| <<<Back 1 day (to 2012/11/18) | 2012/11/19 |
ray_laptop | well, if the changes to 17 file are tucked away somewhere, I sure can't find them. I'm just creating them from memory. I'm not even sure when they went away :-( | 04:31.45 |
saper | I google "git stash ate my changes" a bit | 04:35.49 |
| googled | 04:35.55 |
ray_laptop | saper: me too. Not much help. stackoverflow turned up posts about untracked files, but these were ALL tracked files | 04:42.44 |
| saper: thanks for your suggestions. I guess the 'fatal' errror really was fatal. Maybe that's what erased my changes :-( | 04:43.49 |
saper | no, I think it did not | 04:44.00 |
| I am not sure how stash works but maybe your blobs and trees got created | 04:44.32 |
ray_laptop | saper: the two "recent" objects don't seem to have my changes when I use git checkout _____ to load them | 04:50.21 |
| I give up on getting my changes back. I could spend more time than just doing them over (I think) | 04:51.56 |
kens | Robin_Watts : tor8, a question on stack overflow (actually 2 questions...). Can one of you answer ? | 08:14.02 |
| http://stackoverflow.com/questions/13447962/internal-link-support-in-mupdf | 08:14.03 |
Robin_Watts | kens: Answered. | 09:52.00 |
kens | THanks Robin_Watts | 09:54.57 |
Robin_Watts | np. Thanks for the pointer. | 09:56.04 |
kens | I'm kind of relieved you didn't understand his second question either ;-) | 09:56.24 |
unixSnob | why does imagemagick disagree with pdfinfo on geometry? | 10:43.29 |
| i want to use /CropBox, but i don't want to guess on the coordinates | 10:44.34 |
kens | Maybe you should ask the ImageMagick folks | 10:54.49 |
| Or post an example file | 10:54.54 |
| I don't know what infor ImageMagick gives, but as far asw I'm aware, pdfinfo.ps uses the Box entries from the PDF file | 10:56.10 |
unixSnob | ultimately i intend to crop using /Cropbox, but how do I come up with the coorinates? | 11:12.17 |
| i know imagemagick users can use "display" to get coordinates visually.. but these coordinates don't match up | 11:12.51 |
| what do ghostscript users use? | 11:12.58 |
Robin_Watts | unixSnob: I think kens has answered your question. With gs, if you call pdfinfo.ps you should get exactly the values present in the file. If you know of a file where this isn't the case then please let us know. | 11:14.54 |
| If you want to know what ImageMagick is doing... then ask the people that deal with ImageMagick. | 11:15.16 |
unixSnob | what's pdfinfo.ps? | 11:19.00 |
| i have pdfinfo.. is that different? | 11:19.13 |
kens | where did you get pdfinfo ? | 11:19.54 |
unixSnob | kens: it may be part of the debian installation.. not sure | 11:20.24 |
kens | Well if its not part of GS, we can't help you | 11:20.35 |
Robin_Watts | unixSnob: There used to be a pdfinfo file as part of mupdf. | 11:20.39 |
| but poppler also provides a pdfinfo too. | 11:20.47 |
unixSnob | kens: i installed ghostscript, but i did not get a pdfinfo.ps command as part of the package, and apt-cache search does not find a pdfinfo.ps | 11:22.07 |
kens | pdfinfo.ps is a PostScript program, not a command | 11:22.24 |
unixSnob | i also did a "apt-file search pdfinfo.ps" and "locate pdfinfo.ps". apparently it does not come packaged with the debian ghostscript package | 11:23.52 |
| i have ghostscript 8.71~dfsg2-9.. maybe that's too old? | 11:24.44 |
Robin_Watts | That is pretty old, yes. | 11:24.55 |
unixSnob | debian is always quite behind.. a price we pay for stability | 11:25.22 |
| i'm probably stuck with imagemagick display.. but as soon as I ask the imagemagick folks how to using display coordinates for ghostscript, they're going to send me here | 11:27.26 |
Robin_Watts | unixSnob: No, you ask the imagemagick folks how they get the numbers they report from the numbers in the pdf file. | 11:30.42 |
chrisl | unixSnob: and if they tell you they come from Ghostscript, you need to ask them for the *exact* command line parameters they pass to Ghostscript to extrac the numbers, before we can possibly help. | 11:33.38 |
| s/extrac/extract | 11:35.32 |
Robin_Watts | tor8: Some patches for review on my repo. | 12:28.06 |
unixSnob | i need more options. there is no chance image magick can do this properly | 12:43.51 |
| the document is sideways to ghostscript, but upright to imagemagick | 12:44.12 |
| so the different tools don't even orient it the same | 12:44.23 |
| is pdfinfo.ps the only ghostscript tool that can display a PDF? | 12:44.43 |
| ghostscript faq suggests ggv, but Debian does not have that either | 12:46.52 |
Robin_Watts | pdfinfo.ps is not a ghostscript tool for displaying a PDF. | 12:49.21 |
| pdfinfo.ps is a postscript program that works in conjunction with ghostscripts PDF interpreter to extract information about a PDF. | 12:49.45 |
| ghostscript itself can display a PDF (using the display device), but it's not really a fully fledged viewer - you can't move backwards pages or pan and zoom on them. | 12:50.32 |
| If you want a PDF viewer, why not consider mupdf? | 12:50.54 |
| There are viewers built on top of gs itself, such as ghostview. | 12:51.30 |
unixSnob | Robin_Watts: well it has to be ghostscript-based, because i need the viewer to give me coordinates for selected regions that I can give to ghostscript | 12:51.41 |
Robin_Watts | No. | 12:51.48 |
unixSnob | i'm finding that different viewers are using different coordinate systems | 12:52.00 |
Robin_Watts | You just need the viewer to give you coordinates that can be mapped back into PDF document space. | 12:52.22 |
| Because that's what gs works in. | 12:52.33 |
| Most viewers take things like "/Rotate" into account. | 12:52.51 |
| and if you're feeding in values to ghostscript, then you need to take that into account too. | 12:53.18 |
unixSnob | the /CropBox command is flipping my image sideways | 13:05.11 |
| is it possible that my translation of coordinates is doing that? | 13:05.37 |
Robin_Watts | What do you mean by "flipping my image sideways" ? | 13:27.11 |
| Reflecting it in the Y axis? No, that should not be possible. | 13:28.01 |
| But without knowing exactly what you're doing, I can't really comment with any accuracy. | 13:29.21 |
| lunchtime | 14:03.15 |
tor8 | Robin_Watts: s/factor/power/ in downscaling? | 14:32.10 |
| or factor_pow2 | 14:32.28 |
| or something less confusing :) | 14:32.33 |
| Robin_Watts: "Add image subsampling function." you've got an #ifdef 0 section that doesn't belong and a missing newline at the end of the file | 14:35.01 |
| other than that, LGTM | 14:35.10 |
Robin_Watts | I'll reply to the mupdf support license thing. | 14:39.47 |
paulgardiner | Robin_Watts, tor8: couple of commits on paulg/master. Nasty Android threading stuff. | 14:42.44 |
Robin_Watts | Morning mvrhel_laptop. Happy belated birthday. | 15:04.12 |
mvrhel_laptop | Robin_Watts: Good morning and thanks! | 15:04.28 |
Robin_Watts | paulgardiner: I will look now, unless tor8 already is. | 15:14.24 |
paulgardiner | Thanks | 15:14.38 |
Robin_Watts | paulgardiner: Could you maybe add some comments to the code describing what the different locks/condition variables are about? | 15:18.28 |
paulgardiner | Yeah that would be sensible | 15:18.46 |
Robin_Watts | I understand that you've got a thread in the java that does nothing but start up, wait for alerts etc? Some description of that would be nice. | 15:19.03 |
| Some description of what alert_reply/request/active etc are would be nice too. | 15:19.43 |
henrys` | nice to have a short week, US folks are off Th and Fr | 15:23.23 |
kens | ah, Thanksgiving | 15:23.38 |
mvrhel_laptop | henrys: I am going to take Wed. off too | 15:59.37 |
paulgardiner | Robin_Watts: comments added as requested. | 16:05.02 |
Robin_Watts | paulgardiner: Thanks. | 16:05.09 |
henrys | mvrhel_laptop:okay - I'll be out monday and tuesday going to flint ... buffing up the kevlar | 16:07.56 |
Robin_Watts | henrys: today/tomorrow or next week ? | 16:08.43 |
kens | thinks next week, after the Thanksgiving weekend | 16:09.09 |
Robin_Watts | That would make more sense, yes. | 16:09.19 |
henrys | next week | 16:09.47 |
| I'll send out email | 16:10.03 |
Robin_Watts | paulgardiner: You have alerts_init after fz_open_document | 16:22.29 |
paulgardiner | Hmmm | 16:22.52 |
Robin_Watts | Is it possible that if fz_open_document fails, alerts_fin() will be called without alerts_init having be called. | 16:22.53 |
| Will that matter? | 16:22.58 |
| I also wonder if we should have something like chrome does with alerts. | 16:24.13 |
| With CHrome, if you get a second alert pop up quickly, it offers an "ignore all further alerts from this webpage" option, thus ensuring you can't get into an infinite loop of alerts popping up. | 16:25.01 |
| Maybe we should have a 'close document' option on the alerts or something, so people don't get mupdf jammed on their phone ? | 16:25.38 |
| What's the purpose of the switch(result.iconType) ? | 16:27.01 |
| I'm guessing there is code in there missing ? | 16:27.08 |
paulgardiner | I intend to choose an icon appropriate to the type of alert there, but was going to add that later. | 16:27.39 |
Robin_Watts | ok. | 16:27.48 |
| A 'FIXME' would be nice then (but don't bother unless you're updating other things. | 16:28.09 |
| If you're updating other things, then there are some typos in comments. | 16:28.38 |
paulgardiner | I think a get-out button may be worth adding, but I'd leave it for now. It would happen only if there was a loop in the javascript, and you can always get out by holding menu on android. | 16:29.00 |
Robin_Watts | Does it make sense to have an alertWaiter with no core? | 16:29.48 |
| cos alertWaiter assumes core is non NULL. | 16:29.58 |
| and if core can become non NULL after onStart is run, then nothing will have run core.startAlerts | 16:30.34 |
paulgardiner | onCreate errors out if it fails to init core. | 16:32.24 |
Robin_Watts | I'm looking at onStart | 16:32.47 |
| which checks core before calling core.startAlerts | 16:32.57 |
| but calls createAlertWaiter regardless. | 16:33.04 |
paulgardiner | Yeah, I think the test isn't needed. Maybe I should remove it. | 16:34.03 |
Robin_Watts | only if you're changing other things. | 16:34.31 |
| Otherwise it seems fine. | 16:35.02 |
paulgardiner | I'd better think about fz_open_document failing. | 16:35.54 |
| It's ok, although a bit lucky. alert_fin is safe to call without a call to alert_init provided doc is NULL, which it will be. | 16:38.54 |
Robin_Watts | paulgardiner: OK. Will push that in a mo. | 16:41.35 |
paulgardiner | ta | 16:41.47 |
unixSnob | where can I upload a PDF to? to post examples for this channel | 16:43.43 |
kens | ANy public location and gie us a URL | 16:43.57 |
| If its private, open a bug and attach it and tell one of us the bug number | 16:44.24 |
unixSnob | most of the pastebin type services are text only | 16:44.44 |
| I found one for images, but PDF is not accepted | 16:44.54 |
kens | transferbbigfiles.com | 16:45.05 |
Robin_Watts | paulgardiner: pushed. | 16:53.11 |
paulgardiner | Robin_Watts: Here's the apk http://intranet.glidos.net/~paul/MuPDF20121119.apk | 18:09.34 |
Robin_Watts | And that works for you? | 18:09.49 |
paulgardiner | Yep :-( | 18:17.33 |
mvrhel_laptop | tkamppeter: I closed Bug 693438 as the command line used is not really correctd | 18:21.19 |
tkamppeter | mvrhel_laptop, OK. Richard will get notified then. | 18:27.40 |
kens2 | OK that's enough for the night.... | 19:50.50 |
| I hate TrueType comosite glyphs. | 19:51.04 |
mvrhel_laptop | Robin_Watts: is your birthday tomorrow? | 21:37.30 |
| bbiaw. need to run a few errands | 21:38.08 |
| Forward 1 day (to 2012/11/20)>>> | |