| <<<Back 1 day (to 2014/02/26) | 2014/02/27 |
Robin_Watts | marcosw1: You about? | 00:18.57 |
MacWinner | Robin_Watts, thanks.. it's so nice.. saved my life.. and CPU | 00:22.48 |
| when i run this command: pdf2json ae5a8885d4809cd6089202062f14be9a9e0108e8.pdf -enc UTF-8 -compress ae5a8885d4809cd6089202062f14be9a9e0108e8.js | 01:00.11 |
| i get this error: Error (883705): Unknown filter 'Crypt': Error: Copying of text from this document is not allowed. | 01:00.34 |
| however, mudraw is able to render the PNGs fine.. was wondering if anyone has run into this before? I need this to extract metadata | 01:01.17 |
toothrot | macwinner, seems like a pdf permission thing | 01:45.53 |
stars | Hi all | 08:04.55 |
| I have a question about mupdf... how do we make "fz_print_text_sheet" of stext-output.c to write to separate style sheet? | 08:06.15 |
kens | stars you will have to wait a while for an answer, the MuPDF developers are not here right now | 08:06.40 |
stars | kens, Thanks! | 08:06.54 |
kens | But they will read the logs when they get in, either stick around, come back later, or read our logs when convenient for your answer | 08:07.17 |
stars | kens, how long it takes for those developer to be online today.... any idea... | 08:07.17 |
| sure kens | 08:07.27 |
| ! | 08:07.28 |
kens | stars, should be a couple of houirs, maybe less | 08:07.34 |
stars | kens, Great! Thanks! | 08:07.46 |
Robin_Watts | stars: Morning. | 10:14.17 |
| oh, he's gone. | 10:15.04 |
kens | I did say he oculd read the logs, but I was hoping he'd stick around | 10:15.41 |
Robin_Watts | Well, the answer is "you change the code".# | 10:15.58 |
kens | I'm afraid I didn't understand the question well enough, but I suspected you might need to ask more questions | 10:16.49 |
sebras | morning tor8! | 10:23.36 |
Robin_Watts | http://xkcd.com/now | 10:23.45 |
Mo | Hi, I'm a fan of the mupdf reader. Now I noticed there is zathura. What does it do better? It seems it is using mupdf-lib for rendering. | 10:35.57 |
Robin_Watts | AIUI, yes. | 10:36.34 |
| I can't comment on what zathura does better as I've never used it. | 10:41.07 |
sebras | Mo: maybe it is better to ask the zathura developer (who visits here sometimes). | 10:41.34 |
| Mo: the rest of us here are mainly concerned with the original mupdf reader. | 10:41.48 |
Mo | sebras: Ok, what is his name to look out for? | 10:42.02 |
| I see, then I stay with mupdf since it matches my VIM like commands and is fast and compact. I have lot of large PDF data to maintain and read. | 10:42.49 |
| This comment does match my opinion exactly: https://forums.gentoo.org/viewtopic-p-6983432.html?sid=8687fa00963c9caa72c085e196876b16#6983432 : " but I don't miss anything when using mupdf. Conceptually, it makes more sense to have a viewer that can also look at ps, djvu, docbook, etc., but the truth is that I don't currently look at any of those. " | 10:46.45 |
| Then, someone else says there: " Mupdf is a PDF (1.7) and XPS rendering lib written in C from the ground up (sort of... based off of libart). There are no plans to implement interactive features (form filling). It does have a minimal interface but it is actually intended as a library. " | 10:47.37 |
tor8 | good morning sebras | 10:49.26 |
sebras | Mo: I'm not sure what his nick is. might be ManDay, I forget. | 10:49.28 |
| Mo: some devs are working on form filling these days though. but not in the x11 viewer. | 10:50.16 |
kens | chrisl you htere ? | 11:16.46 |
Robin_Watts | stars: hi. | 11:25.27 |
| stars: short answer - you change the code. | 11:25.37 |
chrisl | kens: is it urgent, or can it wait another 20 minutes or so? | 11:29.56 |
kens | not urgent | 11:30.01 |
chrisl | Okay, I'll ping in a little while | 11:30.13 |
kens | NP | 11:30.17 |
chrisl | kens: back | 11:47.50 |
kens | chrisl ran into a funny with pprintld1 and friends | 11:48.01 |
| they don't (unsurprisingly) handle the %"PRId64" syntax | 11:48.18 |
| I was wondering if it was worth adding some new ones that do | 11:48.30 |
chrisl | I've never seen pprintld*..... | 11:48.39 |
