From karl at freefriends.org Wed Dec 1 00:27:19 2021 From: karl at freefriends.org (Karl Berry) Date: Tue, 30 Nov 2021 16:27:19 -0700 Subject: [tex-k] Valgrind: Kpathsea library leaking memory? In-Reply-To: Message-ID: <202111302327.1AUNRJ3I011037@freefriends.org> Is the problem with kpathsea library or with valgrind? Kpathsea. In the latter case, how to make valgrind not to report errors concerning kpathsea? I don't know. Maybe you should ask on a valgrind list, after seeing if the answer has already been given on the net. I'm sure you are not the first one to want to suppress warnings for bunches of functions. A cursory look at valgrind's man page turns up --ignore-fn and various other suppression-related things. Good luck. -k From Laurence.Finston at gmx.de Wed Dec 22 12:38:36 2021 From: Laurence.Finston at gmx.de (Laurence Finston) Date: Wed, 22 Dec 2021 12:38:36 +0100 Subject: [tex-k] Minor bug report Message-ID: mft --help --> Usage: mft [OPTION]... NAME[.mf|.mp] Translate MFNAME to TeX for printing, using the mftmac.tex (or mptmac.tex) macros. Output goes to basename of NAME extended with `.tex'. -change=CHFILE apply the change file CHFILE as with tangle and weave -metapost assume NAME is a METAPOST source file -style=MFTNAME use MFTNAME instead of plain.mft (or mplain.mft) (this option can be given more than once) -help display this help and exit -version output version information and exit Email bug reports to tex-k at tug.org. *********************************************** Line 2 is somewhat confusing: NAME appears in line 1 but both NAME and MFNAME are referred to in line 2. It would seem that MFNAME is NAME.mf and MPNAME would be NAME.mp but I've never heard of a convention like this before. If it's correct, please ignore. From karl at freefriends.org Wed Dec 22 23:16:50 2021 From: karl at freefriends.org (Karl Berry) Date: Wed, 22 Dec 2021 15:16:50 -0700 Subject: [tex-k] Minor bug report In-Reply-To: Message-ID: <202112222216.1BMMGoEK028014@freefriends.org> Hi Laurence, Line 2 is somewhat confusing: NAME appears in line 1 but both NAME and MFNAME are referred to in line 2. Indeed. Clearly I failed to update the help message completely when .mp files came into play. I will fix. Thanks for the report. -k From igor.liferenko at gmail.com Thu Dec 30 03:51:51 2021 From: igor.liferenko at gmail.com (Igor Liferenko) Date: Thu, 30 Dec 2021 09:51:51 +0700 Subject: [tex-k] DVItype in CWEB Message-ID: Hi all, DVItype is now available in CWEB: https://github.com/igor-liferenko/dvitype $ make $ ./dvitype input.dvi output.typ --Igor From andreas_tex at freenet.de Thu Dec 30 11:04:46 2021 From: andreas_tex at freenet.de (Andreas Scherer) Date: Thu, 30 Dec 2021 11:04:46 +0100 Subject: [tex-k] DVItype in CWEB In-Reply-To: References: Message-ID: > DVItype is now available in CWEB > https://github.com/igor-liferenko/dvitype Nice effort! A few comments: * Why does 'ctangle' come from '/bin' in the two Makefiles? * Could 'web2w' somehow be modified to break long lines to a more decent length? If I use 'cweave 4.6' (not from TL) it bails out with several complaints like ! Input line too long. (l. 1729) For example, the list of array variables in section 72 of dvitype.web gets stolidly expanded to a really long line in the CWEB form that really should be broken into separate lines. Similarly, the convoluted switch/case statement in section 75 exceeds 'standard' cweave's capabilities (l. 17{72,74,77,79,81,84}). * Several WEB sections should be eliminated in dvitype.ch: ! Never used: ! Never used: <|pool_ptr| initialization> ! Never used: <|str_pool| initialization> ! Never used: <|str_ptr| initialization> ! Never used: <|str_start| initialization> Happy New Year 2020, too! ;-) Andreas