[ Thread Index |
Date Index
| More lists.liballeg.org/allegro-developers Archives
]
Shawn Hargreaves <shawn@xxxxxxxxxx> writes:
> - is it safe to assume that sbin will always be pathed, so that install-info
> will work? I say this because it isn't pathed on my system (even as root),
> but I'm not sure whether it was this way by default, or whether I screwed
> something up while editing rc files :-)
Makefile runs "$(INSTALL_INFO_DIR) --version" first to test that it
works. This fix alone would make install-info "work" for me (at least
make will not stop with error, because I have no install-info in
/sbin, it is in /usr/local/bin). Perhaps, else branch in the above
test should suggest to call make with INSTALL_INFO_DIR defined with
full path of install-info (which might be missing at all). And try
'which install-info' or 'whereis install-info', maybe you have another
copy in the different path.
>
> - this more portable version of BYTES_PER_PIXEL() will be marginally slower
> than the old implementation, and this macro is used in a lot of places.
> Since these changes are only relevant for machines where sizeof(short) != 2
> or sizeof(long) != 4, would you have any objection to me only using your new
> version when ALLEGRO_I386 is not defined?
No objections. New macro is necessary, because on all platforms
bitmaps are accessed as arrays of unsigned char, short, or long. But
on i386 they are also accessed as 1 byte, 2 bytes or 4 bytes in asm
routines.
>
> The X code is working brilliantly for me, btw: I haven't seen any problems
> at all yet.
Thank you, thank you. :)
--
Michael Bukin