IRC Logs

Log of #ghostscript at irc.freenode.net.

Search:
 <<<Back 1 day (to 2013/02/11)2013/02/12 
Robin_Watts henrys: What do you mean "1 page with a scrollbar" ?00:01.14 
  You mean convert the ENTIRE document to a single page with a scrollbar?00:01.26 
  We can convert a single page to html fast enough that people can jump around anywhere within the document.00:01.52 
  But do you know how long it would take to convert a large doc? Imagine the 1310 pages of the PDF reference manual.00:02.21 
  Not to mention the memory you'd need.00:02.26 
henrys I thought you could parse out the page and convert that to an html page that possibly was larger than the screen00:03.16 
Robin_Watts At the moment, every PDF page is converted to 1 html page, that is displayed in a standard android webview. That does have a scrollbar, at least on my devices, when it's needed.00:03.20 
  henrys: that's exactly what we do.00:03.29 
  We parse a single PDF page into a single HTML page and display it in a webview.00:03.50 
  IF there is lots of text on a page (or you zoom in a lot) that page will be taller than the screen. When you drag the page around, a scrollbar will appear on the right so you can see where you are in the page.00:04.26 
  When you stop dragging, the scrollbar disappears.00:04.43 
henrys let me look again I must be confused00:04.58 
  I am in the midst of doing something else but I'll get back to it before the meeting tomorrow00:05.20 
Robin_Watts ok.00:05.31 
  paulgardiner: hey.11:15.28 
  henrys tried the reflow stuff yesterday, and seemed confused by it.11:15.40 
  I think he was particularly confused by the fact that some pages (the ones with less content in) were much shorter than the screen.11:16.15 
  One possibility would be for us to enforce a minimum height for each page (of the screen height).11:16.44 
  Turns out css has a property for that: "min-height:480px" which we could apply to the div.page11:17.23 
  Or we could inject that value via javascript.11:17.59 
paulgardiner But, wouldn't that give tiny fonts in same cases?11:18.09 
Robin_Watts paulgardiner: It's a *minimum* height for the page.11:18.28 
paulgardiner ah I see11:18.40 
Robin_Watts hence it won't affect font size at all - it will just ensure we have a page the size of the screen.11:18.54 
paulgardiner I can enforce that in the java easily11:18.58 
Robin_Watts Just something to think about. he said he'd play some more before the meeting.11:19.10 
paulgardiner I wondered whether having the pages in a vertical column might be better than the current horizontal row, seeing as they are all different heights but with the same width11:20.21 
Robin_Watts paulgardiner: possibly, but isn't that a more radical chance for our view system ?11:20.52 
paulgardiner yeah, fairly.11:21.05 
  I find none of the oddness shows when you use smart motion (forgotten what you call it)11:21.46 
Robin_Watts I call it smart motion :)11:22.09 
paulgardiner Good name!! :-)11:22.17 
Robin_Watts I put up: http://ghostscript.com/~robin/MuPDF-8.pdf last night.11:22.42 
  I put up: http://ghostscript.com/~robin/MuPDF-8.apk last night.11:22.49 
paulgardiner Hmmm! Text selection is well broken with that one11:25.34 
Robin_Watts really? I thought I tested it.11:26.06 
  Gawd, yes it is. Rats.11:26.36 
paulgardiner Strange for selection to be broken while refow works perfectly. I haven't yet understood how that happens11:27.54 
Robin_Watts paulgardiner: I've been messing with bboxes etc, remember :)11:28.15 
paulgardiner ah of course :-)11:28.26 
Robin_Watts Ahem. I had a stray -1, which meant that the first chars bbox of every span was calculated from random values.11:38.53 
  ok, debug build works.11:41.03 
  paulgardiner: New version up in the same place. This seems to work.11:53.41 
paulgardiner Same name?11:54.42 
Robin_Watts yup.11:54.48 
paulgardiner Yes that looks better11:56.13 
Robin_Watts Fab, thanks.11:56.52 
  I'm gonna have a quick crack at drawing the icons in Xara rather than in Paint.net11:57.20 
  hopefully we'll get better antialiasing.11:57.28 
paulgardiner tor8, Robin_Watts: you prefer function names like fz_annot_type to fz_get_annot_type, but what if that gives the function the same name as the type?12:27.19 
Robin_Watts tor8 objects to 'get's in names.12:28.00 
paulgardiner exactly12:28.17 
Robin_Watts So your function would be fz_annot_text() or something ?12:28.24 
