| <<<Back 1 day (to 2014/05/27) | 2014/05/28 |
JoeSaan | Hello, does anyone know a way to convert pages of a PDF that are all black to Greyscale while leaving the color pages alone? | 02:26.13 |
| I have a PDF that's mostly black and white and some genious has every page as CMYK so it costs a fortune to print (as my printer thinks every page is color). | 02:26.55 |
marcosw_ | henrys: for the logs. You recent update to mavericks has stopped clusterpush runs from fetching the users code from casper. I tried logging in but am being logged off immediately. | 04:32.05 |
vinay | Hi.. | 07:49.53 |
Guest43672 | m trying to generate libmupdf.so file, but this file is not generating. i am using code from this link "git clone --recursive git://git.ghostscript.com/mupdf.git" | 07:52.46 |
| throwing some warnings.. | 07:53.08 |
| anybody their?? | 07:53.34 |
| please help me in this.. | 07:53.45 |
kens | Firstly, don;t wait for replies, this channel is logged so you may wait some time for a response | 07:54.13 |
Guest43672 | okay.. | 07:54.27 |
kens | secondly, you were told yesterday that this works for developers here. You are going to have to provide information before anyone cna help you | 07:54.35 |
Robin_Watts | henrys: I have seen no email from Brad. | 08:36.40 |
tkamppeter | Robin_Watts, I have posted also http://bugs.ghostscript.com/show_bug.cgi?id=695266 now. | 08:42.52 |
Robin_Watts | yeah, I saw, thanks. | 08:45.38 |
| henrys: Oh, "Good" Brad. Yes, I saw the email, but I wasn't planning to attend. I've not been involved in the Good stuff so far, so I can't believe I'd have anything to add to the conversation. | 08:50.35 |
pedro_pc | wishes they weren't called "Good" ;) | 09:01.35 |
tkamppeter | I am now on a conference and have forgotten to brin the "git push" URL for GS from home. Can someone e-mail it to me? Thanks. | 09:55.27 |
Robin_Watts | tkamppeter: Presumably you have your private key with you? | 10:01.26 |
| For me I have: golden robin@ghostscript.com:/home/git/ghostpdl.git | 10:01.48 |
| but please remember to clusterpush before pushing. | 10:02.04 |
| If you're not set up to clusterpush then do a git format-patch HEAD~1 to get a patch file out. | 10:02.59 |
| Send that patch to me, and I will clusterpush and can even commit it for you. | 10:03.11 |
tkamppeter | Robin_Watts, I have sent it to you now. Please commit it when it works. | 10:10.36 |
| Robin_Watts, is there somewhere documentation about clusterpush? | 10:11.11 |
Robin_Watts | tkamppeter: A few bits and bobs, but it's not really a public thing. | 10:15.16 |
| Didn't I talk you through it a while ago? | 10:15.23 |
tkamppeter | Robin_Watts, some years ago, and probably I have lost all traces of it. Can you send me an e-mail about it? | 10:23.42 |
Robin_Watts | tkamppeter: I'll see what I can dig up. Might not be this week though. | 10:24.27 |
tkamppeter | Robin_Watts, no problem any time in the next two weeks would be OK for me. | 10:24.59 |
Robin_Watts | tkampeter: (For the logs) tests passed, looks good, pushed. | 10:50.05 |
aksr | hi guys, is it possible to jump to the previously visited page in mupdf? for example, page 16, 53g = page 53, jump_to page 16 | 11:45.13 |
| reading man pages, i would say not, still... | 11:45.32 |
tor8 | aksr: press 'm' before jumping and you can go back to the 'm'arked place with 't' | 11:45.54 |
| that is: "m53g" then "t" | 11:46.26 |
aksr | i know about that | 11:46.36 |
| ..but what if i accidentally go to the beginning or something like that | 11:46.56 |
tor8 | aksr: yeah. we automatically create a mark when clicking on a link | 11:47.15 |
| I guess it would be convenient to automatically create a mark when using 'g' as well | 11:47.26 |
aksr | yes | 11:47.39 |
| just a sec, let me check something | 11:47.47 |
tor8 | aksr: and it should be a pretty easy fix. might as well add it, I've been somewhat bothered by it too | 11:48.40 |
aksr | tor8: will you add it or..? | 11:49.02 |
tor8 | aksr: yeah. | 11:52.57 |
aksr | nice | 11:54.01 |
| is postscript still used to describe something "manually"? | 11:57.32 |
Robin_Watts | tkamppeter: tested and pushed. | 11:57.51 |
tor8 | aksr: patch is on tor/master | 11:57.59 |
| Robin_Watts: several commits for you to review on tor/master if you got a mo | 11:58.12 |
aksr | thank you | 11:58.20 |
Robin_Watts | tor8: will look. | 12:03.15 |
aksr | tor8: looking at your continuous scrolling commit, i have a few comments: | 12:05.37 |
Robin_Watts | tor8: The continuous scrolling one... It's clearly better than we have now, but it's not as nice as the 'smart scrolling' would be. | 12:05.57 |
| or 'smart motion', or whatever we called it in the android app. | 12:06.32 |
aksr | Robin_Watts: if it's not a problem, i would like to say my opinion on that matter :) | 12:06.33 |
tor8 | Robin_Watts: agreed. but barring lots of extra work on the viewer it was that or rip out the continuous pan-drag-scrolling to retain symmetry | 12:06.37 |
aksr | i would like to implement continuous scrolling on other keys, for example J and K, j and k should stay on the same page, why: | 12:07.16 |
tor8 | I'm of half a mind to remove all "continuous" scrolling operations from the desktop viewer, they are mostly confusing | 12:07.58 |
aksr | you don't know when you're at the end of the page, when page is zoomed | 12:08.17 |
| that's why j&k is good | 12:08.26 |
tor8 | there are already separate commands to go to next/prev page and retain the scrolled position/scroll to top/bottom | 12:08.30 |
aksr | J&K is consistent with other applications, too. | 12:08.49 |
| yes,.. | 12:08.59 |
Robin_Watts | tor8: All lgtm. | 12:09.31 |
aksr | i like current mupdf's behaviour with j&k | 12:09.42 |
tor8 | Robin_Watts: what do you think? make only uppercase J&K scroll, or just remove the lot of continuous scrolling? | 12:10.26 |
| sebras: ^ | 12:10.32 |
Robin_Watts | I have no opinion, as I never use it. | 12:10.50 |
tor8 | if I were to redesign it, I'd have three sets of commands: hjkl/arrows to nudge the page view (within a page), back/forward to use the smart navigation, and then hard page flips | 12:12.23 |
| but I'm not really willing to spend a lot more time on this current incarnation of the viewer. I'm itching to rewrite the lot after the next staff meeting instead. | 12:12.51 |
| Robin_Watts: anyway, thanks for reviewing | 12:14.57 |
Robin_Watts | np. | 12:15.07 |
tkamppeter | Robin_Watts, thanks. | 12:29.47 |
kens | Robin_Watts : what's the name of the marker for this jpeg funny ? Just writing a comment.... | 12:50.54 |
Robin_Watts | DNL | 12:51.18 |
| Define Number of Lines | 12:51.21 |
kens | OK thanks | 12:51.22 |
| Robin_Watts : will having the height of the image available in s_DCTD_init() be sufficient ? | 13:03.31 |
henrys | jogux_mac, Robin_Watts : miles will be through the roof when he sees that. | 13:12.21 |
kens | suspected this would be the case.... | 13:12.37 |
jogux_mac | henrys : :-( | 13:12.54 |
| I suspect we don't want to talk about it on IRC... | 13:13.16 |
henrys | jogux_mac: probably best not to | 13:14.54 |
henrys | hates to say but I do like they're web site | 13:15.50 |
| s/they're/their | 13:16.03 |
| pedro_pc: so if you run "strings" on the so do you see familiar constants and such? | 13:20.13 |
| paulgardiner: meeting in 2 hours right? | 13:23.18 |
