| <<<Back 1 day (to 2013/07/08) | 2013/07/09 |
vtorri | chrisl: ping | 08:03.46 |
chrisl | vtorri: pong | 08:04.36 |
vtorri | chrisl: do you remember i had problems to compile ghostscript ? | 08:06.55 |
chrisl | Yes | 08:07.07 |
vtorri | (msys + mingw) | 08:07.08 |
chrisl | And I couldn't reproduce the issue | 08:07.20 |
vtorri | i think that i have found the problem | 08:07.23 |
chrisl | The mingw development headers? | 08:07.34 |
vtorri | before compiling ghostscript, i have compiled other libraries (for another project), including win-iconv, libjpeg, libpng and libtiff | 08:08.09 |
| they are all installed in /opt/win64/deps | 08:08.21 |
| as ghostscript configure asks for iconv, i have added -I/opt/win64/deps/include to gcc options | 08:09.09 |
| hence, it seems that the headers for libjpeg and others are first used from /opt/win64/deps/include and not the ones provided in ghostscript source | 08:10.11 |
| and as the versions are different, there are conflicts | 08:10.34 |
chrisl | Yes, there would be..... | 08:10.55 |
| Personally, I'd build without iconv, and avoid the problem | 08:12.02 |
vtorri | iconv is used for what in shotscript ? | 08:13.25 |
| -s+g | 08:13.29 |
chrisl | The two openprinting devices (oprp and opvp) | 08:14.14 |
vtorri | so i don't need it if i just want to use libgs for postscript rendering, right ? | 08:14.52 |
chrisl | Right | 08:15.05 |
vtorri | thanks | 08:20.15 |
chrisl | vtorri: I mentioned the mingw headers above: I found that if I had the headers for building mingw installed, Ghostscript wouldn't build - there are different, incompatible contents in them. I never got an answer as why that should be the case, nor what the recommended solution was (Ghostscript was not the only build that failed in that setup) | 08:22.25 |
Srini | Hi all | 08:22.53 |
| Is there someone from Mupdf? Need to understand one problem with mudraw | 08:23.16 |
vtorri | chrisl: i don't know, i never install the headers needed for mingw's compilation | 08:23.30 |
kens | Srini just ask your question, this channel is logged and the developers all read the logs | 08:23.44 |
chrisl | vtorri: fair enough, I just wanted to mention it, just in case. | 08:23.54 |
vtorri | ok | 08:24.00 |
Srini | kens, How do I submit a pdf so that I can explain the problem.... | 08:24.39 |
kens | Srini if you need to do that then you can open a bug report, or yes you will have to wait until someone can answer you directly, and email it to them. | 08:25.08 |
Srini | kens, Thanks! | 08:25.26 |
kens | Srini I'd suggest you give at least an outline, so the relevant developer knows you want them (and not someone else) | 08:25.27 |
| I see one here at the moment anyway (even though he's not saying anything) | 08:26.04 |
Srini | the problem is if the line spacing between the paragraphs and lines are the same and the paragraph opens with an indent... mudraw does not create paragraph tags in the html... | 08:26.24 |
kens | OK so that's text extraction, you need Robin_Watts_ for that, he'll be along within the hour I expect. | 08:26.57 |
Srini | ok! I will wait for him... | 08:27.15 |
kens | If you can hang around he'll ping you when he gets in (I'll poke him to read the logs) | 08:27.27 |
Srini | kens, Thanks! | 08:27.41 |
Robin_Watts_ | Srini: Hi | 08:48.16 |
| If the line spacing between paragraphs and lines are the same, that's one big clue about where the paragraphs are lost to us. | 08:48.53 |
kens | Robin_Watts_ : I think Srini is suggesting that the indent at start could be a clue. | 08:49.17 |
| Obviously its all heuristic.... | 08:49.26 |
Robin_Watts_ | I thought we tried to spot indents as a clue too, but it's all fuzzy maths. | 08:49.39 |
| Feel free to open a bug, and provide us a sample file, but I can't promise a fix any time soon. | 08:50.03 |
Robin_Watts | http://arstechnica.com/information-technology/2013/07/us-agency-baffled-by-modern-technology-destroys-mice-to-get-rid-of-viruses/ | 14:23.41 |
kens | And this was down to the Chief *Information* Officer ? boggling.... | 14:25.08 |
henrys | Robin_Watts, tor8 meeting time soon. Do we need to have one this week? | 14:57.56 |
Robin_Watts | henrys: I'm still bashing on progressive stuff. | 14:59.22 |
tor8 | henrys: I don't have anything new. we might want to decide when to merge in the pagetree stuff, but it still needs a bit more work before it's polished enough to go | 14:59.50 |
Robin_Watts | I've made progress, but (as mentioned on here the other day) I want to try to use the hints. | 15:00.00 |
henrys | I assume you've seen svgwrite customer requests. | 15:00.21 |
Robin_Watts | tor8: The pagetree stuff works now. | 15:00.26 |
| and if we don't push it, we'll never get reports about it being broken, so it'll stay in "it seems to work, but we're not confident enough to push it" limbo. | 15:01.00 |
| so I'd vote to push it onto master now. | 15:01.11 |
| henrys: I have. I was planning to ignore them for now, as the basic answer to all of them is "yes, it's an unfinished device. We'll get to it as time allows." | 15:01.50 |
henrys | he's also going on about signatures in ghostscript and we'll likely point him to mupdf. | 15:02.41 |
mvrhel_laptop | good morning | 15:02.54 |
Robin_Watts | Morning | 15:02.57 |
henrys | hi mvrhel_laptop | 15:03.06 |
Robin_Watts | mvrhel_laptop: Congratulations on getting the mupdf thing out to the store! | 15:03.15 |
tor8 | Robin_Watts: okay. the remaining bits that I want to fix are not critical for pushing it | 15:03.25 |
Robin_Watts | tor8: which bits are they? | 15:03.34 |
henrys | yes an app in the store is awesome! | 15:04.22 |
tor8 | Robin_Watts: creating a new pagetree from scratch if one doesn't exist using the pdf_insert_page functions | 15:05.09 |
henrys | mvrhel_laptop: have you been asking around at MS regarding the 8.1 pdf renderer? | 15:05.24 |
tor8 | and a function to rebalance / recreate it | 15:05.31 |
| the second part needs more care, given the fact that we can now inherit resources from higher up in the page tree hirearchy | 15:06.06 |
mvrhel_laptop | henrys: I am going to ping one contact via email today and I am hoping to see a VP from MS tonight at a baseball game | 15:06.10 |
Robin_Watts | tor8: For linearized files, we need to flatten resources (and there is code to do that) | 15:06.45 |
henrys | FWIW zeniko should have an NDA soon. | 15:06.51 |
| so we'll give him a casper account and he'll be able to regression test. | 15:07.44 |
Robin_Watts | so whenever we rebalance /recreate a tree we could flatten the resources too. | 15:07.50 |
| henrys: fab. He's wanted access to our regression files for ages. | 15:08.13 |
henrys | in the maintenance section of the agenda there is a link to bugs by engineer - tor8 leads the pack by a healthy margin ;-) | 15:11.41 |
| fuzzing may distort those numbers - I haven't looked at details. | 15:12.46 |
mvrhel_laptop | hmm I am in second | 15:13.08 |
Robin_Watts | henrys: I fear tor takes the pain for all the enhancement bugs for mupdf. | 15:13.16 |
henrys | yes I know | 15:13.48 |
| everyone knows mvrhel_laptop has the most real bugs ;-) | 15:14.31 |
Robin_Watts | I thought the only real bugs were customer bugs? :) | 15:15.20 |
henrys | they are no bugs only misunderstandings. | 15:15.47 |
| I didn't have anything else, do you see anything on the agenda worth discussing? | 15:19.08 |
Robin_Watts | Are we supposed to have booked flights to Chicago yet? | 15:19.46 |
henrys | I have a meeting with miles after the meetings, I'll ask. | 15:20.08 |
| have him send out an email. | 15:20.18 |
| I assume Raed is going to want digital sigs on something windows like, is that going to be horrific once it is working on android. | 15:21.08 |
| ? | 15:21.10 |
Robin_Watts | Either we'll need an openssl build for windows rt, or we'll need to do a stub to use WSA (is that right?) presumably. | 15:21.54 |
mvrhel_laptop | is this in ghostscript or mupdf? | 15:22.24 |
Robin_Watts | On the agenda, in the MuPDF Output devices section, we have "svg read and write". | 15:22.31 |
| mvrhel_laptop: mupdf. | 15:22.34 |
mvrhel_laptop | ok | 15:22.39 |
henrys | he asked for gs but we are going to push him toward mupdf. | 15:22.45 |
| like svgwrite | 15:22.54 |
Robin_Watts | MuPDF -> Output Devices -> read - yes, that should be mine. | 15:23.01 |
| sorry. | 15:23.05 |
| MuPDF -> Output Devices -> write - yes, that should be mine. | 15:23.12 |
| But svg read shouldn't be in the Output Devices section, and it shouldn't be assigned to me, AIUI. | 15:23.40 |
| at least, I'm assuming it's not assigned to me, and hence I'm ignoring it :) | 15:24.03 |
henrys | okay I'll move it and reassign to tor8 | 15:24.11 |
Robin_Watts | If it gets bounced back to me, that's fine, but at least then I'll know that my ignoring of it is bad. | 15:25.00 |
henrys | I have to think openssl will be okay on windows - some famous last words to end the meeting. | 15:25.03 |
Robin_Watts | henrys: I would imagine that OpenSSL will build on Win32 file. | 15:25.40 |
chrisl_r61 | henrys: the problem might be WinRT - a lot open source projects seem to be ignoring it for now, at least | 15:25.53 |
Robin_Watts | Whether it will build on WinRT (and WinPhone) is another question. | 15:25.53 |
| henrys: I'm not saying that it *is* a problem, just that it might be. | 15:26.34 |
| tor8: Shall I rebase the pagetree stuff onto the latest master and push it then ? | 15:27.25 |
henrys | maybe we should do a look at what is natively available first? | 15:27.32 |
mvrhel_laptop | silly question, but can we use any of this | 15:27.43 |
| http://msdn.microsoft.com/en-us/library/windows/apps/windows.security.cryptography.core.aspx | 15:27.45 |
henrys | we meaning mvrhel_laptop | 15:27.56 |
| I need to make and drink coffee before the next meeting. brb | 15:28.45 |
Robin_Watts | mvrhel_laptop: Yes, I suspect so. | 15:29.21 |
mvrhel_laptop | I am not up to speed on what we need openssl for with regard to signatures. I probably need to understand that first | 15:29.26 |
Robin_Watts | Though it's most likely the certificate stuff we need: http://msdn.microsoft.com/en-us/library/windows/apps/windows.security.cryptography.certificates.aspx | 15:29.35 |
mvrhel_laptop | ok right | 15:29.45 |
Robin_Watts | mvrhel_laptop: You might be well advised to wait til paul gets back from holiday and speak to him about what is required. | 15:29.57 |
| to avoid you going on a snark hunt. | 15:30.08 |
mvrhel_laptop | ok. yes. good idea | 15:30.15 |
| brb | 15:40.01 |
henrys | meeting time | 16:00.09 |
chrisl_r61 | henrys: I think there may have been some miscommunication when we got the "update" from URW back at the end of 2011...... | 16:00.59 |
henrys | chrisl:did you see the response from urw? | 16:01.00 |
chrisl_r61 | Yes ;-) | 16:01.06 |
| henrys: I'm thinking they took your wording of "original" to mean the first release from way back when, not the current, but unmolested URW release (guessing!) | 16:02.10 |
henrys | chrisl_r61: but the fonts are still visibly wrong aren't they? | 16:02.15 |
chrisl_r61 | Well, I felt that glyphs in an italic font ought to be italic, given that it appeared to be a regression, I didn't check those with Adobe | 16:03.24 |
| I checked the "M" width against Adobe, though | 16:03.47 |
henrys | chrisl_r61:no initially he was confused because he thought I wanted fixes to the AFPL fonts but then I told him GPL. | 16:03.54 |
| chrisl_r61: likely to get involved I'll look some more after the meeting. | 16:04.57 |
chrisl_r61 | henrys: back in 2011 (March 16), Peter wrote: "attached, please find the original 35 PST1 fonts that we supplied." and I don't think that's what we were exepecting | 16:05.07 |
henrys | mvrhel_laptop: is the v4 to v2 project still a go - we discussed it at the meeting? | 16:07.10 |
| I guess I don't have anything for the gs meeting other than what is on the agenda. | 16:08.01 |
| anybody else? | 16:08.09 |
mvrhel_laptop | henrys: I need to get to going on it, but yes it is still a go. | 16:08.26 |
| kens: is this holding you up right now in anyway? | 16:08.45 |
chrisl_r61 | henrys: leave the fonts stuff with me for a day or so.... | 16:08.58 |
mvrhel_laptop | when is the freeze going to occur for the release? | 16:09.00 |
henrys | the agenda items now have "owners" if anyone objects to that please let me know. | 16:09.09 |
mvrhel_laptop | v4 to v2 is not going to make it into this release though | 16:09.19 |
chrisl_r61 | mvrhel_laptop: the freeze is loosely "planned" for the first week or so in August - unless some sh*t hits the fan! | 16:10.03 |
henrys | yeah just a bunch of items in "discussed at meeting" and I want to figure out if they go to real projects or the "someday" bin. | 16:10.18 |
Robin_Watts | henrys: Are we expecting takane-san to appear on here at any point? | 16:11.30 |
henrys | chrisl_r61: how important is this ufst problem you added at the meeting and do you have a way to demo a specific problem? | 16:12.08 |
| Robin_Watts: I will talk to miles about that today. I'd like him to and I've already invited him. | 16:12.43 |
chrisl_r61 | henrys: erm, I don't remember that, hang on..... | 16:12.46 |
Robin_Watts | Are we still all setting up on skype? Should we share skype ids somehow? | 16:12.56 |
mvrhel_laptop | I am back on west coast now so I can keep an eye out for him in the evenings | 16:13.16 |
| I have Robin_Watts and ray_laptop on skype I know | 16:13.47 |
| and scott | 16:13.50 |
henrys | oh thank you - it is recommended (not required) that all staffers have Skype - audio at least. | 16:13.51 |
Robin_Watts | There is an entry for "Discussed adding document (word etc) interpreters" in the workflowy. At the meeting I was tasked with looking into the idea of licensing that in from my previous employers. I've talked to some people about this (including the ex-chief engineer and a current engineer there) and I've really got cold feet about that approach. | 16:14.34 |
| mvrhel_laptop: I have henry, tor and marcos too. | 16:15.32 |
chrisl_r61 | henrys: is this the UFST threading problem? Basically, the code in there is incomplete and slightly wrong. Assuming UFST itself is threadsafe, I need to spend a week or two fixing it up. | 16:15.38 |
ray_laptop | Robin_Watts: you mentioned some reservations at the meeting | 16:15.41 |
Robin_Watts | ray_laptop: Yeah, they crystallised more the more I spoke to people about it. | 16:16.13 |
henrys | chrisl_r61: I thought UFST threading was broken. | 16:16.19 |
mvrhel_laptop | sounds like it is best to avoid then | 16:16.22 |
ray_laptop | I'd be really surprised if UFST was thread safe | 16:16.54 |
chrisl_r61 | henrys: I *think* I found a trivial change to let it build in "thread safe" form, but I had no way to test it, and ran out of time to look at it | 16:17.07 |
henrys | did you tell them the alternative is we'll write our own and give it away ;-) | 16:17.12 |
Robin_Watts | LibreOffice split from OpenOffice over the issue of MS office document loaders, AIUI. | 16:17.20 |
| and 'new' contributions to LibreOffice are dual LGPL and MPL licensed. | 16:17.40 |
chrisl_r61 | Robin_Watts: that was just one aspect of it, but yes.... | 16:17.57 |
ray_laptop | well, at least those licenses are friendly | 16:18.03 |
Robin_Watts | so it's possible that the loaders are MPL licensed, and hence we may be able to use them as a basis. | 16:18.07 |
| ray_laptop: LGPL is not friendly for our purposes. | 16:18.37 |
ray_laptop | Robin_Watts: why not? "using the Lesser GPL permits use of the library in proprietary programs" | 16:19.50 |
chrisl_r61 | ray_laptop: only via dynamic linking | 16:20.09 |
henrys | does anyone know what google apps uses? | 16:20.41 |
chrisl | Whatever it is, I haven't been very impressed with it! | 16:21.33 |
ray_laptop | it's probably written in Java | 16:22.29 |
henrys | all things considered the item belongs in the someday bin. | 16:22.59 |
kens2 | Its a hard problem | 16:23.21 |
ray_laptop | for now we should keep looking for possible solutions | 16:23.43 |
| the part of the problem that concerns me is that it is an on going problem. A solution that's pretty good today will be obsolete a couple of years later and the library developers may have gone away | 16:24.49 |
Robin_Watts | ray_laptop: At least with the latest iterations, the file formats are shared, more or less. | 16:25.44 |
| rather than all being random binary blobs in secret code as they used to be. | 16:26.02 |
chrisl | And XML, so a bit easier to reverse engineer changes | 16:26.09 |
henrys | I think we'd need to hire to cover the project. | 16:26.28 |
ray_laptop | Robin_Watts: agreed. XML based is (a bit) nicer | 16:28.12 |
henrys | okay meeting done, I'll be around for a bit. | 16:28.25 |
kens2 | OK goodnight all | 16:28.54 |
ray_laptop | My family had a good time at Miles' place on the river! | 16:28.55 |
| bye, kens | 16:28.59 |
henrys | chrisl_r61: I don't follow about the fonts from URW's perspective they've given us the latest release of the GPL fonts. | 16:29.14 |
chrisl | henrys: Yes, but I don't think they gave us the latest release back in 2011 | 16:29.44 |
marcosw | bye everyone. I'm off to uni but will be back online later. | 16:29.50 |
ray_laptop | bye, marcos | 16:29.59 |
henrys | ray_laptop:oh nice, is there a lot for kids to do there? | 16:30.25 |
chrisl | henrys: I *think* the fonts they gave us back in 2011 were the ones from the original GPL release way back in 1999, or whenever it was.... | 16:30.54 |
henrys | well I think it's probably easier to look at the fonts see what is really broken wrt adobe and go from there. | 16:33.02 |
ray_laptop | Robin_Watts: My kids loved the river (canoeing and inner tubes and kayak). When we went to Yosemite last year, they were swimming for at least 3 hours every day in the Merced. The Russian River is slower, but still fun for them | 16:33.17 |
henrys | look at the latest fonts they sent. | 16:33.18 |
ray_laptop | Robin_Watts: rivers are much more fun than pools :-) | 16:33.53 |
Robin_Watts | ray_laptop: Nice. My memory from a few years ago is that it's a nice area. | 16:34.01 |
chrisl | henrys: I just checked the italic ones, and the current Adobe fonts appear to match the ones URW just sent - i.e a bunch of "symbols" are not italicised | 16:34.08 |
Robin_Watts | ray_laptop: Not where I'm going on holiday they aren't :) | 16:34.15 |
ray_laptop | Robin_Watts: yes. And they also had fun driving the golf cart all around. And picking and eating blackberries | 16:34.50 |
| Robin_Watts: where are you going (I've forgotten if you mentioned it) | 16:35.11 |
Robin_Watts | Namibia. | 16:35.21 |
henrys | Robin_Watts: where are you going now? - there ought to be discovery channel program that follows you and Helen around on vacation. | 16:35.35 |
ray_laptop | Robin_Watts: I see. yep, I wouldn't want to swim there | 16:35.48 |
Robin_Watts | Previously, on "Stupid White Tourists" | 16:36.08 |
mvrhel_laptop | :) | 16:38.05 |
chrisl | henrys: so the ones they sent back in March 2011 did *not* have those 2006 changes mentioned | 16:39.39 |
henrys | chrisl:right they sent us the originals | 16:47.04 |
| that is technically what we asked for one would think they'd give us the latest release but they didn't | 16:47.48 |
chrisl | Yes, so the only thing that *might* be a problem is the "M" glyph one, and I need to double check that | 16:48.25 |
mvrhel_laptop | henrys: So one of my contacts at MS who I mt. bike with is going to track down someone for me. He said it may be a few degrees of separation from him and who ever he finds. I also pinged another contact who is at the VP level. I have not heard back from him yet. | 18:04.52 |
| Robin_Watts: this is interesting | 18:09.38 |
| http://www.freelancer.com/projects/microsoft-windows-eight/Compile-MuPDF-WinRT-for-Metro.html | 18:09.40 |
| tor8: I think you can close Bug 693668 | 18:14.03 |
Robin_Watts | mvrhel_laptop: I think we should contact them. | 18:14.53 |
mvrhel_laptop | You mean the bug submitter Robin_Watts ? | 18:15.19 |
Robin_Watts | I mean the poster for that project. | 18:15.29 |
mvrhel_laptop | oh yes | 18:15.33 |
| I agree | 18:15.35 |
Robin_Watts | Shall I do that, or would you like to? | 18:15.43 |
mvrhel_laptop | Robin_Watts: you can go ahead if you would like | 18:15.50 |
Robin_Watts | ok. | 18:15.53 |
| oh, they've cancelled the project | 18:16.08 |
mvrhel_laptop | oh. I did not notice that | 18:16.23 |
| Robin_Watts: this one appears to be open | 18:17.48 |
| https://www.elance.com/j/compile-mupdf-on-winrt-metro-app/38328666/ | 18:17.50 |
| sounds like the same one | 18:18.15 |
Robin_Watts | does indeed. | 18:19.44 |
| mvrhel_laptop: Gawd, they don't make it easy to contact people on these sites... | 18:24.19 |
mvrhel_laptop | I only stumbled upon this while doing a google on winrt mupdf | 18:24.40 |
Robin_Watts | mvrhel_laptop: The second site's project has been assigned to someone. | 18:24.44 |
mvrhel_laptop | interesting | 18:24.55 |
| well, that should be an easy project for them | 18:25.06 |
| Forward 1 day (to 2013/07/10)>>> | |