Re: Sangband is back.
- From: The Wanderer <inverseparadox@xxxxxxxxxxx>
- Date: Sat, 07 Apr 2007 13:41:29 -0400
I'm not sure if this should go privately (since technically it's a bug
report), or if publicly is appropriate (since others may well have
encountered the same problem). Posting publicly, because the last time I
sent a private bug report to a maintainer of a variant I got no response
at all.
Leon Marrick wrote:
Sangband is back.
For pretty much all of last year, my health fluctuated between poor
and downright ugly; pretty much nothing got done between February and
December (except for an alpha-quality demo written in August that I
should have shared but didn't). Things got so bad, in fact, that on
several occasions I gave serious consideration to packing it in and
letting Sangband go to someone who could actually maintain the thing.
But I didn't. And, starting early this year, things have gotten
better. Enough better that a new release of Sangband is here. It's
the long-awaited Version 1. I'm not going to post up a changelog but
instead let you imagine what it took for me to be so bold.
Though, of course, anyone capable of reading the sources can run a diff
and see for themselves.
(I can read the sources, but not nearly well enough to tell what effect
the changes will have on actual gameplay in most cases. From what little
I can tell, they mostly seem to each fall into one of two categories:
either "minor tweak" or "extensive surgery".)
Source (for Linux, Windows, etc.):
http://runegold.org/sangband/files/sang100src.rar
This does not compile under Linux for me, using Makefile.std, even after
I copy the files back from the unix/ subdirectory (since that is easier
than modifying the references in the Makefile appropriately). main-gcu.c
fails on undefined references to screen_rows, which is not defined
anywhere in the source; I would guess that it is supposed to be defined
in ncurses.h, but the copy on my system does not define that symbol.
In the last version I tried compiling (0.9.9), which I believe I was
able to compile successfully, screen_rows is referenced in all of:
birth.c
cave.c
changes.txt
cmd3.c
cmd4.c
cmd6.c
dungeon.c
externs.h
files.c
info.c
main.c
main-crb.c
main-gcu.c
main-gtk.c
main-ibm.c
main-mac.c
main-win.c
main-x11.c
monster1.c
obj_make.c
skills.c
store.c
talents.c
util.c
variable.c
wizard2.c
In the sources extracted from that RAR file, it is mentioned only in:
changes.txt
mac/main-crb.c
mac/main-mac.c
unix/main-gcu.c
unix/main-gtk.c
unix/main-ibm.c
unix/main-x11.c
I expect that all of these main-*.c files would fail in exactly the same
way as main-gcu.c My guess would be that the Windows port has been
modified to account for the fact that the variable is no longer
available, but the other ports have not.
I also tested compiling with makefile.sdl (any reason for the unusual
capitalization?), since main-sdl.c is the only main-*.c file other than
main-win.c to not reference screen_rows. That does compile successfully,
but fails in the linking stage with various missing FreeType symbols.
This may potentially be a problem with inconsistent SDL and/or FreeType
libraries on my system.
--
The Wanderer
Warning: Simply because I argue an issue does not mean I agree with any
side of it.
Secrecy is the beginning of tyranny.
.
- Follow-Ups:
- Re: Sangband is back.
- From: powera
- Re: Sangband is back.
- References:
- Sangband is back.
- From: Leon Marrick
- Sangband is back.
- Prev by Date: Re: Sangband is back.
- Next by Date: Re: Sangband is back.
- Previous by thread: Re: Sangband is back.
- Next by thread: Re: Sangband is back.
- Index(es):
Relevant Pages
|
