| | <<<Back 1 day (to 2018/09/19) | 20180920 |
| sebras | tor8: ugh. | 02:51.34 |
| | tor8: the page singleton list implementation is broken. the prev pointer might not be set if the page is not fully loaded. | 03:09.11 |
| | tor8: I have a commit for that. | 03:09.20 |
| | Robin_Watts: I need help in resolving the compilation errors I get from watts when testing this commit. | 03:17.11 |
| | Robin_Watts: is the problem that the command line states MuPDF.sln while the file is really called mupdf.sln? at least the latter file is present in platform/win32 in the repo, so msgc ought to be able to find it..? | 03:20.40 |
| vtorri | hello | 03:22.13 |
| mubot | Welcome to #mupdf, the channel for MuPDF. If you have a question, please ask it, don't ask to ask it. Do be prepared to wait for a reply as devs will check the logs and reply when they come on line. | 03:22.13 |
| sebras | Robin_Watts: seems like win32 path case sensitivity depends on the file system. | 03:22.25 |
| | vtorri: hello | 03:22.28 |
| vtorri | is there API break between mupdf f1.12 and 1.13 ? | 03:22.35 |
| sebras | vtorri: is there any special function you're thinking of? | 03:23.10 |
| vtorri | no | 03:23.16 |
| sebras | vtorri: I don't remember off hand. I don't remember any, but there might be. | 03:23.53 |
| vtorri | it's just that i use mupdf in my project and i have to deal with linux distro which provide mupf 1.12 or 1.13 | 03:23.58 |
| | usually archlinux has latest mupdf code | 03:24.50 |
| | ok, thank you | 03:25.03 |
| sebras | vtorri: ok, if you stay awat from newly added and internal APIs the likelihood is lower that you will encounter API breaks at least. | 03:25.52 |
| | s/awat/away/ | 03:26.01 |
| vtorri | thank you | 03:28.31 |
| Robin_Watts | sebras: Yes. | 09:22.05 |
| sebras | Robin_Watts: so... it is due to case sensitivity? | 09:22.57 |
| Robin_Watts | I believe Ray's problems are, yes. | 09:23.26 |
| | I will see if I can set up a case sensitive filing system on a USB stick later. | 09:23.58 |
| sebras | Robin_Watts: wait... this one failed on watts, not on ergs. | 09:29.21 |
| Robin_Watts | Try it again now ? | 09:31.21 |
| tor8 | sebras: sebras/master LGTM | 09:45.50 |
| Robin_Watts | reruns it for sebras. | 09:55.44 |
| | and again... | 10:15.11 |
| sebras | Robin_Watts: now it works, thanks! | 10:38.07 |
| Robin_Watts | No problem. I am just updating the twiki on what needs to change for ergs too. | 10:38.37 |
| sebras | tor8: pushed. | 10:42.20 |
| Robin_Watts | Did someone update freetype recently? | 10:56.48 |
| sebras | Robin_Watts: I did. | 10:56.55 |
| Robin_Watts | 4>Creating library... | 10:57.02 |
| | 4>ftbase.obj : warning LNK4006: _FT_Get_Font_Format already defined in ftfntfmt.obj; second definition ignored | 10:57.04 |
| | 4>ftbase.obj : warning LNK4006: _FT_Get_X11_Font_Format already defined in ftfntfmt.obj; second definition ignored | 10:57.05 |
| | Should we really be including both ftbase and ftfntfmt in the same build ? | 10:57.37 |
| sebras | oh, these are only wornings...? | 10:57.40 |
| | warnings. | 10:57.50 |
| Robin_Watts | yes. nothing to ponic about :) | 10:57.54 |
| sebras | Robin_Watts: I didn't alter the list of included c files though. | 10:58.36 |
| | Robin_Watts: ftbase.c does an #include "ftfntfmt.c" | 10:59.13 |
| Robin_Watts | So I should remove ftfntfmt.c from the list of files? Simples! | 11:00.10 |
| sebras | the same goes for ftlcdfil.c | 11:00.12 |
| Robin_Watts | That's already not there, | 11:00.49 |
| sebras | Robin_Watts: it is still there in Makethird. | 11:01.13 |
| Robin_Watts | That would be an ecumenical question. | 11:01.28 |
| | What is the procedure for fixing stuff in jbig2? | 11:02.23 |
| | Does that involve commiting to gs and getting it reexported? | 11:02.34 |
| sebras | yes. | 11:02.51 |
