IRC Logs

Log of #ghostscript at irc.freenode.net.

Search:
 <<<Back 1 day (to 2013/02/19)2013/02/20 
mvrhel_laptop henrys: I think ray and I have a handle on what needs to be done on the soft mask. Almost all of it is in the interpreter so ray is going to attempt the required fix00:15.32 
henrys mvrhel_laptop:sounds like you can head back to the windows stuff01:22.53 
mvrhel_laptop henrys: yes. working on the zoom feature now01:37.16 
rooligan Hi! Some developer from MuPDF here? :) Would like to know, if you'd like a translated website10:07.52 
saper rooligan: translatewiki.net?10:09.03 
rooligan saper, I doubt MuPDF uses a translation service like this10:10.12 
saper afair not10:12.02 
Robin_Watts rooligan: hi10:12.12 
rooligan hi Robin_Watts 10:12.25 
Robin_Watts rooligan: Are you volunteering to translate MuPDF for us ?10:12.40 
rooligan Robin_Watts: yes10:12.55 
Robin_Watts rooligan: into what languages?10:13.08 
rooligan Robin_Watts: I'd translate into German10:13.28 
Robin_Watts tor8 handles the website generally. It's a kind offer.10:14.02 
  tor8: ping10:14.32 
tor8 Robin_Watts: hiya10:14.48 
saper is the webpage in git?10:14.48 
Robin_Watts saper: it is not, AIUI.10:15.10 
  tor8: rooligan is offering to translate MuPDF.com into german for us.10:15.23 
saper since then the translation could be done with a simple "pull reqest"10:15.34 
  request10:15.36 
tor8 I have a (local) git for the mupdf.com web pages, just so I can revert if I clobber too much10:16.37 
  translations would be fab, but I think we'll have to redesign the navigation a bit and decide exactly which pages to translate10:17.15 
  Robin_Watts: there's a set of patches for you to review on tor/master as well10:18.52 
Robin_Watts tor8: I will look.10:19.26 
  tor8: In pdf_annot.c11:05.09 
  at around line 298 you introduce an fz_try/fz_catch11:05.25 
  the sole purpose of which is to set link to be NULL if anything there throws.11:05.50 
  We can do that more efficiently by adding fzvar(link); at the top of the function, and then just setting link = NULL; before the pdf_array_get11:06.29 
  No try/catch required.11:06.35 
  pdf_free_annot should be fixed to cope with annot being NULL on entry.11:07.34 
tor8 Robin_Watts: okay, to both.11:07.46 
Robin_Watts And then we don't need the if at line 448ish.11:07.48 
paulgardiner Robin_Watts: That's weird. I didn't know that fz_var could ever be of use in a function not using try/catch.11:08.46 
Robin_Watts oh, you're right of course.11:09.14 
paulgardiner Ah good. I don't have to feel confused all day now. :-)11:09.42 
Robin_Watts actually....11:09.54 
paulgardiner But still I'd imagine you are right that try catch would be unnecessary. I often put them in and then later realise they aren't needed because x = f() doesn't update x when f throws11:11.04 
Robin_Watts Having the try/catch in that function means that if one annot fails to load, we go ahead and load the next anyway.11:11.11 
  paulgardiner: This is one that zeniko (via tor) added.11:11.28 
  It's possible that that is desirable behaviour, in which case the try/catch should stay.11:11.50 
  tor8: sorry to mess you about.11:12.11 
paulgardiner Yeah. Actuall, I thought I added a try/catch in the body of the loop that loads annots... but maybe I'm thinking of a different place11:13.01 
Robin_Watts In pdf_colorspace.c at line 271 you add a second fz_try/fz_catch.11:13.23 
  I think the two can be rolled together.11:13.33 
  Put an fzvar(file) before the try, and then the two merge with no problems.11:14.03 
tor8 Robin_Watts: feel free to clean them up more :)11:14.27 
Robin_Watts ok.11:14.48 
  Interesting.11:15.46 
  zeniko has exposed the restrict stuff in the headers.11:15.58 
  I thought about that, but couldn't see a reason to do so.11:16.07 
tor8 he said he got warnings with VS200811:16.41 
Robin_Watts that's a good reason :)11:16.55 
tor8 oh FFS, now google's added some frigging ads for the chrome book to the new tab page...11:19.48 
  time to switch browser. again.11:19.56 
Robin_Watts I don't follow the change to fz_parse_xml to skip the BOM.11:20.23 
  Oh, because the presence of the BOM means it's already utf8 so no conversion ?11:21.28 
tor8 yeah, but the conversion shouldn't be necessary so I don't think that patch is strictly required11:22.18 
  although we called the conversion function with the wrong 'n' previously11:22.31 
  could probably rewrite the whole conversion thing to be clearer11:22.58 
  in fact, let me go ahead and do that11:25.21 
Robin_Watts runs. will look again when I get back.11:33.25 
chrisl tor8: what's the git incantation to update the thirdparty stuff in the mupdf repo?12:31.15 
Robin_Watts chrisl: git submodule update --init 13:02.42 
chrisl Thanks13:03.25 
  Robin_Watts: And there's this that might benefit from your attention: http://stackoverflow.com/questions/14973019/open-linearized-pdf-on-mupdf-in-android13:05.04 
Robin_Watts chrisl: Answered.13:09.14 