  I'm having trouble visualising it. Can you give me a concrete example ?12:29.11 
paulgardiner It returns the type, so it should be fz_annot_type12:29.24 
  but that is also the type it returns.12:29.44 
Robin_Watts Ah, you have a function to return the type of an annotation. Gotcha.12:29.50 
paulgardiner "type" is literal in all this12:29.52 
tor8 yeah. this is where our naming convention falls down.12:30.18 
paulgardiner I think C allows it, but maybe I still shouldn't do it.12:30.55 
Robin_Watts I'm tempted to agree.12:31.11 
paulgardiner You can certainly have a struct and a typedef share a common name.12:31.11 
tor8 paulgardiner: C has different name spaces for structs and enums12:31.21 
  but typedefs and functions are all in the global name space12:31.30 
paulgardiner Ah ok12:31.35 
Robin_Watts fz_annotation_type? Not nice :(12:31.53 
tor8 the tag (i.e. "foo") in struct foo { } lives in a shared namespace with other struct tags12:32.00 
paulgardiner fz_annot_type_type :-) 12:32.18 
tor8 type_of() :)12:32.25 
paulgardiner fz_type_of_annot?12:32.46 
tor8 or make an exception to the rule and just use get here12:32.47 
paulgardiner I'm thinking get wont hurt too12:33.00 
Robin_Watts fz_annot_type fz_type_of_annot(fz_annot *);12:33.02 
tor8 yeah, type_of_annot looks reasonable to me12:33.15 
Robin_Watts Looks horrid to me :( I'd rather see us use fz_annot_get_type personally.12:33.52 
paulgardiner Maybe "of" then. Ta12:33.55 
Robin_Watts actually, I take that back. I have no preference.12:34.27 
tor8 I'm fine either way. just try to avoid "get" as a filler word when you don't need it.12:34.32 
paulgardiner Ok good12:34.43 
Robin_Watts OK. Reflow icons look much nicer now.13:12.43 
  tor8, paulgardiner: ping14:27.45 
paulgardiner hi14:27.55 
ghostbot hola14:27.55 
Robin_Watts I've just put a new version of the apk up (same URL)14:27.57 
  This one spots things like the first span on a line being entirely made of digits, and guesses that it's a list item.14:28.34 
  So it makes a MUCH nicer job of tables of contents etc.14:28.43 
  but it highlights a problem.14:29.00 
  Currently when we output the spans in a line we don't put spaces in between them.14:29.37 
  actually... the only purpose in having multiple spans on a line now is because there are gaps.14:30.41 
paulgardiner I have examples of words concatenated in other files14:30.43 
Robin_Watts So we should just always put a space out.14:30.53 
  paulgardiner: Can you make said files available please?14:31.06 
paulgardiner yep14:31.29 
  Contents does look stonkingly good now14:32.02 
Robin_Watts has hiccups.14:32.50 
paulgardiner Contents in Grahame's thesis not so good, perhaps because the gaps aren't as big?14:33.41 
Robin_Watts Yeah. I currently have the threshold for making a new span or not based on spacing set to the gap being larger than 1.2 times the 'size' of the text.14:34.46 
  And we insert a space into normal text if the gap is more than 0.2 times bigger than the 'size' of the text.14:35.34 
  So I could reduce 1.2 to 0.8 or something.14:35.51 
  Want to throw me grahames thesis and I'll look at it?14:36.13 
paulgardiner http://intranet.glidos.net/~paul/thesis.pdf14:39.08 
  Hang on. Still copying14:39.30 
  Done14:40.38 
henrys Robin_Watts:sorry yesterday I interpreted what I was seeing incorrectly, but I do see some things -14:44.38 
Robin_Watts henrys: no problem. go for it.14:44.56 
henrys first page of the 1.7 manual - toggle between fill and no fill and it changes pages14:46.00 
Robin_Watts between reflow and non-reflow you mean?14:46.33 
henrys yes14:46.58 
Robin_Watts doesn't for me :(14:47.05 
paulgardiner Hmmm, I can't make it do it either.14:48.38 
Robin_Watts but I'll try it on the tablet rather than the phone in a mo.14:48.38 
  Anything else ?14:48.41 
henrys yes I seem to skip forward to the table of contents page 7 or so when I go back to no reflow14:48.42 
  the small white box14:49.00 
  which is subjective we said we'd wait for others14:49.22 
Robin_Watts You mean the fact you get a small reflowed page on pages with no content ?14:49.49 
  Making pages be at least as high as the screen (or maybe at least 80% high or something) seems reasonable to me.14:50.37 
henrys right14:50.54 
Robin_Watts (or maybe, at least as high as the original page is ?)14:51.01 
