| 1 |
Hi, |
| 2 |
|
| 3 |
After Ciaran's comment on -core, I wondered how imlate would fare on |
| 4 |
x86. For those who are not familiar with it, imlate list all packages on |
| 5 |
one architecture that are outdated compared to another architecture. All |
| 6 |
other architectures use x86 as a reference.. That clearly didn't work |
| 7 |
for us. So I hacked it to compare x86 to all other Linux keywords. |
| 8 |
|
| 9 |
The results are surprising, 214 packages are more recent on non-x86 |
| 10 |
architectures. But its pretty hard to know if its because they have arch |
| 11 |
specific patches or because the maintainer's arch is not x86... We |
| 12 |
should really have a way to specify in the ebuild or metadata.xml what |
| 13 |
arch is the maintainer arch. |
| 14 |
|
| 15 |
I also ran the notx86 (notamd64 renamed) script on the tree (it finds |
| 16 |
packages that dont have an x86 keyword at all)... and I handfiltered the |
| 17 |
result for obvious errors.. |
| 18 |
|
| 19 |
This is all available at http://dev.gentoo.org/~tester/imlate/ |
| 20 |
|
| 21 |
Ah yea, and I also tried my new imlate script on amd64 and it finds |
| 22 |
about 50% more packages than the regular version.. |
| 23 |
|
| 24 |
We really need to form an x86 team... |
| 25 |
|
| 26 |
-- |
| 27 |
Olivier Crête |
| 28 |
tester@g.o |
| 29 |
x86 Security Liaison |
Archives