Complete.Org: Mailing Lists: Archives: freeciv-dev: September 2003:
[Freeciv-Dev] Re: tile API
Home

[Freeciv-Dev] Re: tile API

[Top] [All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
To: "freeciv-dev@xxxxxxxxxxx" <freeciv-dev@xxxxxxxxxxx>
Subject: [Freeciv-Dev] Re: tile API
From: Bursig Rafal <bursig@xxxxxxxxx>
Date: Tue, 23 Sep 2003 19:22:00 +0200

Dnia 2003.09.23 18:47, Per I. Mathisen napisał(a):
On Tue, 23 Sep 2003, Gregory Berkolaiko wrote:
> Why me?

You're a map kind of guy.

> Anyway, I just run two autogames with the new patch and without.
...
> I would discount the increase in time to inevitable variations.
> But it means that if there is an increase in speed, it's
insignificant.
>
> Rafal, maybe you can send us the full patch and then we measure the
speed change having full info rather then a small part of it.

I tested the full patch. The difference in speed was insignificant.
The patch now in question probably accounts for it all, I'd guess. I'm for committing it without the new iterators and leave the rest be.


Per you don't test entire patch becouse I can't create it with current speed of changes in cvs.

I made some test with my p233mmx and old tree but it is little old and some slowdows was already fixed in cvs. (warmap init and etc)

without patch
9347.45user 1.98system 2:35:52elapsed 99%CPU (0avgtext+0avgdata 0maxresident)k 0inputs+0outputs (444major+35537minor)pagefaults 0swaps
civgame+1991.sav.gz
 with patch
6231.83user 2.19system 1:43:55elapsed 99%CPU (0avgtext+0avgdata 0maxresident)k 0inputs+0outputs (446major+40368minor)pagefaults 0swaps
civgame+1999.sav.gz

IMHO new itarations are very important and can fully replace old ones.
But most important is that we change our thinking and only use x,y api when we can't use tile api.

Rafal



[Prev in Thread] Current Thread [Next in Thread]