kens | Hmm I htink that was the name, they were called from pdfwtite, maybe they are a pdfwrite-specific | 11:49.14 |
| Nope, they are used in spprint.c as well | 11:49.52 |
| THough apart form that it looks like only the vector devices | 11:50.13 |
chrisl | Well, they break down to gs_sprintf() so you could add 64 bit versions. Or just change the existing ones? | 11:51.13 |
kens | They seem to be defined in spprint.h | 11:51.13 |
| Well I was considering adding 64-bit ones, so as not to break teh existing ones :-) | 11:51.34 |
| But if I'm the only one using them..... | 11:51.41 |
chrisl | That's what I was thinking. A quick grep is probably in order | 11:52.24 |
kens | I did a grep, and it seems to be just me | 11:52.39 |
| And I just fixed my cases | 11:52.52 |
| So I guess I won't bother | 11:52.59 |
chrisl | I confess, I don't really see the point of them. | 11:53.15 |
kens | To be honest, me neither, they take a format and replace it with %ld | 11:53.30 |
| It looks like its just a convenience for writing to a stream | 11:54.02 |
chrisl | Yes, but a rather half-assed way to do it | 11:54.20 |
kens | I couldn't possibly comment :-D | 11:54.35 |
chrisl | Well, it strikes me as an invitation to stumble across the problem you just found! | 11:55.04 |
kens | I'm inclined to agree, me and the customer both in fact. | 11:55.31 |
| But its not a commonly used thing, so I think I'll just leave it | 11:55.47 |
henrys | mvrhel_laptop: here early | 13:38.39 |
mvrhel_laptop | hi henrys | 13:38.47 |
| it is late here . need to go to bed | 13:38.55 |
| had 3 meetings today. very tired | 13:39.21 |
| one more meeting tomorrow | 13:39.26 |
| then flying home | 13:39.30 |
Robin_Watts | mvrhel_laptop: Did you get to meet Ali? | 13:39.33 |
mvrhel_laptop | Robin_Watts: yes. he is an interesting guy | 13:39.47 |
| that was the 4th meeting today actually | 13:39.54 |
Robin_Watts | ah, fab. Never met him myself. But I'm glad that managed to be fitted in. | 13:40.05 |
| would have been a shame to have missed him given you were there. | 13:40.27 |
mvrhel_laptop | yes. He was relatively close to where we are | 13:40.43 |
Robin_Watts | Have a safe flight home if we don't speak before. | 13:40.47 |
mvrhel_laptop | thanks | 13:40.51 |
henrys | mvrhel_laptop: can't wait to hear all about the trip | 13:41.20 |
mvrhel_laptop | there will be a lot to talk about. This staff meeting is going to be an interesting one | 13:41.44 |
henrys | that's one possible adjective | 13:42.45 |
mvrhel_laptop | :) | 13:42.52 |
henrys | mvrhel_laptop: do you notice folks worried about the economy there? I keep reading disturbing news about the japanese economy: http://www.economist.com/blogs/banyan/2014/02/japans-economy | 14:01.19 |
mvrhel_laptop | henrys: I have been reading this stuff in the economist too, but I have not talked with anyone here about it | 14:04.27 |
henrys | mvrhel_laptop: and Japan having a trade deficit - that was really surprising to me. | 14:04.28 |
mvrhel_laptop | yes | 14:04.35 |
| off to print my boarding pass and then off to bed... | 14:04.45 |
henrys | good night safe travels | 14:04.57 |
Robin_Watts | henrys: I note that the economist won't let you read more than an article or two before it cuts you off. | 14:07.39 |
| If you then delete the economist cookie on your browser, you can read again... secure :) | 14:08.00 |
henrys | same for the new york times, denver post etc. | 14:08.33 |
Robin_Watts | I would have expected them to do something with IP addresses too. | 14:09.30 |
henrys | if you want it to be per person and an address is shared by a family ... | 14:11.23 |
Robin_Watts | I guess. | 14:11.55 |
kens | tkamppeter you may find this feature useful: | 14:54.01 |
| http://git.ghostscript.com/?p=ghostpdl.git;a=commit;h=b53c81e38925d05b00ba9068aebc5fcbc519a685 | 14:54.01 |
tkamppeter | kens, yes this is great, thanks. It will make the pdftops filter (at least with Ghostscript as renderer) much simpler, perhaps even eliminating the need of CUPS' pstops filter being called after GS. | 15:26.02 |
kens | tkamppeter ,sorry connectin dropped, I'd very much be interested in any feedback you have regarding the feature, particularly if its before the next release (which is, unfortunately, quite soon now) | 15:33.23 |
