Review: UnAngband

[TJA]

PART I.
Compilation

unzip unangband-062-wip5b-src.zip
cd unangband-062-src
cd src

Edit Makefile.std

Confige Makefile.std to say:
NORMAL_CFLAGS = -Wall -pipe -O2 -D"USE_GCU" -D"USE_NCURSES"
LIBS = -lncurses

make -f Makefile.std
make -f Makefile.std install
cd ..

Make produces a strange error again:
util.o(.text+0x213): In function `path_temp':
util.c: warning: the use of `tmpnam' is dangerous, better use `mkstemp'

Also a strange error:

xtra1.c: In function `calc_bonuses':
xtra1.c:3694: warning: comparison is always false due to limited range of data type

./unangband

Works, but needs knowledge of what to change.

Like in Portralis, i needed to fiddle around to find the right seetings - at first i got bad results because no output modules (use via "-m") could be found.
But then, i got it

First impression: Questions could just be answered with <Return>, which is great IMHO

[Bandobras]

Originally Posted by TJA

Confige Makefile.std to say:
NORMAL_CFLAGS = -Wall -pipe -O2 -D"USE_GCU" -D"USE_NCURSES"
LIBS = -lncurses

What was the problem with the original Makefile.std? It works for me...

Originally Posted by TJA

Make produces a strange error again:
util.o(.text+0x213): In function `path_temp':
util.c: warning: the use of `tmpnam' is dangerous, better use `mkstemp'

Also a strange error:

xtra1.c: In function `calc_bonuses':
xtra1.c:3694: warning: comparison is always false due to limited range of data type

These are not errors, but warnings, as you can read...

Originally Posted by TJA

Like in Portralis, i needed to fiddle around to find the right seetings - at first i got bad results because no output modules (use via "-m") could be found.

What was the problem, exactly?

[TJA]

Originally Posted by Bandobras

What was the problem with the original Makefile.std? It works for me...

At first, could not find termcap:

gcc -Wall -pipe -O2 -D"USE_X11" -D"USE_GCU" -D"USE_NCURSES" -o unangband birth.o cave.o cmd1.o cmd2.o cmd3.o cmd4.o cmd5.o cmd6.o info.o dungeon.o files.o generate.o init1.o init2.o load.o melee1.o melee2.o monster1.o monster2.o object1.o object2.o randart.o save.o spells1.o spells2.o store.o tables.o util.o variable.o wizard1.o wizard2.o xtra1.o xtra2.o borgdumb.o z-form.o z-rand.o z-term.o z-util.o z-virt.o main-cap.o main-gcu.o main-x11.o main-xaw.o main.o maid-x11.o -lX11 -lncurses -ltermcap -L/usr/X11R6/lib
/usr/lib/gcc-lib/i586-suse-linux/3.3.5/../../../../i586-suse-linux/bin/ld: cannot find -ltermcap
collect2: ld returned 1 exit status

Removing -ltermcap was enough to get it compiled and installed - i just wanted to delete X11 as well ....

Originally Posted by Bandobras

These are not errors, but warnings, as you can read...

Now ... normaly i would say "YES".
But THIS warning says something about an algorithm and how it may work or not.
So, this is more a content-warning - so maybe more interesting then a syntax warning.

Originally Posted by Bandobras

What was the problem, exactly?

It was not straight forward for me to reduce the compilation to curses/ncurses. Some time (4 or 5 compiles) i got a binary that offered NO output version for the "-m" argument. The --help option did list nothing to choose from ...
After i got it, it was easy of course.

I love "configure" - should have a look at how to create such a thing

[Bandobras]

Originally Posted by TJA

At first, could not find termcap:

Ok, now I see your point. I have no clue if termcap is of any use here. Does anybody know if "-ltermcap" should be removed from the makefile or rather that TJA should better install termcap or a version of ncurses with termcap.h instead (my Debian version has it)? At my machine it seems to work OK without termcap, but perhaps it only shows its usefulness when run remotely on some strange terminal?

BTW., why don't you use the X11 version?

[TJA]

I have termcap - it will not be found:

rpm -qa | egrep "(termcap|curses)"
ncurses-devel-5.4-68
termcap-2.0.8-879
ncurses-5.4-68

There does not seem to be a termcap-devel package available, as everything is included in this package already:

/usr/include/curses
/usr/include/curses/curses.h
/usr/include/termcap
/usr/include/termcap/termcap.h
/usr/lib/curses
/usr/lib/curses/libcurses.a
/usr/lib/curses/libcurses.so
/usr/lib/libcurses.so.1
/usr/lib/libcurses.so.1.0.0
/usr/lib/libtermcap.so.2
/usr/lib/libtermcap.so.2.0.8
/usr/lib/termcap
/usr/lib/termcap/libtermcap.a
/usr/lib/termcap/libtermcap.so

Still, the Makefile.std does not work with it ...

No idea why :-O

[takkaria]

Originally Posted by Bandobras

Ok, now I see your point. I have no clue if termcap is of any use here. Does anybody know if "-ltermcap" should be removed from the makefile or rather that TJA should better install termcap or a version of ncurses with termcap.h instead (my Debian version has it)?

-ltermcap isn't necessary, AFAICT. V doesn't use it at all anymore -- though it is required if you use main-cap.c (something else V doesn't have anymore).

[TJA]

Will you invest time to clean up the Makefiles to reflect the actual requirements?
Or is that nothing you plan to do?

[Bandobras]

I've just removed -ltermcap in SVN. Thanks for pointing this out, TJA and thank you for the tip, takkaria.

BTW., I've just noticed I didn't have -ltermcap in my local copy of Makefile.std already for some time...