| | commiting in gs, waiting a day for it to appear in the extracted repo (or prompt tor8), and then do an update in mupdf. | 11:03.28 |
| | why? | 11:03.32 |
| | have you seen an issue:? | 11:03.37 |
| Robin_Watts | signed/unsigned warnings is all. | 11:04.55 |
| sebras | if you pastebin me the change I'll fix it. | 11:05.28 |
| Robin_Watts | I have no change yet :) | 11:06.26 |
| sebras | Robin_Watts: but you have a warning? | 11:06.45 |
| Robin_Watts | I do, but not so I can paste at the moment. Give me a mo to rebuild. | 11:06.59 |
| sebras | Robin_Watts: I believe you may be able to remove ftgasp.c too. at least linux builds without it. | 11:09.32 |
| | and ftsynth.c and fttype1.c | 11:09.48 |
| | not sure what impact any of these removals have though. | 11:09.59 |
| Robin_Watts | 4>c:\artifex\mupdf.git\thirdparty\jbig2dec\jbig2_generic.c(391) : warning C4018: '<' : signed/unsigned mismatch | 11:10.46 |
| | 4>c:\artifex\mupdf.git\thirdparty\jbig2dec\jbig2_generic.c(560) : warning C4018: '<' : signed/unsigned mismatch | 11:10.47 |
| | 4>c:\artifex\mupdf.git\thirdparty\jbig2dec\jbig2_generic.c(671) : warning C4018: '<' : signed/unsigned mismatch | 11:10.49 |
| | 4>c:\artifex\mupdf.git\thirdparty\jbig2dec\jbig2_generic.c(734) : warning C4018: '<' : signed/unsigned mismatch | 11:10.50 |
| | 4>c:\artifex\mupdf.git\thirdparty\jbig2dec\jbig2_halftone.c(67) : warning C4018: '<' : signed/unsigned mismatch | 11:10.52 |
| | 4>c:\artifex\mupdf.git\thirdparty\jbig2dec\jbig2_halftone.c(482) : warning C4018: '<' : signed/unsigned mismatch | 11:10.53 |
| | 4>c:\artifex\mupdf.git\thirdparty\jbig2dec\jbig2_halftone.c(483) : warning C4018: '<' : signed/unsigned mismatch | 11:10.55 |
| | 4>c:\artifex\mupdf.git\thirdparty\jbig2dec\jbig2_halftone.c(514) : warning C4018: '<' : signed/unsigned mismatch | 11:10.56 |
| | 4>c:\artifex\mupdf.git\thirdparty\jbig2dec\jbig2_halftone.c(515) : warning C4018: '<' : signed/unsigned mismatch | 11:10.58 |
| | sebras: The first 3 commits after your one should be good to go: http://git.ghostscript.com/?p=user/robin/mupdf.git | 11:12.46 |
| sebras | Robin_Watts: Re: "Be consistent in use of #if FZ_ENABLE_..." the commit message seems to have been interrupted by dinner..? | 11:14.09 |
| tor8 | Robin_Watts: "Be consistent in use of #if FZ_ENABLE_..." LGTM | 11:16.32 |
| sebras | tor8: how do I parse "Using #ifdef FZ_ENABLE_ means we build code in, even if we have"? | 11:17.37 |
| tor8 | Robin_Watts: ftfntfmt.c and ftlcdfil.c are both included by ftbase.c so should be removed from Makethird too | 11:17.50 |
| | sebras: the message looks truncated yes, but the code is good and the intent is clear even if the message is aborted :) | 11:18.25 |
| | there's a FIXUP commit for the freetype stuff on tor/master | 11:18.35 |
| sebras | tor8: ok. | 11:18.39 |
| tor8 | I expect Robin will want to fix the message though. | 11:19.01 |
| | sebras: the non-WIP commits on tor/master could do with a new review | 11:19.16 |
| sebras | tor8: mupdf-x11 builds find here without ftbbox, ftdewbug, ftgasp, ftsynth and fttype1. | 11:19.31 |
| | tor8: perhaps it was the case that they were necessary at one point, but not anymore. | 11:19.56 |
| tor8 | sebras: most of the symbols in those files are disabled with the scripts/freetype/slimftoptions.h and slimftmodules.h defines | 11:20.16 |
| | it may well be we can skip building those files altogether | 11:21.44 |
| | but then things have broken in the past when doing that ... freetype isn't always 100% accurate in their ifdeffery when adding functions | 11:22.13 |
| sebras | right. | 11:22.26 |
| Robin_Watts | sebras: ha! the next line was "#defined FZ_ENABLE_WHATEVER to be 0, as we do in config.h" | 11:24.58 |
| | and of course git thought it was a comment line. | 11:25.09 |
| | I'll not fiddle more than I need to with files being included/excluded. | 11:26.32 |
| tor8 | sebras: yes, ftbbox, ftdebug, ftgasp, ftsynth, and fttype1 are not required | 11:28.01 |
| | but I still think we should leave them in, because someone linking to us may also use freetype | 11:29.04 |
| | and we shouldn't cripple them unneccessarily | 11:29.17 |
| sebras | right. | 11:34.00 |
| Robin_Watts | So, any opinions on the next 2 commits? | 11:46.03 |
| | (or did I miss that?) | 11:46.15 |
| sebras | Robin_Watts: yes, tor8 provided a FXIUP commit on tor/master | 11:46.52 |
| | for "Fix harmless "repeated symbol" warning in windows build." | 11:46.58 |
| Robin_Watts | You want me to roll that into mine? | 11:51.02 |
| sebras | Robin_Watts: that seems reasonable to me, yes. | 11:51.29 |
| Robin_Watts | I'll do that after lunch. And the other commit? | 11:54.19 |
| moolc | sebras: thanks for 94d6166428a679baa5a34fc5faa18a2aa26cee4a made me stop pulling my hair | 11:56.51 |
| sebras | moolc: I have no hair, so it didn't affect me that much. :) | 11:59.12 |
| | moolc: but yeah, I ought to have caught that myself when reviewing it. I was quite cautious, but I couldn't see the path that caused the issue. | 11:59.39 |
| | Robin_Watts: I built with -Wsign-compare before/after your patch to see what got fixed. | 12:00.25 |
| | Robin_Watts: several were fixed, but the one in pdf-xref.c causes the ?: operator to generate signed in one case and unsigned in the other. I'm not sure if we care, but it did generate a new warning. | 12:01.16 |
| moolc | sebras: well... thanks to https://www.moser-profiline.de/en/produktfinder/_leaf/2921 i have very little hair too... then again it has been a couple of days since i've been my own barber.. so enough have grown to make things unpleasant ;) | 12:01.43 |
| sebras | Robin_Watts: actually it seems like we might have 150 | 12:01.50 |
| | about 150 of thesewarning in the entire codebase. | 12:02.03 |
| paulgardiner | Are there now better ways to build for VS2015 than when we first used mupdf within Windows MuSO? I have a copy of the project files in muso that have been updated from the vs2005 ones to work with vs2015. I've just resynced the submodule and of course any changes to the vs2005 project files aren't represented within my vs2015 copies. | 12:07.00 |
| tor8 | paulgardiner: no. you'll need to reimport I think. | 12:14.31 |
| | we did experiment with premake to generate VS project files | 12:14.41 |
| | but that's not pushed to master (and I'm not sure if we're going to, either) | 12:15.02 |
| sebras | Robin_Watts: ok, I prepared a jbi2dec patch for you on ghostpdl:sebras/master using that I expect that you don't get any more signed/unsigned mismatches? | 12:26.07 |
| Robin_Watts | sebras: I will try it in a mo. | 12:59.20 |
| | paulgardiner: I am looking at a script to update from VS2005 to VS2017 for gsview at the mo. | 12:59.56 |
| | it may be possible to do the same for you. | 13:00.03 |
| | I'll let you know how I get on. | 13:00.09 |
| paulgardiner | That would be just wonderful | 13:00.20 |
| Robin_Watts | paulgardiner: What configurations do you build? Do you make any changes? | 13:19.20 |
| | mvrhel adds a couple of predefines, which means I've added DebugGSView and ReleaseGSView configurations to the basic MuPDF solution. | 13:20.12 |
| paulgardiner | I think just Debug and Release with no changes. | 13:26.13 |
| Robin_Watts | paulgardiner: OK. | 13:32.50 |
| paulgardiner | I have my project files in a directory outside the mupdf submodule. bin2coff.vxproj seems determined to create bin2coff.exe inside the mupdf submodule and then other parts of the build can't find it. | 14:55.56 |
| | Ah, sorted I think | 15:16.30 |
| | Now complaints of jconfig.h missing. Is that also a generated file/ | 15:18.27 |
| | Forward 1 day (to 2018/09/21)>>> | |
Log of #mupdf at irc.freenode.net.