[ Thread Index |
Date Index
| More lists.liballeg.org/allegro-developers Archives
]
On Thursday 06 January 2005 13:29, Elias Pschernig wrote:
> Makes sense. Where do we actually need "long" at all?
In most cases, we don't really. In some cases, it should really be size_t,
in others it should really be int...
> And for after 4.2,
> I think we should switch to the C99 types (int32_t int64_t).
Absolutely.
> I don't know if this is a good time to rename them. Using "w" and "l"
> for 16/32 bit seems to make as much sense to me as using "i" and "m" for
> little and big endian.
Well, i for Intel and m for Motorola... makes sense to me ;)
Personally, I get rather confused with w and l. But the point was more to
demphasize the assumption that long = 32 bit, which isn't really true
anymore. The function should be named for what it does: it doesn't write a
long (well, ok, in practice it does actually), it writes a 32 bit integer.
I think it would make sense to do this now, but if not, then we should do
something like that for after 4.2 anyway. Whatever small things we can do
to make 4.2 more safe I think we should. It'll probably be the `most
recent stable version' for a while after release.
> Actually, when looking at the 3 version above, I vote for keeping the
> current names :)
Can't say I particularly like packf_iputi32 myself... maybe I'd prefer
packf_iput32.
Evert