lot of |
| 27 |
space), the operations are quite CPU intensive. Either the GIT |
| 28 |
server gets overwhelmed by the raw number of clients using it or |
| 29 |
the slower clients (arm, mips, sparc, hppa...basically anything not |
| 30 |
x86, ppc, ppc64, amd64) get screwed by the raw amount of CPU and |
| 31 |
RAM necessary to unpack a checkout from these packs. |
| 32 |
|
| 33 |
4. git-daemon (and git over ssh) both are very stupid when it comes |
| 34 |
to generating packs for transfer, since often two or three |
| 35 |
fetches can be going on but the packs are not shared between |
| 36 |
fetches. This only makes the already shakey server performance |
| 37 |
even worse, as the same packs are generated N times instead of |
| 38 |
once. |
| 39 |
|
| 40 |
If GIT gets repository slicing (ability to check out and update |
| 41 |
slices of the repository) as well as history slicing (only take the |
| 42 |
last six months of history, for example) I think it would be a |
| 43 |
better canidate. The conversion to GIT from CVS was also lengthy |
| 44 |
(approximately two weeks) althought many projects attempted a switch |
| 45 |
this summer and tools have improved in speed. |
| 46 |
|
| 47 |
Note: Both history and repository slicing are in the works for GIT, |
| 48 |
but there is no date of completion for them. |
| 49 |
|
| 50 |
-- |
| 51 |
gentoo-dev@g.o mailing list |
Archives