chrisl | henrys: this development can't be that much of a surprise, given what we already knew of their history...... | 13:25.38 |
paulgardiner | henrys: yep. Will you be in on it? | 13:25.51 |
henrys | paulgardiner: if that's okay with you guys I'd like to listen in. | 13:29.24 |
paulgardiner | Yeah, hoped you would | 13:29.43 |
henrys | chrisl: response on skype | 13:30.54 |
chrisl | tries to remember how skype works...... | 13:31.23 |
pedro_pc | herys: would nice strings like "CompactText foreachLayoutFragment(): vertical text NOT SUPPORTED!" do? :) | 13:32.22 |
| or "UE2FileViewerApp_getNewsfeedUrl version string == NULL" | 13:33.05 |
| this isn't a difficult one to prove ;) | 13:33.25 |
henrys | pedro_pc: no attempt at obfuscation | 13:34.00 |
| I guess | 13:34.04 |
pedro_pc | not that smart - a lot of debugging in it too | 13:35.04 |
| to be honest, they believe they have a god-given right to usethe code regardless of who 'owns' it | 13:35.42 |
Robin_Watts | kens: yes please. | 13:35.55 |
pedro_pc | so they'll not be too fussed about obfuscation imho | 13:36.03 |
Robin_Watts | ownership is a strange concept to the founders. Everyone can invest in their companies, but any money that comes in is theirs to do with as they see fit (massages, first class travel, cars, etc) | 13:37.08 |
pedro_pc | not to mention chauffeur-delivered smoothie-makers | 13:37.34 |
kens | OK I have that working now. In s_DCTD_init the stream_DCT_state ss->data->commopn has a new member 'Height' which contains the image height. Its zero if we didn't get a DecodeParms dictionary with a Height in it (will only ever be the case for PDF input currently). I need to do some testing then I'll throw it over to you if it works | 13:37.45 |
pedro_pc | hmm, sounds odd when you say it out loud ;) | 13:37.51 |
henrys | Robin_Watts, pedro_pc so I guess you guys won't be clicking the careers link on the web page? ;-) | 13:38.38 |
pedro_pc | its always good for a laugh :) | 13:38.58 |
pedro_pc | just finds them fantastic entertainment value these days | 13:39.22 |
Robin_Watts | henrys: I prefer the boring "get paid every month" approach of Artifex to the "will you get paid at all this quarter?" approach at Picsel. | 13:39.39 |
henrys | Robin_Watts: I was just watching the ted talk that money doesn't motivate people, picsel must have seen that too. | 13:40.22 |
| their cloud move makes a lot of sense to me I was going to propose that at the meeting as a step forward with SOT | 13:43.34 |
chrisl | henrys: "money doesn't motivate people" generally only applies once you have enough income to be comfortable..... | 13:45.14 |
henrys | chrisl: miles told me he watched it and found it convincing. I was really spinning up stuff to try and convince him it was all bollocks ... ;-) | 13:52.10 |
chrisl | henrys: there's a lot of truth in it - but we prefer not to let "management" be too aware of it! | 13:53.17 |
pedro_pc | its more the flipside that lack of money or feeling you're being ripped off is de-motivating, and that's often confused :) | 13:54.25 |
chrisl | pedro_pc: the other thing is that often the only form of recognition that "management" understand is financial, so lack of financial gain implies lack of recognition, and that's often demotivating even if you are comfortably off...... | 13:56.25 |
pedro_pc | nods | 13:59.03 |
kens | henrys seems you were not quite correct :-) | 14:14.04 |
pedro_pc | henrys: the android apk and native libraries for entwrx are now in my home folder on casper in ~pete/entwrx-ripoff | 14:16.21 |
| in case they're needed later ;) | 14:16.43 |
henrys | kens: yes I'm surprised | 14:32.46 |
kens | Hmm server going offline, reconnect time | 14:33.15 |
henrys | cool they wrote my security marketing piece http://www.entwrx.com/securewrx.php | 14:38.05 |
pedro_pc | grins | 14:41.12 |
rayjj | henrys: you'll have to fix the spelling, however "customised" and "Internationalisation" | 14:52.23 |
pedro_pc | rayjj: fix? ;) | 14:53.43 |
Robin_Watts | "English (simplified)" rather than "English (traditional)" :) | 14:55.00 |
rayjj | pedro_pc: I bet you folks in the UK get tyred of hearing about the funny way you spell ;-) | 14:55.05 |
| Robin_Watts: I like that :-) | 14:55.17 |
pedro_pc | rayjj> with my surname I can't really complain about people's spelling ;) | 14:55.33 |
| or their abuse of apostrophes | 14:55.54 |
Robin_Watts | pedro: If people misspell your surname on an invoice, mispell their name on the cheque :) | 14:56.16 |
pedro_pc | :) | 14:56.28 |
Robin_Watts | paulgardiner_lap: Have you guys had your meeting yet ? | 14:57.11 |
rayjj | Robin_Watts: so, on dialog pages are they cheque boxes ? | 14:57.14 |
Robin_Watts | nope. | 14:57.23 |
kens | And we don't cheque the output either.... | 14:57.34 |
| At least we cna tell the difference between bank cheques and spot checks | 14:58.24 |
| or bank checks (ie regulator visits)for that matter | 14:58.54 |
rayjj | just as long as Miles sends you the correct type :-) | 14:59.05 |
Robin_Watts | http://www.youtube.com/watch?v=om7O0MFkmpw&feature=kp | 14:59.11 |
| I'm sure I've shared that before. | 14:59.16 |
kens | Wire transfer is the same both ways ;-) | 14:59.17 |
Robin_Watts | pete, paulgardiner_lap: If you haven't had your meeting yet, and it's being done by phone rather than skype, then I recommend: | 14:59.51 |
| http://www.telediscount.co.uk/index2.php | 14:59.56 |
| as that will get you transatlantic calls for 1p a minute just by dialing the prefix number. | 15:00.15 |
| much cheaper. | 15:00.19 |
rayjj | kens: oh, it's not "Wyre transpher" (wouldn't want anyone to confuse it with an action between an electrical supply house and an electrician ;-) | 15:00.41 |
kens | Only in Middle English..... | 15:01.05 |
paulgardiner | Robin_Watts: thanks, although they've given us several UK dialin numbers | 15:01.46 |
Robin_Watts | ah, ok. | 15:02.00 |
rayjj | kens: I liked reading Middle English when in school (I was taking German and so I was one of the few that knew how to pronounce it correctly when reading to the class) | 15:02.23 |
pedro_pc | Robin: yeah, but thanks anyway | 15:02.25 |
kens | Its also true that that klindof spelling was around with Shakespeare, where people pretty much made it up as they went along. Spelling wasn't very formalised | 15:03.14 |
pedro_pc | kens> think we're headed back that way in the next generation or so... | 15:04.04 |
Robin_Watts | zomg, u r soo rite. | 15:04.19 |
pedro_pc | :) | 15:04.34 |
kens | My wife will be happy,she has a 'tangential' approach to spelling.... | 15:05.00 |
Robin_Watts | Helen is a spelling Nazi. I get her to proofread stuff. | 15:05.35 |
rayjj | Robin_Watts: my mom is as well. She can't read books without the errors distracting her. | 15:06.54 |
kens | Melanie and I are both in he 'what do you mean how do spell that ?' camp | 15:07.12 |
| Melanie has been correcting Stella's spelling since she was about 6 | 15:07.28 |
rayjj | the gs knockout stuff was totally messed up. I've gotten the painting working into the TOS, but now the transparency group 'pop' composition is getting it wrong (even though the input is correct compared to the non-knockout case) | 15:16.05 |
| close, though | 15:16.11 |
henrys | pedro_pc: meeting? | 15:33.02 |
el_mendi | How can I run a page trough a pdf-op-filter' | 16:27.03 |