paulgardiner Robin_Watts: that last option would might be a pain14:52.27 
Robin_Watts paulgardiner: I'll believe you.14:52.57 
henrys okay it is jumping forward and it is not going back to non reflow when I select non reflow14:53.43 
  I hope michael can try his 1014:54.02 
Robin_Watts henrys: Could you try installing the new version?14:54.21 
  I doubt it will have been solved, but...14:54.35 
  paulgardiner: How big is grahames thesis? It's taking ages.14:55.34 
paulgardiner henrys: could it be a screen calibration problem? What if you hit the button well away from the main screen?14:56.01 
henrys yes by the way we agreed you'd send those to tech mainly for ray and michael14:56.18 
paulgardiner Robin_Watts: not sure. loads of diagrams14:56.23 
Robin_Watts henrys: yes, we did, sorry.14:56.43 
henrys if you do I'll try to harp on the priority it will be good to get feedback14:57.24 
paulgardiner Robin_Watts: Oops! 160 MB14:58.48 
henrys Robin_Watts:no I don't think so.14:59.21 
  the footnote on page 1 of the manual is an image and should disappear?14:59.38 
Robin_Watts The footnote on page 1 of what manual? pdf_reference17.pdf has no footnote for me.15:00.21 
henrys I am looking at PDF 32000-1:2008 first edition15:01.43 
Robin_Watts IS that in our test repo?15:02.05 
  I have it.15:03.35 
  henrys: For me, the footer appears as the first line on the reflown page.15:06.17 
henrys mvrhel_laptop:your ears must have been ringing curious how reflow works on the 1015:08.36 
  Robin_Watts:ah yes that is what I have.15:11.31 
Robin_Watts Page 5 shows something a bit wierd. It's breaking "Transparent Imaging Model" into "Transparent Imagi" "ng Model". I wonder why?15:12.22 
paulgardiner Adobe Reader manages to put the footer at the bottom, but foxit does the same as we do15:14.25 
Robin_Watts At the moment,we don't do any reordering of lines/spans.15:14.55 
  So it's expected behaviour.15:15.04 
  Ordering lines top to bottom would be simple enough.15:16.38 
  but that might break multi-column stuff.15:16.58 
henrys paulgardiner: not a big thing15:33.09 
paulgardiner henrys: yeah, it's not like it makes the reflowed version hard to read15:34.21 
henrys overall I'm really impressed with it. I wish we realized earlier on we could do this.15:35.50 
  I wonder if the reflow button should be highlighted when reflow is enabled15:36.16 
Robin_Watts henrys: Yes. it should.15:36.32 
  or it should change to an alternative form of the icon.15:37.04 
henrys yes just some indication15:37.25 
paulgardiner I tried inverting the colors, but it fills its bounding rectangle to well, so that doesn't look good.15:37.32 
  too well, I mean15:37.45 
Robin_Watts I could do a version with no arrows?15:38.16 
paulgardiner Yes, or we could tint it like the link-enabling icon maybe15:38.46 
Robin_Watts oh, tinting it would be the best.15:38.56 
henrys maybe there should be one icon15:38.57 
Robin_Watts we should probably disable the icons that won't work in reflow mode. Like search and selection etc.15:39.29 
henrys bent arrows when no reflow and straight when reflow - 15:39.40 
Robin_Watts henrys: bent arrows are for reflow.15:39.55 
  They indicate that the line continues below.15:40.09 
henrys well you only want to select reflow when you aren't in reflow mode15:40.24 
Robin_Watts Right. so the bent arrows one would mean "enter reflow mode". And the straight one would mean "enter non-reflow mode"15:40.53 
henrys yes15:41.02 
Robin_Watts but tinting the icon seems the best solution.15:41.04 
henrys okay15:41.33 
henrys resists saying we shouldn't have a button that has no use.15:42.05 
Robin_Watts henrys: Which button is that ?15:42.33 
henrys if you are in reflow you should never need to use the reflow button right?15:43.07 
paulgardiner henrys: we have, but it's also invisible.15:43.12 
Robin_Watts henrys: In reflow mode you hit the reflow button to exit reflow.15:43.44 
  Thus the button has a use.15:43.54 
henrys ah that's why my device is misbehaving.15:44.19 
Robin_Watts and it should change state to say "reflow is active".15:44.37 
henrys I wouldn't have expected that at all and didn't - straight lines no reflow curved done reflow15:44.43 
Robin_Watts The button next to the reflow button is the "Outline view" button.15:45.00 
  That swaps to show the outlines (aka bookmarks) for a document.15:45.19 
  The outline button arguably should only appear if we have outlines in a document :)15:45.34 
  So you were entering reflow mode, then exiting into the outline view? 15:45.59 
henrys yes so my bug is a cockpit error15:46.18 
Robin_Watts Well, it's probably an indication that we need better icons or something :)15:46.40 
mvrhel_laptop good morning15:47.03 
Robin_Watts Possibly, when you change mode, we could do with a popup that says "Reflow mode enabled" or something.15:47.19 
  (plus "Reflow mode disabled", and "Outline view")15:47.59 
paulgardiner outline and reflow icons are two similar15:48.16 
Robin_Watts paulgardiner: That's easy to do, right? Just steal the code you used for "Text copied to clipboard".15:48.24 
mvrhel_laptop so on the nexus 10, I just have a button that I press and it does reflow on the current page15:48.30 
  have to get kids out door . bbiab15:48.39 
henrys Robin_Watts: interesting to see what others think maybe just me being daft15:48.40 
Robin_Watts mvrhel_laptop: Using MuPDF?15:48.44 
paulgardiner Robin_Watts: yes15:48.45 
Robin_Watts henrys: No, we can't disregard your views on this. You only ever get to look at something for the first time, once.15:49.13 
  we all know what the buttons do, because we've been using the app as they've been introduced. You've spotted the obvious thing that they are less than ideal, so we should try to rectify it.15:49.54 
mvrhel_laptop no using the canned app15:50.13 
  Robin_Watts: ^^15:50.16 
paulgardiner Patch to highlight the icon is on paulg/master15:50.20 
mvrhel_laptop I have not tried ours since you been doing reflow work15:50.51 
Robin_Watts mvrhel_laptop: When you get back, I'll get you a new version to try.15:51.36 
henrys paulgardiner: is it best to just work through and skip the meeting like last week?15:59.06 
paulgardiner henrys: whatever suits you.15:59.36 
Robin_Watts If we had a meeting, what would we discuss?15:59.58 
  paulgardiners report makes sense to me. I have nothing to add/suggest.16:00.21 
henrys I'm good too. torn?16:00.31 
  s/torn/tor816:00.39 
paulgardiner I'm happy so long as I'm concentrating on the right areas.16:00.51 
Robin_Watts paulgardiner: Would it be easy to disable the outline icon if there is no icon view?16:01.15 
paulgardiner I've been mostly back to annotation creation last few days, but keeping my eye out for stuff on the reflow side16:01.29 
Robin_Watts s/icon view/outline view/16:01.32 
  actually, I can try that.16:01.38 
henrys does adobe do anything intelligent with the table of contents in the pdf manual wrt reflow?16:02.07 
paulgardiner Robin_Watts: can be done, but may not be easy because of having to enquire asynchtonously16:02.11 
tor8 nope, nothing much to add. it all looks like good progress.16:02.41 
paulgardiner Robin_Watts: no that was wrong. It seems that we have a hasOutline check we can call synchronously16:04.32 
  Robin_Watts: sorry wrong again. It looks to me like we already remove the icon when not needed16:06.54 
Robin_Watts paulgardiner: Oh, marvellous!16:08.09 
  paulgardiner: Hmm. calc.pdf shows the outline icon.16:12.47 
  calc.pdf HAS outlines, but the icon does nothing.16:14.31 
paulgardiner Works mostly. Don't know why calc plays up16:15.27 
Robin_Watts I wonder if it's to do with '/Count -2'16:15.38 
paulgardiner Where's that then?16:16.21 
Robin_Watts 97 0 obj has /COunt -316:17.06 
paulgardiner ah right. Weird16:17.29 
Robin_Watts oh, no, the use of negative ints is documented.16:17.52 
Guest42318 Hi everyone16:26.49 
Robin_Watts Guest42318: Hi16:27.59 
Guest42318 sorry to bother but anyone does hav used anything to detect if a page in a pdf file is colored or not?16:31.02 
Robin_Watts Guest42318: short answer, not trivially.16:31.34 
  but you need to define what you mean a bit more.16:31.52 
Guest42318 i kown, was such a pain in the last month16:32.29 
Robin_Watts Do you mean "The page has no marks specified in color" (i.e. it's all specified in terms of greyscale images/greyscale colorspaces etc)16:32.30 
  or do you mean "The final rendered image ends up all as greyscale" ?16:32.54 
Guest42318 i just need to now if one page is all black or not16:33.03 
  gray scale excluded16:33.12 
henrys Guest42318: http://stackoverflow.com/questions/12299574/ghostscript-color-detection16:33.18 
Robin_Watts Guest42318: That statement only serves to make me more confused.16:33.44 