| Morning Ray | 15:49.05 |
kens2 | Goodnight folks | 17:31.44 |
Robin_Watts | tor8, sebras: Did either of you have a look at the commit yesterday? | 18:48.40 |
| I'm still fiddling with the filter stuff (I've got it down to 2000 diffs and no SEGVs now). | 18:49.07 |
marcosw | Robin_Watts: were you looking for me yesterday? | 18:57.38 |
Robin_Watts | I was, but I sorted my question. | 18:57.53 |
| A suggestion: | 18:57.57 |
| Can we put the full build lines used by the cluster in the logs? | 18:58.10 |
| I spent ages hunting for why my binary was working and the cluster one was failing, and it turned out to be that the cluster one has XCFLAGS="-DCLUSTER" | 18:58.58 |
| which I should have known, cos it's probably my fault it's there, but to find out what was going wrong required me digging through the cluster code to find the invocation, because all that's in the logs is: make V8_PRESENT=1 | 18:59.44 |
marcosw | sure, I think that will be quick and easy. | 19:00.19 |
Robin_Watts | Those are words a computer programmer should *never* speak :) | 19:07.48 |
chrisl | Along with "of course it will be done by then....." :-) | 19:09.35 |
Robin_Watts | I *think* it might just be patterns that I've broken. | 19:18.33 |
MacWinner | is there any relationship between mupdf and xpdf? | 19:34.43 |
henrys | MacWinner: no | 19:38.45 |
MacWinner | ahh.. thanks.. | 19:39.18 |
henrys | MacWinner: you reported a file significantly faster in mupdf vs. gs with png output. Can you post that as a bug at bugs.ghostscript.com? | 19:39.58 |
MacWinner | also, I have a PDF that contains form fields that were filled in with some text.. when I use gs to render the PDF into PNGs, the form fields seem to render fine.. but using mudraw, the form fields do not render, though the rest of the PDF does.. any ideas on how to get the filled in field data to render as well? | 19:40.46 |
| henrys, sure.. I thought it was a known thing that mudraw is much faster? | 19:40.58 |
henrys | not to a raster file output - screen we know about | 19:41.40 |
MacWinner | henrys, it was more than 1 file.. it was pretty much all files I tested with were much faster.. Though GS was about the same if I was rending very low quality images | 19:41.47 |
| henrys, k.. I'll post the commands I used and dig up some example PDFs | 19:42.13 |
henrys | MacWinner: well if you could just post a single obvious example | 19:42.38 |
MacWinner | will do | 19:43.10 |
Robin_Watts | We can add some delay loops to mupdf. | 19:43.24 |
MacWinner | haha | 19:45.05 |
| Robin_Watts, is lack of rendering of form field data a known issue with mupdf? | 19:45.35 |
| i'm using 1.3 built from source | 19:45.43 |
Robin_Watts | MacWinner: Depends on the exact nature of the form data. | 19:46.03 |
henrys | or maybe Robin_Watts should work on gs more ;-) | 19:46.07 |
Robin_Watts | 1.3 should have form support in. | 19:46.14 |
| MacWinner: Can you point me at a file? | 19:47.19 |
henrys | Robin_Watts: has there been a large improvement along the line because last comparision we were pretty close. | 19:47.23 |
| ? | 19:47.26 |
Robin_Watts | It depends on the file I think. | 19:47.39 |
MacWinner | Robin_Watts, can I pm you a link? I need to make sure the file doesn't have anything confidential in it | 19:47.54 |
Robin_Watts | By the time you've cranked gs up to do antialiasing, MuPDF is faster for screen output. | 19:48.04 |
henrys | of course if gs it might be gs is banding | 19:48.05 |
Robin_Watts | MacWinner: Sure. | 19:48.08 |
chrisl | Command line options will also make a big difference - I suspect that mupdf's antiliased output will be rather faster than gs's..... | 19:49.13 |
Robin_Watts | Down to 1134 diffs... | 19:49.36 |
MacWinner | as a lay person, mudraw just did what I needed without needing to futz around with command line params.. generating large images, then shrinking them down to get around anti-alias stuff | 19:50.17 |
| i was doing a bunch of steps with gs and imagemagick before | 19:50.53 |
chrisl | Hmm, why? Ghostscript does antialiasing. | 19:51.24 |
MacWinner | chrisl, I'm sure it does.. i couldn't get it to work after a lot of messing around and lots of googling.. Seems like a bunch of people on the forums have their own set of command line switches, but none rendered text properly for me | 19:52.16 |
| chrisl, do you have a pointer to maybe some example commands that worked for you | 19:52.34 |
chrisl | -dTextAlphaBits=2 (or 4) | 19:52.50 |
MacWinner | i think I tried that.. it was a couple months ago | 19:53.17 |
chrisl | Having said that, I *hate* antialiased text, and never use it myself | 19:53.21 |
| Well, TextAlphaBits enables antialiasing of text in GS - if fuzzy characters are your thing | 19:54.25 |
| MacWinner: what version of gs are you using? | 19:54.43 |
MacWinner | chrisl, default version with centos6.. lemme check | 19:55.00 |
| 8.70 | 19:55.16 |
Robin_Watts | Oh, so 5 years old then? | 19:55.39 |
MacWinner | lol.. i guess so.. didn't realize | 19:55.57 |
chrisl | Oh lord, the current version is 9.10, and we'll be doing a new release in a few weeks.... | 19:56.04 |
henrys | chrisl: I was think 9.12 | 19:56.54 |
| s/think/thinking | 19:57.05 |
chrisl | henrys: superstitious? | 19:57.26 |
MacWinner | so I'm confused.. what's the relationship between mupdf and gs? | 19:58.59 |
Robin_Watts | both are developed by the same company. | 19:59.10 |
henrys | chrisl:nope but lots of folks are | 19:59.11 |
MacWinner | or why would you use one vs the other (assuming I'm using the latest versions of both) | 19:59.34 |
chrisl | henrys: sure, OTOH, we might get a big increase in hit count if we went with the obvious increment ;-) | 19:59.54 |
chrisl | thinks: time for Robin's sports car vs truck analogy..... | 20:00.33 |
Robin_Watts | I'm with henrys. We fly too much to possibly give the TSA an excuse :) | 20:00.58 |
| chrisl: It's a good analogy! | 20:01.12 |
chrisl | MacWinner: one obvious thing is if you need Postscript input..... | 20:01.16 |
| Robin_Watts: I never said otherwise | 20:01.24 |
henrys | right they won't let you on the plane with that release on your laptop | 20:01.27 |
Robin_Watts | careful. Tor will be bumping the mupdf release number a lot... | 20:01.51 |
chrisl | No, I'm just joshin', I'm happy with 9.12 | 20:02.02 |
MacWinner | chrisl, ahh.. but if rendering PDFs to PNGs, should both pretty much perform the same and have the same fidelity? | 20:02.04 |
Robin_Watts | MacWinner: MuPDF is entirely written in C. | 20:02.41 |
chrisl | MacWinner: well, mostly, yes. There are things mupdf doesn't do that gs does, and things mupdf handles better than gs. | 20:02.53 |
Robin_Watts | The PDF handling in gs is written partly in C, partly in postscript. | 20:03.00 |
| gs has much better handling for print output, separations etc. | 20:03.30 |
| mupdf has a lower footprint, and handles antialiasing in a much more natural way. | 20:03.45 |
MacWinner | interesting.. i'm using a tool called flexpaper, and they recommend mupdf | 20:04.20 |
chrisl | Actually, now I think about it, antiliased text is probably a lot slower in gs, because we don't cache glyphs when doing antiliasing | 20:04.41 |
henrys | this comes up frequently enought that I think we should have a side by side comparison so folks can choose what suites them best. | 20:04.44 |
MacWinner | that's when I first discoverred it.. i'm surprised I didnt' find it earlier since I spent a lot of time searching for this type of tool a couple months ago | 20:04.50 |
chrisl | Anyway, it's a bit late here, and I'm starting to get a headache - goodnight, all! | 20:09.12 |
MacWinner | gnite | 20:09.44 |
Robin_Watts | MacWinner: I've got to go for dinner now. Can you check back in tomorrow? | 20:09.56 |
MacWinner | Robin_Watts, sure thing.. thanks so much for looking at it | 20:10.16 |
Robin_Watts | This may need to go to paulgardiner for a look, and he may not be about until Monday. | 20:10.29 |
MacWinner | No worries.. I'll leave #ghostscript as a perminanent fixture in my IRC channel list | 20:13.47 |
Robin_Watts | handy OâReilly discount code: EP499 - seems to make all eBooks $4.99. | 20:53.14 |
| Forward 1 day (to 2014/02/28)>>> | |