| ? | 16:27.11 |
Robin_Watts | el_mendi: just a mo. | 16:28.57 |
| Are you using the current git source? | 16:29.03 |
el_mendi | hmm... no, I'm using the 1.3 source | 16:29.41 |
| but I guess there is no problem to update it to the current | 16:30.12 |
Robin_Watts | You would need to be using 1.4 at least. | 16:31.03 |
| and 1.4 has some problems, so you really want the current git source. | 16:31.12 |
el_mendi | ok | 16:32.31 |
| and then? | 16:32.48 |
Robin_Watts | then I'll point you at some stuff. | 16:33.27 |
el_mendi | ok thank you | 16:34.34 |
henrys | paulgardiner, pedro_pc : seemed like we spent 3 engineering hours listen to them mumble out loud what _they_ ought to do, maybe I'm cranky from lack of coffee... | 16:42.37 |
pedro_pc | henrys: nope, fair summary | 16:42.52 |
paulgardiner | henrys: I did notice an aspect of that. | 16:43.01 |
pedro_pc | i'm happy that at least they've clarfied some of the unfathomable stuff for me | 16:43.38 |
paulgardiner | ... the mumbling that is, not your crankiness. ;-) | 16:43.54 |
pedro_pc | like "mm, we hacked the relationships into the NOC' and 'yes, that broken behaviour is expected behaviour' ;) | 16:44.10 |
paulgardiner | If we have access to the old account by Friday it will have been worth the time just for that alone. | 16:45.03 |
pedro_pc | absolutely. Wish 'd asked them for access before ;) | 16:45.28 |
paulgardiner | I do feel as though some questions have been answered. | 16:45.44 |
henrys | paulgardiner: yes the old account seems to be the key to much of this. | 16:46.00 |
pedro_pc | yeah, I've cetainly been in many less fruitful meetings before (asuming they do what they said) | 16:46.14 |
kens | Robin_Watts : coping with broken PDF files has cost me some time on this bug (695116) but I'm nearly there. | 16:49.03 |
| hopeing this last cluster push will be clean | 16:49.20 |
Robin_Watts | kens: Ta. Does this have an impact for postscript files too? | 16:58.16 |
| or do we not care? :) | 16:58.27 |
| Postscript files generally give dimensions for the images other than in the image data, right? | 16:59.11 |
chrisl | Postscript is strictly streamable, so...... | 16:59.33 |
Robin_Watts | chrisl: So? | 17:00.20 |
chrisl | Well, a marker that effectively requires you to seek wouldn't be good | 17:00.41 |
Robin_Watts | if images are given dimensions outside the stream, then no seeking is required. | 17:01.10 |
| width height bits_per_sample matrix datasrc image | 17:01.38 |
chrisl | And I'm not sure if it makes a difference, but Postscript doesn't use JPEG, it's just a DCT encoded stream | 17:01.46 |
| Besides, if it happens in Postscript we're screwed - we don't have control over how the dictionary is setup nor the filters | 17:02.24 |
Robin_Watts | chrisl: They do use JPEG. They may not use JFIF :) | 17:02.44 |
chrisl | True.... | 17:02.56 |
Robin_Watts | If we do it by amending the dict, then no, we can't make it work in postscript. | 17:03.21 |
| If we extend the stream interface, we could make it work. | 17:03.34 |
chrisl | No, we still couldn't guarantee it would work.... | 17:03.49 |
Robin_Watts | why not? | 17:03.56 |
chrisl | What if I decode the DCT data into a string, then render the image with the string as a the data source | 17:04.29 |
Robin_Watts | oh, right. | 17:05.26 |
| I see. The separation of data source and image handler is more than momentary in PS. | 17:05.57 |
chrisl | Or, indeed, just decode the data, and don't render it to an image at all...... | 17:06.05 |
Robin_Watts | OK, that's a relief. Kens solution is the best we can hope for then, I think. | 17:06.19 |
chrisl | Did you ever discuss a more general solution with the libjpeg people? | 17:08.59 |
Robin_Watts | I didn't. | 17:09.17 |
| In order to do a more general solution, libjpeg would need to break its API. | 17:09.45 |
| libjpeg says that once you've called jpeg_read_header you have the width/height. | 17:10.09 |
chrisl | It would need to extend it's API, not necessarily break it | 17:10.27 |
Robin_Watts | They would need to read the first scan entirely into memory before they knew that they weren't going to get a DNL. | 17:10.50 |
chrisl | Still seems like a bug - it is the reference implementation, after all | 17:11.33 |
Robin_Watts | it is the defacto implementation, yes. | 17:12.05 |
| It's not endorsed or released by the spec authors in any way, AIUI. | 17:12.31 |
chrisl | I thought they claimed "reference" status, and hence their long refusal to handle broken files | 17:12.50 |
Robin_Watts | I don't believe so. If they dared to claim reference status, someone would have thrown a 65535x65535 file at them and laughed as they failed to cope. | 17:13.51 |
chrisl | Robin_Watts: can you recall: does the Postscript compaction code handle all comments, or does it just strip lines beginning with "%"? | 17:14.40 |
Robin_Watts | I'd assume that it would handle all comments. | 17:15.20 |
| as otherwise trying to extend it to use binary encoding would have been trying to walk before it could run. | 17:15.39 |
chrisl | Okay then, I can leave my comments where they are, then..... | 17:16.25 |
kens | Bum, still one file with a diff :-( | 17:35.17 |
| Robin_Watts : there is no impact for PostScript at all. As chris says, in the general case we can't do anything reliably. | 17:39.42 |
| OK the diffs are a known indeterminism | 17:40.07 |
Robin_Watts | kens: cool. So are you going to commit what you have? Or forward me a patch? | 17:40.58 |
kens | It mihgt be easier if I just commit it, it stand-alone as it is, with no side effects | 17:41.26 |
| It'll take me longer if you want a patch, as I'll have to work out how toi use git format-patch | 17:41.42 |
| A|ny preferences ? | 17:42.15 |
Robin_Watts | commit away. | 17:42.23 |
kens | OK one minute then while I write up a log | 17:42.33 |
| OK Robin_Watts there you go. In s_DCTD_init (or any other convenient place) the stream_DCT_state structure ->data.common.Height now contains the original height of the image,or 0 if it wasn't given (ie PostScript) | 17:54.13 |
kens | is off now, (running away quick before Robin looks at the commit ;-) | 17:56.00 |
| Goodnight all | 17:56.04 |
Robin_Watts | kens: (For the logs) Thanks! | 17:56.47 |
| tor8: 1 commit on robin/master for review. | 19:51.24 |
| henrys: So the latest message on support saying that he'd sent a question on the 23rd of May... | 23:06.03 |
| I can see an email on the 23rd saying "Can I ask a question?" but I can't see any actual question ever being asked. | 23:06.22 |
henrys | Â Robin_Watts well ray responded to hcl in the bug report and marcos was supposed to report back to the customer. I let it go because I assumed HCL got what they wanted. | 23:09.59 |
| I guess it would be best if marcosw followed up | 23:10.42 |
tor8 | Robin_Watts: LGTM | 23:21.08 |
Robin_Watts | tor8: Ta. | 23:21.41 |
tor8 | Robin_Watts: there's also a minor fix on tor/master to prevent some needless reparations | 23:22.13 |
Robin_Watts | tor8: lgtm | 23:23.01 |
tor8 | saw a beautiful view of saturn today | 23:23.04 |
| Robin_Watts: ta. | 23:25.06 |
marcosw | 1 | 23:53.16 |
Robin_Watts | marcosw: Bug 695118 is now fixed if you want to tell customer 870. | 23:57.55 |
| Bug 695116, sorry. | 23:58.09 |
| Forward 1 day (to 2014/05/29)>>> | |