  Thanks.13:09.20 
chrisl Cool - I'm doing stackoverflow duties whilst kens is away.....13:10.18 
Robin_Watts 26 days in and we've gained 548 Android MuPDF users.13:12.47 
  tor8: ping13:41.43 
  Are you redoing that conversion? Or should I work on the patches as they are?13:42.00 
tor8 Robin_Watts: I'm redoing the conversion. if you want to work on the others, just remove that one from the list and rebase away13:42.31 
Robin_Watts will do.13:42.42 
  tor8: Revised version of the commit I was wibbling about:13:53.29 
  http://git.ghostscript.com/?p=user/robin/mupdf.git;a=commitdiff;h=4b9cebf7983cdfe9ee087ce5f2e454ce661fa20413:53.30 
  If you're happy then I'll push all the commits except for the one you're working on.13:53.43 
tor8 Robin_Watts: LGTM.13:54.50 
Robin_Watts Cluster test shows 15 differences... mostly in 2 files. bmpcmping now.14:07.35 
tor8 Robin_Watts: more understandable version of UTF-8 fix on tor/master14:07.53 
Robin_Watts tor8: Seems fair enough.14:10.59 
  Presumably it's illegal to have more than one BOM in a string ?14:11.13 
tor8 Robin_Watts: one would hope so!14:11.59 
  and putting a BOM in UTF-8 text should be a hanging offense14:12.11 
Robin_Watts OK. Nice progressions on those things.14:12.17 
tor8 cool!14:12.23 
Robin_Watts All [pushed.14:13.17 
  tor8: Time for a new tc ?14:16.31 
  rc ?14:16.34 
tor8 I do believe so14:16.43 
Robin_Watts Cool.14:16.48 
  And we should put news.txt on the MuPDF web site this time :)14:17.00 
tor8 I seem to have misplaced my calc.pdf ... Robin_Watts or paulgardiner do you have a link?16:25.59 
Robin_Watts http://ghostscript.com/~robin/calc.pdf16:28.17 
paulgardiner tor8: I just stuck a copy in http://intranet.glidos.net/~paul16:28.20 
tor8 huh, that's funny. dropbox just gave me 23GB extra space for free just for using the dropbox app on my htc desire...16:30.29 
Robin_Watts didn't dropbox used to give 2Gig for free? and would give more by you recommending it.16:34.29 
  Then I think they updated to offer more free, to compete with the likes of Google Drive.16:34.47 
  so probably they just spotted you adding a new app and upgraded you or something.16:35.15 
  ray_laptop, mvrhel_laptop: Morning. How goes the Hunting of the Smask?16:55.21 
ray_laptop Robin_Watts: I discussed an idea I had with mvrhel_laptop yesterday16:56.37 
  Robin_Watts: I am working on the changes now.16:56.56 
Robin_Watts cool.16:57.08 
mvrhel_laptop Robin_Watts: ray_laptop should have a fix soon if everything works as planned16:57.38 
ray_laptop Robin_Watts: basically, the 'SMask' object in the ExtGstate in the interpreter currently has two values -- a 'null' or a 'dict' that when the SMask needs to be rendered.16:58.00 
  Robin_Watts: by setting to other values, we can indicate that "yes" this transparency group currently has an SMask in it's maskbuf so (1) we don't need to re-render it and (2) if a 'gs' op sets a new SMask, we can pop the old one17:00.07 
  at least that's the theory for what I'm trying.17:00.46 
Robin_Watts My memory of this is now so poor that I can formulate no real opinion on that, but it sounds plausible.17:07.12 
  tor8: seen zenikos reply to bug 69363917:16.27 
  ?17:16.29 
tor8 just read now. do you want to fix?18:08.49 
Robin_Watts I think we should fix it. I will do so unless you want to.18:09.43 
  Gah. I hate floating point.18:10.32 
  tor8: nod at this... http://git.ghostscript.com/?p=user/robin/mupdf.git;a=commitdiff;h=ae2b0f60c4d636b6cea32a372494e514409fde7c18:27.00 
  oops. ignore that.18:27.17 
  Better one... http://git.ghostscript.com/?p=user/robin/mupdf.git;a=commitdiff;h=682611d4b5abde493999ec669a465414a18a7c2f18:28.09 
tor8 "many thanks" to who?18:30.29 
  otherwise, *nod*18:30.33 
Robin_Watts "More fixes from zeniko"18:30.55 
  Reworded and pushed.18:32.21 
chrisan Hello, relatively new to compiling on linux... trying to install mupdf on centos 6.3, do I need to copy the thirdparty libraries to a ./thirdparty folder inside the main mupdf source folder, or should it go elsewhere?21:25.05 
  nm, was missing some X11 devel dependencies, all good..21:31.55 
wolf1oo If there are any mupdf devs in here... how does one properly search for a term with an underscore in it with mupdf? As soon as I add in the underscore it doesn't correctly match anything21:37.47 
chrisan Am I missing something with saving pdf to image? `mudraw -o %d.jpg -m pages02_03.pdf` doesn't seem to output files for me, but its "doing" something. http://pastie.org/private/tfud7zgznk7vorit04wgjw21:51.21 
  OK.. apparently jpg is not working, I can do ppm or png. Any pointers :) ? So far I'm having good luck just typing here and eventually figuring stuff out so perhaps my success will continue22:47.14 
wolf1oo chrisan: read the man pages. The supported image formats are pgm, ppm, pam, and png22:49.23 
chrisan oh.. derp. well thank you22:49.52 
wolf1oo haha no problem22:50.01 
 Forward 1 day (to 2013/02/21)>>> 
ghostscript.com
Search: