Log of #ghostscript at irc.freenode.net.

Search:
 <<<Back 1 day (to 2016/11/08)20161109 
Draco_ Is this where I can contact MuPDF devs? 00:46.12 
Robin_Watts yup.00:52.26 
  .topic Ghostscript (and MuPDF) development and discussion | live channel log at http://ghostscript.com/irclogs/ | Info at http://www.ghostscript.com and http://www.mupdf.com00:53.14 
  tor8: So, I think I've changed my mind about this usage/intent/OCG stuff.10:37.41 
  I think the right way to do it *is* to have state in the pdf_document.10:37.55 
tor8 ok.10:38.09 
Robin_Watts When we load the PDF document, we set up the default layer state.10:38.17 
  Then we have functions to enumerate the different configs within it, and to select one.10:38.57 
  Then we have an API to pull out the 'display' (or menu) for that config, and to select/deselect elements within that menu.10:39.43 
  Then we can set the usage, and the intent, and everything just falls out fairly neatly, I think.10:43.12 
tor8 okay.10:48.08 
  sounds reasonable enough. so something similar to one of my pastebins from yesterday?10:48.49 
Robin_Watts possibly.10:49.02 
tor8 but with more understanding of the arcane wonders of the PDF spec... :)10:49.12 
Robin_Watts my understanding of this stuff is fluid at best.10:49.36 
tor8 Robin_Watts: so the fz_paint_affine indentation/brace commit, did you approve that in the end?11:01.24 
Robin_Watts tor8: I did.11:12.02 
tor8 Robin_Watts: thanks. I'll tag an RC and make some tarballs.11:12.33 
  and see if I can get my windows build system up and running again after my system failure :(11:12.48 
Robin_Watts tor8: ping17:16.11 
  how would you feel about us having a value in the context (possibly in the AA context) for the minimum thin line size?17:16.40 
  Any lines thinner than that size would be bumped up to it.17:16.49 
  https://github.com/samshadwell/TrumpScript19:47.52 
malc_ hah20:00.14 
Robin_Watts tor8, sebras: Opinions on: http://git.ghostscript.com/?p=user/robin/mupdf.git;a=commitdiff;h=622e0fd98dd33271c653136b4699af756689f86620:31.55 
  mvrhel_laptop, fredross-perry: You too, cos you guys are probably the ones likely to have to use this.20:32.19 
fredross-perry Robin - this ocg thing is PDF 2.0?20:46.12 
Robin_Watts fredross-perry: Nope. 1.522:38.01 
fredross-perry ok22:38.11 
Robin_Watts In 1.5 they define "View" and "Design" as Intents and say "there may be more".22:38.27 
  In 2.0 they say "nope, just those two"22:38.35 
fredross-perry Ah.22:38.42 
Robin_Watts fredross-perry: Basically the idea of this stuff is so you can turn layers on and off in a PDF.22:39.07 
  So you can have say 1 document with lots of different layers, 1 per language.22:39.29 
  Or you could have technical blueprints where you can have different layers for water and electricity and aircon etc.22:39.52 
fredross-perry Right. Is this code a revision to something we already had, or new?22:39.55 
Robin_Watts fredross-perry: We had code before to turn the OCGs on and off for the default cases, but never anything to allow us to configure which ones to turn on more than that.22:40.33 
  That's what I've added now.22:40.39 
fredross-perry So what's a "usage" vs "intent"?22:40.44 
Robin_Watts That's a horrible question.22:41.04 
fredross-perry Or a good question with a horrible answer.22:41.15 
Robin_Watts yeah.22:41.20 
  The spec is unhelpful on this point - or at least the words keep bouncing off my brain.22:41.55 
  Usage is "View" or "Export" or "Print".22:42.04 
  In a viewer, you use "View". If you're exporting to a non-layer supporting format, you are supposed to use "Export", and if you're printing you use "Print"22:42.40 
  So "Print" might have registration marks etc, I guess.22:42.52 
  Usage kinda makes sense.22:43.03 
  Intent is trickier.22:43.10 
  Each OCG has an Intent value (either View, or Design, or All).22:43.36 
  Depending on which Intent is in use (which comes from the selected layer configuration (or OCCD - Optional Content Configuration Dictionary)), that affects whether OCGs are considered for visibility or not.22:44.30 
  I don't quite understand why OCGs alone aren't enough and why they need an Intent too.22:44.46 
  I guess we'll find out if anyone ever actually gives us a file for which it matters.22:45.01 
fredross-perry Hmmm.22:46.17 
Robin_Watts The immediate question is, "is this enough of an API for you to code layer selection in GSView". I hope it should be.22:46.38 
  The code fragment here is a quick test: http://git.ghostscript.com/?p=user/robin/mupdf.git;a=commitdiff;h=67984051b3aa9d826e664f882858e2ade6e3ae7922:47.00 
  That enumerates the names of the layer configs and prints out the "configuration ui".22:47.21 
  (i.e. text labels, checkboxes, radioboxes etc for a given configuration)22:47.39 
  and then toggles some to check it's working.22:47.50 
fredross-perry So what the "configuration ui"? then?22:49.13 
Robin_Watts fredross-perry: Ok, so each PDF file can have multiple configurations.22:57.36 
fredross-perry is "ui" just a meaphor for values and toggles?22:58.05 
Robin_Watts Each configuration suggests how the groups in question should be presented to the user.22:58.15 
  labels, nested groups, checkboxes and radioboxes, yes.22:58.30 
  with the radioboxes being appropriately mutually exclusive.22:58.46 
  so I offer a mechanism to read a) how many "rows" in the UI there are, b) to read the rows out, c) to select/toggle any given row.22:59.35 
  It seemed the easiest way to work without having to require tons of logic in the caller.22:59.55 
fredross-perry I'll understand more when I start to use it, but it makes sense. A viewer could literally enact the "ui" in a real UI.23:01.14 
Robin_Watts I'm open to suggestions for better namings.23:01.25 
  yeah.23:01.28 
  I would absolutely expect a viewer to enable the "ui" in a UI.23:01.55 
  s/enable/enact/23:02.07 
fredross-perry Thanks. This'll make more sense as I try to make use of it. We'll need an array of files that use these, and (probably) a tool for creating our own. Or maybe Adobe does that already.23:05.11 
 Forward 1 day (to 2016/11/10)>>> 
ghostscript.com
Search: