$LastChangedDate: 2024-04-02 21:01:17 +0200 (Tue, 02 Apr 2024) $

==== BUG REPORTING

For sending a bug report, please drop me an email with a subject line beginning with '[FIM]'. Contact information is in the doc/FIM.TXT file, or in the bottom of this file.

Please DO NOT dare sending bug reports about installation unless you have read the INSTALL file and you are sure all the dependencies on your system have been resolved.

For example "[FIM] problems with resolution 457x3778" would be a good, informative subject for me.

When submitting a bug report, please follow the following rules:

be as much informative as possible when describing your experience.
make a report about the current environment: ( make ; make report ) 2>&1 | gzip > fim.`date +%Y%m%d%H%M`.log.gz and send me the created .tgz file
please give me information about the exact way you experienced the problem
please give me information about the resolution/video mode you are working with; from the bash shell, execute the following and send me the output: ( for f in /sys/class/graphics/fb0/* ; do echo $f ; cat $f ; done )
please send me the config.log file, too (especially if it is a build bug). it is of vital importance! (the config.log file is created by the configure script when you run the ./configure script)
if you are using an ad hoc method for installing (say, the ebuild for the Gentoo Linux ), consider sending the logfile created and named in the error messages.

==== CORE DEBUGGING

Consider debugging with gdb or letting me do this for you! by running

$ ulimit -S -c unlimited

before Fim, in the case Fim should crash, it will generate a file called 'core'. That file, if sent to me (but contact me before sending, because it could be big!), will be helpful for debugging Fim and discover the causes of the crash.

p.s.: please compile the code configuring in this way: ( use ./configure --help='-O0 -ggdb' ) and don't forget to send me the executable file, too :)

==== CONTACT

You can mail your report to dezperado_FOobAr_autistici_Baz_org, by replacing FOobAr with a '@' and Baz with a '.'.

==== KNOWN BUGS

20240402        the search history and the command history are not properly separated
20240322        if using libcaca (-o ca=W) mode with a version older than pre-0.99.beta20 (like 0.99.beta19), pressing continuously 'n' has been observed to lead to an endless loop
20240320        in gtk (-o gtk) mode, pressing 'n' continuously has been observed failing to refresh the images along the way
20240320        in gtk (-o gtk) mode, if terminating via the window system while in the console, --final-commands COMMANDS  will not be executed
20150131        in aalib mode PageUp / PageDown do not seem to work.

20150123

When in framebuffer mode, never, for ANY reason, press C-z (^Z) with the intention of putting fim in the background. It will blind your console and allow only blind typing, which is a very dangerous situation!

20110603 Window splitting is disabled.

==== MINOR ISSUES

20091208        if fim is configured to take input from piped xfig (as xfig-3.2.5b), it could be - passively - subject to
                xfig's buffer overflow on files like:

0 1 2 3
4
1 2 3 4 5 6 7 ZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ

                ( see http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=559274
                  and http://bugs.debian.org/cgi-bin/bugreport.cgi?msg=5;filename=xfig_poc.f;att=1;bug=559274)
                As long as xfig (fig2dev) is accessed through the popen() system call, there will be no harm to fim from xfig crash.
20091015        In 16 bit graphics mode, some lines which should be white actually look blue (seems not fim' fault; happens on both on ARM and x86)
20091013        fim -i < file.pdf # does not work on potentially multipage files
20091009        fim does not support 1,2,4 bit modes
20090328        problems with --std=c++98 in legacy src/FbiStuffJpeg.cpp code (now commented)
20070919        while command suffers from being not interruptible in some cases (when the image is small and not 'panable')
20070912        Debugging it with the framebuffer under gdb hangs the console (even with -p).
20070904        problems could occur with the commands giving output when no std::cout << "" is returned soon after (see Commandconsole.cpp mehods for this..)
20071128        tried:
        Linux host 2.6.19-gentoo-r5 #1 SMP Mon Sep 24 15:08:36 CEST 2007 ppc 7447A, altivec supported PowerBook5,4 GNU/Linux
        and gave problems with the PAGE_MASK macro, originally in /linux/include/asm-i386/page.h :)

20080504        Seems like running the aalib driver under screen maps -3072 units down the arrow keys.
                I do not see this problem in aalib under ssh.
20080127        Under screen, Fim behaves well but it doesn't handle well console switching:
                after a console switch, it doesn't redraw the image if not triggered to do so.
20080106        keys 'Insert,Cancel' are not recognized in the commandline mode.
                This issue will be fixed after the fbi core will be cleaned up and revised.
20070811        On CENTOS 64 bit, i reported linking problems : -ltermcap is required, too.
                I Don't know exactly how to fix this issue.
2007*           Bugs shared with fbi, not fully solved:
                - do not switch console or interrupt the program while it is busy: it could hang your console
                  and force you to make tricks to get back the terminal (like typing blindly through consoles,
                  or trying to get to some X instance via ALT-F7-like sequences, or even restart the computer),
                  and blind typing could be dangerous, if you do now know what you are doing!

==== OUTSTANDING FIXED ISSUES

20110602 fim versions before svn version r373 segfault in 32 bit framebuffer mode, on 64 bit builds

==== FIXED FBI/FBGS ORIGINATING BUGS

Replaced popen with pipe calls : http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=422034, in rev. 213.

Original typo in fbgsh : http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2006-3119 fixed following the patch indicated on http://bugs.gentoo.org/show_bug.cgi?id=141684 :

        -gs     -dSAVER -dNOPAUSE -dBATCH
        +gs     -dSAFER -dNOPAUSE -dBATCH

It could let a PostScript file execute evil code, according to the advisory.

==== UNFIXED EXTERNALLY ORIGINATING BUGS

20090329 On my debian box's aalib, the ASCII art driver leaks 224 bytes of memory. Actually, it happens the same when compiling aalib-1.4.0's aafire.c program against debian aalib and running it in console mode (no X).
After clean aalib recompilation, these problems vanish.

==== SOLVED ISSUES

20110712 in sdl (x11) mode, when fim is spawned in a non-terminal environment, the readline mode does not work in the proper way. detecting this situation may be helpful to disable the readline mode at all in these cases, until a readline-centric fix is found. (seems fixed with r688)