henrys don't know if we are talking about gs or mupdf though16:34.06 
Guest42318 i don't try any of them yet16:34.29 
  i don't work with c for a long time16:34.48 
henrys but the inkcov device is pretty cool.16:34.58 
Guest42318 i try some c# libs but no success16:35.03 
Robin_Watts henrys: Indeed.16:36.15 
henrys Got my new desk, Sabrina couldn't stand my cinder block supported standing desk. Got the geek desk… pretty nice, now I can sit for half the day and stand half16:36.45 
Guest42318 let's see, I need to build a service that receive a pdf file and answer how many none black pages it contains16:36.52 
henrys Guest42318: who is your customer?16:37.43 
Guest42318 the service is for a printer house ( don't know a better translation ) and is for normal people to use16:39.24 
Robin_Watts Guest42318: So this will run on a website or something? People will upload their PDFs, and you'll then know how many pages need to be printed in color, and how many in greyscale ?16:40.19 
Guest42318 tecnicaly grey will be considered as color16:41.19 
  only black means16:41.26 
  my problem is that all tools i find need some kind of rasterization16:41.56 
Robin_Watts Oh, so you literally mean "full back or full white".16:42.06 
Guest42318 and that is computer extensive and take to much time16:42.15 
Robin_Watts OK, well the first thing you should know is that gs has just moved to the Affero GNU GPL license.16:42.53 
  This means that using gs (or mupdf) on a website means you'll either need to take a commercial license out, or you'll need to abide by the terms of the GNU GPL and make all your modifications available.16:43.42 
  (effectively, under the Affero license, using code on a website is the same as distributing that code).16:44.15 
Guest42318 and we can't buy some kind of licence?16:45.12 
Robin_Watts You absolutely can buy a commercial license.16:45.25 
  I was merely pointing out that such a license might be required - I wouldn't want you to spend hours/days/weeks on this,and then be surprised by needing a license at the end.16:45.55 
Guest42318 oh tks16:46.16 
Robin_Watts BUT... it sounds to me like the inkcov device won't do what you want.16:46.23 
  You could probably implement this by taking mupdf and adding a new 'are_all_the_colors_black_or_white' device to it.16:47.50 
Guest42318 i can't find any buy link on ghostscript.com16:48.25 
mvrhel_laptop Robin_Watts: where do you want me to get the apk?16:48.34 
Robin_Watts Guest42318: There is no 'buy' link. You'd need to talk to our VP of sales, Scott Sackett to discuss prices.16:49.00 
  He'll analyse your needs and come up with a pricing proposal tailored to you.16:49.21 
  mvrhel_laptop: I'm producing a new one at the moment. I'll send an email around in a bit.16:50.01 
  Thanks.16:50.05 
mvrhel_laptop ok. thanks16:50.09 
Guest42318 can u give any range of prices? just to understand the viability?16:57.13 
Robin_Watts I can't mention figures with any accuracy.16:58.24 
  But, would you be prepared to give away any code you write to do this ?16:58.43 
  If so, you need not pay anything.16:58.53 
henrys Guest42318:http://www.artifex.com has contact info on the home page.16:59.27 
  and meeting time ...16:59.44 
kens I'm on holiday next week17:00.47 
  leaving Satirday 16th, back onm the 23rd17:01.01 
henrys mvrhel_laptop: of the screens my vote is just put everything out there and hope to attract new customers, but I do see your point.17:01.02 
mvrhel_laptop henrys: ok. I am fine with that17:01.24 
Robin_Watts mvrhel_laptop: Is there a copyright thing in the headers of the file saying that it's copyright artifex, and released under the Aferro GNU GPL and artifex commercial licenses?17:01.52 
mvrhel_laptop but I wonder how much time do I spend on the email that I got? I can point him to the new code?17:01.58 
  Robin_Watts: good question let me check17:02.20 
henrys anything beyond trivial is a support contract17:02.30 
mvrhel_laptop hmm it just says Copyright 2010-2012 Artifex Software, Inc.17:02.47 
henrys we do want the regular headers.17:03.15 
mvrhel_laptop this is code that someone would never really ship, but a tool to use to create screens17:03.16 
  I will update the header17:03.25 
henrys so release day is thursday17:04.03 
  for ghostscript, what is going on with mupdf?17:04.19 
chrisl Everything's updated and ready to go for GhostPDL....17:04.40 
Robin_Watts There is a mysterious crash in the ios version currently.17:04.48 
  tor8 updated xcode and tried to rebuild, and it looks like some of the magic reference counting is knackered.17:05.14 
  once that's fixed, I think we're in reasonable shape for a release.17:05.40 
chrisl tsk - why would you update tools just before a release? Especially Apple ones - that's bound cause a sh*t storm!!17:06.02 
Robin_Watts chrisl: I'm not sure tor8 had a choice.17:06.15 
henrys is this for the new iOS?17:06.30 
chrisl Ah, the not-so-proverbial "bad Apple"!17:06.42 
Robin_Watts henrys: no idea.17:06.56 
henrys anything else for this meeting?17:08.36 
kens Bug #69357817:08.48 
henrys I can't think of anything but I am not trying hard ;-)17:08.57 
kens Its a cusotmer report and I have no idea what to do with it17:08.59 
  THe PDF file we create opens in Acrobat (and every other reader I have tried) without complaint, but when*pritnting* Acrobat complains about the font17:09.33 
mvrhel_laptop Robin_Watts: when the meeting is over I had a git question for you 17:09.51 
chrisl kens: is it worth using our Acrobat support contract for it?17:10.28 
henrys kens:invalid - what am I missing?17:10.38 
kens henrsay only Acrobat complains, and only when printing17:11.06 
  henrys17:11.10 
  chrisl I was wondering the same17:11.18 
  I cannot see anythign wrong with teh font, and nothing else complains about it, including Acrobat, except when pritning17:11.39 
  If it was not a customer report I would say 'report it to Adobe'17:11.53 
Robin_Watts kens: sounds like a good use of the support contract to me.17:11.58 
henrys agreed17:12.18 
Robin_Watts Hopefully they'll say either "it's a bug in our code", or they'll say "this is what's wrong"17:12.19 
kens In that case ping marcosw17:12.25 
Robin_Watts either of which would suffice to get us moving.17:12.29 
marcosw yes?17:12.30 
kens Robin_Watts : if it only it were that simple17:12.38 
henrys should I text ray do we have something him?17:12.40 
Robin_Watts If you text him, tell him to bring his tablet to work :)17:13.01 
henrys mvrhel_laptop: I'm sure others would benefit from your git question fire away17:13.03 
marcosw our adobe support contract expired. 17:13.11 
kens marcosw see above re bug #693578, can we use the Adobe support contract to find out why Acrobat complains about our PDF file, and only when printing please ?17:13.11 
  ROFL17:13.18 
  THen I have no idea how to tackle this problem17:13.32 
mvrhel_laptop Robin_Watts: so when I try to push from to the ETS git master that you have set up I get the following17:13.43 
Robin_Watts Isn't the support contract like $200 a year or something?17:13.46 
mvrhel_laptop error: insufficient permission for adding an object to repository database ./objectsfatal: failed to write objecterror: unpack failed: unpack-objects abnormal exitTo mvrhel@ghostscript.com:/home/robin/sauce/ETS.git! [remote rejected] master -> master (n/a (unpacker error))error: failed to push some refs to 'mvrhel@ghostscript.com:/home/robin/sauce/ETS.git'17:13.58 
Robin_Watts mvrhel_laptop: Ah, right, yes. That's probably a setup thing on my end.17:14.12 
  just a tick.17:14.15 
chrisl kens: is the file on bug the source or the pdfwrite output?17:15.21 
kens I htink its the source, I made my own output file as I recall17:15.38 
henrys marcosw:so now you have 2 things to do.17:16.10 
kens Yes, they are trying to create a PDF/A file, but it doesn't matter, just running through pdfwrite has hte effect17:16.13 
chrisl kens: OKay, I'm just going to try acroread on Linux17:16.20 
Robin_Watts mvrhel_laptop: Try now ?17:16.24 
henrys we've got to have our support contract right?17:16.29 
kens I will be interestged to hear what that does chrisl17:16.34 
mvrhel_laptop hmm same thing Robin_Watts 17:16.44 
marcosw I will renew the support contract and send the problem file to them.17:17.36 
chrisl kens: same error "Cannot extract...." etc17:17.37 
kens chrisl well at least its consistent17:17.45 
Robin_Watts mvrhel_laptop: And again?17:17.51 
mvrhel_laptop Robin_Watts: Success!17:18.12 
kens marcosw the customer didn't supply a converted file, you need to make one, but its trivially easy to do17:18.19 
Robin_Watts mvrhel_laptop: OK. we can ask tor8 if it ever goes wrong again :)17:18.42 
