| 1 |
On Wednesday 12 November 2003 18:46, Alastair Tse wrote: |
| 2 |
|
| 3 |
> I need some more testing for python-2.3.2 before it is released into |
| 4 |
> ~x86. |
| 5 |
|
| 6 |
This update is looking good here. Fine work. 8) |
| 7 |
|
| 8 |
> 2. run: emerge -u portage python |
| 9 |
|
| 10 |
note that -u will update alot of things with ~x86 that you might not want. |
| 11 |
|
| 12 |
> UCS4 uses significantly more memory than UCS2. |
| 13 |
|
| 14 |
I have compared ucs2 (with the ebuild in portage) and ucs4 (with hacking that |
| 15 |
ebuild to include the --enable-unicode=ucs4 configure switch). I compared: |
| 16 |
1. The sizes of a newly started interpreter using the RES column on 'top' |
| 17 |
2. The size of a "btdownloadheadless" process seeding a knoppix cd image |
| 18 |
3. The size of the .tbz2 binary package. |
| 19 |
|
| 20 |
The empty interpreters are of identical size. The bittorrent process is |
| 21 |
considerably *smaller* with ucs4, and the tbz2 files are largely unchanged in |
| 22 |
size. Numbers below. IMHO thats a clear but suprising win for ucs4. Lets use |
| 23 |
it always. |
| 24 |
|
| 25 |
> Plan for Python's Future |
| 26 |
> ======================== |
| 27 |
|
| 28 |
Portage has a number of packages that still need 2.2 but reference |
| 29 |
#!/usr/bin/python. For example, rdiff-backup. The distutil eclass has support |
| 30 |
for forcing /usr/bin/python2.1 (thanks to zope), but I cant see any way to |
| 31 |
force 2.2. am I overlooking something? |
| 32 |
|
| 33 |
|
| 34 |
|
| 35 |
|
| 36 |
ucs2 ucs4 |
| 37 |
empty interpreter 2928 2928 |
| 38 |
btdownloadheadless 6076 5464, 5420 |
| 39 |
tbz2 size 5954517 5951807 |
| 40 |
|
| 41 |
-- |
| 42 |
Toby Dickenson |
| 43 |
|
| 44 |
|
| 45 |
-- |
| 46 |
gentoo-dev@g.o mailing list |
Archives