| 1 |
On Wednesday 12 November 2003 18:46, Alastair Tse wrote: |
| 2 |
|
| 3 |
That all looks good. Im keen to give it a spin... |
| 4 |
|
| 5 |
> The reason why I'm not making this default is because UCS4 python uses |
| 6 |
> more memory. An example is supybot (Python IRC bot) that uses 8M for |
| 7 |
> UCS2 and 13M for UCS4. |
| 8 |
|
| 9 |
Ive not used ucs4 python yet, but it is one of the things I was looking |
| 10 |
forward to in version 2.3. It would much nicer to leave ucs2 behind. |
| 11 |
|
| 12 |
If ucs4 strings were the only cause of that difference, supybot would need to |
| 13 |
be storing 2.5 million unicode characters. I guess that isnt likely. |
| 14 |
Excluding bugs, I dont see any reason why a program that doesnt use any |
| 15 |
unicode objects would use more memory when running on a ucs4 python |
| 16 |
interpreter. |
| 17 |
|
| 18 |
> But note that this example is not scientific |
| 19 |
> because the machines were different in kernel version, compiler and |
| 20 |
> compiler optimisations. |
| 21 |
|
| 22 |
Those reasons sound much more plausibe to me. Does anyone have a more |
| 23 |
scientific comparison of the effect of the ucs4 option on python? |
| 24 |
|
| 25 |
|
| 26 |
-- |
| 27 |
Toby Dickenson |
| 28 |
|
| 29 |
|
| 30 |
-- |
| 31 |
gentoo-dev@g.o mailing list |
Archives