marcosw I'll contact the customer first, to verify that it's okay to pass the file along (though since it's currently attached to a bug and world readable I'm not sure why).17:18.45 
chrisl kens: does (ha-ha!) preflight give any extra hints?17:18.57 
kens Didn't for me no17:19.04 
chrisl Hmm, so it can "read" the font, but not "extract" it.... how weird17:19.54 
kens just so.17:20.27 
  marcosw I will send you a PDF file that exhibits the problem17:21.30 
marcosw kens: I can duplicate the problem. Acrobat displays the file but it doesn't print.17:21.33 
kens ah OK you can see it, that's fine17:21.44 
marcosw or rather prints with a warning.17:21.50 
kens FOr me it prints, it just complains17:21.52 
tor8 Robin_Watts: missing sticky+group permissions on the git repo?17:22.11 
Robin_Watts yeah.17:22.19 
chrisl kens: But presumably is printing with a substitute font?17:22.24 
henrys ray has been pinged17:22.43 
kens chrisl I don't know to be honest.17:22.44 
henrys alexcher:anything for the meeting?17:23.00 
alexcher henrys: no. I'm looking into 32-bit crashes.17:23.34 
henrys okay17:23.51 
marcosw kens: the problem only appears to occur once. I.e. printing a second time doesn't produce the warning, even closing and reopening the file. Restarting acrobat is required.17:24.29 
kens marcosw that's even more bizarre17:24.41 
marcosw kens: do we want to remove the PDF creator tag so that they don't just tell us to contact ourselves?17:25.29 
kens Hmm, might be a good idea17:25.44 
alexcher kens: what happens after printing to a PS file in Acrobat, converting the result to PDF, and printing again?17:27.35 
kens alexcher, I have no idea.17:27.57 
  I'm not clear how that would help17:28.05 
henrys I do have to leave promptly at the half hour. but I'll be back in an hour or so.17:28.10 
alexcher kens: Acrobat may tweak the font as it likes.17:28.57 
kens Yes indeed it may17:29.06 
  The oiriginal file doesn't have this problem either17:29.25 
chrisl Hmm, the fonts have the full Monotype copyright notice in them....17:30.12 
kens Oh well, I shall leave it in marcos's capable hands and go off. GOodnight all17:32.59 
henrys marcosw:something I forgot - Babysitting an army of monkeys Charlie Miller might be something you'd like (in your copious spare time) anyway that led to the discussion of pairing down the test suit again since the paper discusses that.17:41.00 
  suit suite whatever17:41.16 
  he endorses your original idea of using code coverage.17:44.12 
  got to go17:44.47 
  Robin_Watts: was 8-03 supposed to have a minimum page size or did we decide against that?18:48.30 
Robin_Watts henrys: It hasn't been coded.18:50.16 
  Depending on feedback from Ray and Michael we could add it.18:50.36 
  I know that Raph has a version where text selection (and hence strikeout) is broken, so I'd like to get him a new version soon in light of his most recent email about having passed it on.18:51.19 
henrys otherwise it looks good to me. wish we could do something about the table of contents in the pdf 1.7 manual - I wonder if detecting 4 or more periods would be an adequate heuristic for detecting a table of contents entry18:52.39 
Robin_Watts henrys: If you use the other version of the pdf reference manual, you'll note that it looks much better in recent versions :)18:53.11 
  The PDF32000 looks MUCH crapper18:53.52 
  and in fact has broken typesetting in some places.18:54.02 
mvrhel_laptop Robin_Watts: ok I ran it18:54.11 
  only one comment now18:54.18 
  the native app on the nexus shows the figures too when in reflow18:54.34 
Robin_Watts does it? crumbs. Not seen one that does that yet.18:55.14 
  I'd like to do it, but that's a lot more work.18:55.25 
mvrhel_laptop and when I rotate it jumps out of reflow mode18:55.26 
Robin_Watts so it does.18:56.00 
mvrhel_laptop is that a feature?18:56.36 
Robin_Watts It is now :)18:56.47 
mvrhel_laptop ;)18:56.49 
  Robin_Watts: I take it back about the figues18:59.14 
  I had a docuement earlier that I recall it showing them, but not on the PDF manual18:59.44 
henrys I never heard back from ray I texted him during the meeting19:00.14 
mvrhel_laptop actually this may be a graphic object19:00.17 
  let me check the images19:00.31 
Robin_Watts Images are easier than lineart. but we don't do either currently.19:00.52 
mvrhel_laptop their viewer is sooo slow though19:01.47 
  so yes. their reflow will place images in there19:02.48 
  it is funny to see what happens with patterns and other constructs19:03.05 
Robin_Watts I'd be amazed if they do anything other than drop them.19:03.30 
mvrhel_laptop Robin_Watts: but the reflow with the images may be distracting to some degree19:03.31 
  with some of the complex cases, it has basically taken all images (which may be used in other constructs) and displayed them19:04.33 
  so if you have a pattern template that was an image19:04.49 
  the pattern template is shown19:04.54 
Robin_Watts right, so the images aren't floated in the correct place within the text ?19:04.56 
mvrhel_laptop Robin_Watts: there is that issue too19:05.03 
  anyway I think what you have is good. the only odd thing is the rotate19:05.30 
Robin_Watts I have just fixed that here.19:05.40 
  Uploading a new version in a mo.19:05.47 
mvrhel_laptop great19:05.48 
  I have to head out to meet a friend for lunch19:05.59 
Robin_Watts Thanks.19:06.13 
mvrhel_laptop Robin_Watts: I will test it out this afternoon if that is OK. I think though what you have is looking really good19:06.27 
Robin_Watts Thanks.19:07.48 
henrys Robin_Watts: do you want a file where the native app doesn't delete an image? I have one. I didn't think we were that far along in this exercise.19:08.15 
Robin_Watts henrys: Haven't even started that yet.19:08.34 
henrys okay that's what I thought19:08.50 
  the native app calls reflow reading mode vs. page mode on the nexus 719:09.44 
Robin_Watts I added a popup that says "Reflow mode enabled" and "Reflow mode disabled".19:10.09 
  paulgardiner dislikes it :)19:10.15 
henrys kindle does not reflow at all on the nexus19:10.32 
Robin_Watts henrys: Right. They only support pdf on the nexus app because the kindle devices support pdf.19:11.05 
  and they only do that as a tick list item.19:11.17 
  Hi ray_laptop 19:11.26 
henrys reading mode does seem more user oriented to me.19:12.00 
ray_laptop sorry I missed the meeting. I had a plumbing problem over at our previous house (that is almost not ours anymore), and I don't have any internet there.19:12.48 
Robin_Watts Ok, the version online has been updated with the rotate fix.19:12.52 
henrys but I guess we don't want to be too similar to the other app19:12.53 
  I'd never turned on autorotate19:16.22 
Robin_Watts ok. I'm going to forward this version onto Raph.19:25.27 
ray_laptop mvrhel_laptop: I thought I had already added the gen_stochastic directory to the toolbin/halftone -- I guess it was just laying around in my directory, untracked. Adding it now20:27.07 
  darn -- how do I tell git to not do its 'tab in indent' checking on .sln and .vcproj files (but leave it in place for the .c files)20:40.26 
Robin_Watts ray_laptop: It should only be warning you, not stopping you.20:41.01 
ray_laptop Robin_Watts: it doesn't pull up the editor so I can finish my commit, so that's stopping me20:41.48 
  and the files still show as A on git st20:42.45 
  since I let it run the hooks on the C file to make sure it's clean, do I just use --no-verify ?20:44.08 
  and then do that anytime the .sln or .vcproj file changes (which won't be often)20:45.15 
  with svn there were 'properties' that could be set20:45.42 
  Robin_Watts: -n worked, so I'm just going to push it.20:49.41 
Robin_Watts I don't know, sorry.20:50.29 
ray_laptop Robin_Watts: np. It seemed to work, so I just pushed it. No gripes from the push, so apparently once past my local commit, no more checking is done (or at least no complaining)20:51.56 
  time for lunch20:52.15 
tor8 ray_laptop: it's the core.whitespace setting in git config that gives that error20:54.15 
  there should be a flag to suppress it temporarily for a command, maybe --whitespace=nowarn20:55.25 
  ray_laptop: git commit --no-verify20:56.42 
mvrhel_laptop henrys: are you still there?23:38.10 
henrys yes23:38.41 
mvrhel_laptop I meant to check about the release notes. I wanted to make sure all the color stuff I did ended up in there23:39.33 
  it it too late to send chris something23:39.50 
  henrys^^23:40.02 
henrys no send it as soon as you can though23:40.14 
mvrhel_laptop ok23:40.17 
tklc hey guys: pardon the interruption, I'm doing a research project on CentOS and I need to discover the original author/licensor of ghostscript. Can anyone tell me who that might be?23:57.01 
 Forward 1 day (to 2013/02/13)>>> 
ghostscript.com
Search: