| <<<Back 1 day (to 2012/04/29) | 2012/04/30 |
Tester | I have a problem with missing library of ghostscript on mac os x 10.6. I try to use ghost4j and I can't found the library libgs.dylib. | 09:38.40 |
kens | Is it on your system ? | 09:39.03 |
Tester | no. | 09:39.13 |
kens | Well, three's your problem then | 09:39.19 |
Tester | but I don'zt know where to find the library. | 09:39.21 |
kens | have you asked the ghost4j people ? | 09:40.12 |
Tester | no. | 09:40.25 |
kens | well it seems to me that would be the first place to ask | 09:40.46 |
| chrisl you've got a Mac, do we make dynamic libraries for MacOS ? | 09:41.42 |
| Tester : where did you get Ghostscript from, and what version are you using ? | 09:44.15 |
Tester | I use version 9.05 from the official site. | 09:44.36 |
kens | Built form source ? | 09:44.48 |
Tester | yes | 09:45.44 |
kens | Probably you need to specify a make option to build dynamic librareis tehn | 09:46.03 |
Tester | --enable-dynamic? | 09:46.29 |
kens | Possibly :-) | 09:46.36 |
| I'm not the build engineer, adn I don't have a Mac, so its hard for me to tell | 09:46.52 |
Tester | I try it with this option. | 09:49.06 |
chrisl | kens: no, we don't make dynamic libs for mac. You just make them the same as Unix ones "make so" | 10:28.18 |
Yoshi47 | Robin_Watts, About you comment on my question, I don't mind waiting for the pdf to download, they are only single page or < 5 pages anyways and are on a local subnet so it should be fast anyways, so I just need it to open from http link on a local intranet server. | 10:59.54 |
kens | The MuPDF viewer doesn't open files from URL, tehy must be stored lcoally | 11:00.33 |
| SO you need to copy the file from the server | 11:00.42 |
Yoshi47 | kens, Robin_Watts was saying "We could put an http fetcher into the mupdf app, but as mupdf doesn't currently support opening PDFs that are still downloading, you'd still have to wait for it all to arrive first.20:51.27 | 11:01.19 |
| So there would be no benefit.20:51.37 | 11:01.20 |
| Actually, we should put linearised pdf support on the agenda." | 11:01.20 |
kens | I must have missed Robin's response | 11:01.46 |
Yoshi47 | kens, basically asking if they can add that support to it, he responded on friday afternoon, http://ghostscript.com/irclogs/2012/04/27.html | 11:02.10 |
kens | But as he says, there's no real benefit to adding it to the viewer | 11:02.30 |
| RIght, so as Robin says, the viewer doesn't support linearied PDF | 11:03.08 |
Yoshi47 | well not speed wise because of the download, but it makes it easier to access pdf on the network | 11:03.11 |
| which means viewing while downloading, right? | 11:03.31 |
kens | Yoshi47 : its a 'dumb' app, it can be extended in many ways. | 11:03.31 |
| Yoshi47 : in this case support for linearised meqans opening page 1 and viewing while download continues, yes. | 11:03.55 |
Yoshi47 | yep don't care to wait for page 2 to download to, 90% of these pdfs are 1 page | 11:04.22 |
| and the rest are < 5 pages long | 11:04.31 |
kens | So no real advantage | 11:04.36 |
Yoshi47 | so you wait for it to go into memory | 11:04.38 |
| not speed wise no. | 11:04.43 |
| but really nice if they are being handle off of a intranet application for multiple users on different OSs | 11:05.13 |
kens | Its not my area, but I doubt if it wil be added to the viewer. | 11:05.45 |
Yoshi47 | wonder if i can prepend my command with a downloader into tmp and then open with mupdf and then delete after viewing | 11:06.22 |
kens | You could certainly scrip it | 11:06.52 |
kens | lunches | 11:09.26 |
Robin_Watts | Yoshi47: Yes, either you can download to tmp and then invoke mupdf (which is what should happen on android, and we hope to make happen on ios)... | 11:40.47 |
| ... or you can add that to the app yourself. | 11:41.13 |
| IIRC you're on android ? | 11:41.20 |
Yoshi47 | so how are you doing that on android? | 11:41.27 |
| no on linux | 11:41.29 |
| but also need to work for windows | 11:41.35 |
| android would be nice too | 11:41.50 |
Robin_Watts | Yoshi47, ah ok. | 11:41.51 |
Yoshi47 | haven't got that far | 11:41.53 |
| just found mupdf last week | 11:42.03 |
| so im excited | 11:42.07 |
Robin_Watts | On android, it's just what the OS does - when you try to open a file that the browser doesn't understand, it downloads it. Then it looks for a file that matches the extension/mimetype and invokes the app on that (or should do) | 11:42.49 |
Yoshi47 | ah | 11:43.09 |
| Robin_Watts, thanks for that enlightening comment, I think its working now! browser is downloading first and then handing it off to mupdf, well xulrunner that is. | 11:45.39 |
Robin_Watts | cool. | 11:45.47 |
Yoshi47 | just not small enough | 11:46.13 |
| is there a limit on the minimum size? | 11:46.20 |
Robin_Watts | filesize? | 11:46.52 |
Yoshi47 | no screen size | 11:47.15 |
Robin_Watts | nope. | 11:47.24 |
Yoshi47 | doesn't seem to want to go smaller then like 640x480 | 11:47.37 |
| 54 dpi is the smallest it wants to go | 11:48.38 |
chrisl | Robin_Watts: do you know if I can have the cluster test tiffsep? | 12:08.04 |
Robin_Watts | You can't. | 12:08.14 |
| but it should test pamcmyk, if that's any help. | 12:08.25 |
| psdcmyk, sorry! | 12:08.38 |
chrisl | No, this is changes in the tiffsep device | 12:08.43 |
Robin_Watts | psdcmyk does spots etc. | 12:08.50 |
| The cluster can't test tiffs as writing tiff files involves seeking, and that screws the pooch. | 12:09.11 |
chrisl | Yeh, I thought marcosw had done "something" to let Michael test his changes with tiffsep | 12:09.36 |
Robin_Watts | michaels changes are being tested with psdcmyk. | 12:09.58 |
| tiffsep produces many files per page, and the cluster assumes just one. | 12:10.16 |
chrisl | Oh well, I'll script up something to test these changes, then | 12:11.15 |
Robin_Watts | You may be able to do something with htmldiff.pl in gs/toolbin. | 12:12.50 |
| That was my test script (that bmpcmp was written for) | 12:13.13 |
| It takes a list of files to test in the same format as comes back from a regression report. | 12:13.47 |
chrisl | There shouldn't be any differences in the output pages, only possibly more output pages in rare cases. | 12:13.52 |
Robin_Watts | chrisl: Ah. | 12:15.08 |
| sebras (for the logs): testing a fix for the locking thing now - thanks for spotting that. | 12:15.21 |
Yoshi47 | Robin_Watts, can you try make mupdf go to about 200x200 pixel window with a pdf you have? | 12:43.53 |
Robin_Watts | Look at pdfapp.h | 12:53.57 |
| #define MINRES 54 | 12:54.02 |
Yoshi47 | Robin_Watts, nutts, that means I have to compile it | 13:27.06 |
Robin_Watts | sorry. | 13:35.49 |
| ok. I'm heading off to heathrow in a mo. | 13:47.39 |
kens | see you tomorrow Robin_Watts | 13:48.04 |
Yoshi47 | Robin_Watts, done, compiled successfully, that was easy, good job! | 13:49.54 |
Robin_Watts | ok, that's it. Anything I've still forgotten will have to wait :) | 13:53.08 |
ray_laptop | Hi, all. I made it to the Holiday Inn (finally). | 14:27.30 |
| bbiab. | 14:30.15 |
kens | Hi ray, I hope you didn't have a long delay at immigrartion. | 14:30.30 |
ray_laptop | hi, kens. It was about 25 minutes in immigration. All told, (waiting for the bus and the ride, and getting checked in and on the internet) was about 2 hours | 14:42.21 |
| the flight was right on time | 14:43.18 |
| I'll check back here to see if anyone that's here is doing anything for dinner. | 14:45.55 |
| I assume that the locals will just come in tomorrow AM | 14:46.11 |
Yoshi47 | is there some kinda of conference going on that everyone is going to? | 14:49.23 |
chrisl | ray_laptop: Robin_Watts is at the hotel tonight, kens and I are coming up tomorrow morning | 14:50.06 |
ray_laptop | chrisl: I see. And henrys, marcosw and mvrhel are here, right ? | 14:50.50 |
chrisl | ray_laptop: I believe so. mvrhel was talking about some sightseeing in London, not sure about the others. | 14:51.36 |
ray_laptop | I think Miles is already here too (his flight was about the same time as mine, except from SFO. I don't know Scott's itinereary | 14:52.09 |
chrisl | I'd guess Scott will be sightseeing, too. He sent the warning about which Holiday Inn to find.... so I assume he's there | 14:52.57 |
ray_laptop | Are you guys driving in or coming by public transportation and taking the shuttle from the airport ? | 14:53.46 |
| The airport shuttle is "H4" | 14:54.14 |
chrisl | I'm driving up, and I'm going to collect kens from the coach at Terminal 5 (he gets the coach from Gatwick airport). Robin_Watts is driving | 14:55.02 |
kens | Off for now, see you all tomorrow | 16:34.47 |
mvrhel_laptop | oh looks like there is a major indeterminism in my patch | 16:53.50 |
| I guess I didn't have an easy way to test for that until after the commit | 16:54.04 |
| time to run valgrind.... | 16:55.13 |
| hmm something in the shading code from bmpcmp it would appear | 16:57.10 |
| actually 478-01.ps has a problem | 16:58.29 |
robin_watts_mac | At last. | 17:02.34 |
| henrys``: Are you at the hotel? | 17:02.50 |
mvrhel_laptop | hi robin_watts_mac | 17:02.50 |
robin_watts_mac | hi mvrhel_laptop | 17:03.03 |
mvrhel_laptop | scott and I just got back from the city about 30 minutes ago | 17:03.08 |
robin_watts_mac | ah. I've just been for a run. I have Henry and Sabrinas luggage in the car... | 17:03.39 |
| and internet in my room was messing about, so they had to give me some paswords. | 17:04.02 |
| What did you see in the city? | 17:04.28 |
mvrhel_laptop | The science museum and then I did a bunch of shopping for stephanie and the kids | 17:04.48 |
robin_watts_mac | nice. I haven't been to the science museum for at least a decade, but I used to love it as a kid. | 17:06.09 |
mvrhel_laptop | Jammie Dodgers and the like | 17:06.15 |
| hi ray_laptop | 17:06.24 |
| I was surprised to see that you came in today | 17:06.31 |
robin_watts_mac | hey hey ray. Where in the world are you this fine morning. | 17:06.34 |
mvrhel_laptop | I had a huge price diff with saturday vs sunday departures | 17:06.48 |
ray_laptop | hi, mvrhel_laptop. DO you know if we are going to dinner tonight ? | 17:06.49 |
mvrhel_laptop | I am | 17:06.54 |
robin_watts_mac | Me too. | 17:06.58 |
ray_laptop | mvrhel_laptop: how much was your fare ? | 17:07.07 |
robin_watts_mac | I saw Miles earlier. | 17:07.14 |
mvrhel_laptop | It was around 900 or so | 17:07.17 |
robin_watts_mac | and he's wanting food this evening. | 17:07.28 |
mvrhel_laptop | I think Scott is up in Mile's room now | 17:07.41 |
ray_laptop | robin_watts: yes, I saw you checking in, but was too far away to shout | 17:07.42 |
mvrhel_laptop | ray_laptop: how much was your ticket? | 17:08.07 |
ray_laptop | Miles came in today as well. My fare RT was 1134 | 17:08.20 |
mvrhel_laptop | oh that is not bad | 17:08.26 |
robin_watts_mac | ray_laptop: That was probably after my run. You were wise to stay out of smell range :) | 17:08.32 |
| Hey tor8. | 17:08.34 |
ray_laptop | there wasn't a difference for me with Sat. vs. Sun departure | 17:08.46 |
mvrhel_laptop | not with that price | 17:08.54 |
tor8 | hey robin, just arrived. dog slow internet if going with the "free" option. | 17:09.08 |
robin_watts_mac | wired or wifi ? | 17:09.21 |
mvrhel_laptop | wired with the full rate is not too bad | 17:09.29 |
ray_laptop | tor8: I went ahead with the 9.95 / day for 24-hrs | 17:09.38 |
ray_laptop | doesn't know how to type the funny "L" for pounds | 17:09.58 |
| I kept getting dropped off of the wired connection -- the cable they had in the room had the locking tab broken off, so it was loose. Once I switched to my own cable it was OK | 17:12.36 |
mvrhel_laptop | someone should call Miles and see what time he wants to eat | 17:13.26 |
robin_watts_mac | I nominate michael. | 17:14.02 |
mvrhel_laptop | ok done | 17:15.32 |
| we are going to the magpie | 17:15.39 |
| in a few minutes | 17:15.41 |
| miles and scott will be down in the lobby | 17:15.54 |
robin_watts_mac | heads for lobby then. | 17:15.58 |
| tor8: ping??? | 17:16.02 |
mvrhel_laptop | anyone talk to henry? | 17:16.03 |
| ray_laptop ping | 17:16.11 |
robin_watts_mac | Henry isn't here yet. | 17:16.11 |
mvrhel_laptop | oh ok | 17:16.15 |
robin_watts_mac | I hope tor8 can here the pings from the shower. | 17:16.23 |
ray_laptop | I'm here | 17:16.27 |
mvrhel_laptop | so lobby | 17:16.32 |
| in a couple minutes | 17:16.37 |
ray_laptop | I'll head to the lobby as well. Thanks | 17:16.40 |
robin_watts_mac | or at least henrys hadn't checked in 10mins ago. | 17:16.46 |
mvrhel_laptop | oh he is not even at the hotel yet | 17:17.00 |
| marcosw: are you still here or are you off to the show? | 17:17.11 |
| ok. see you all in the lobby | 17:17.54 |
tor8 | robin_watts_mac: pings received :) | 17:18.24 |
| robin_watts_mac: see you downstairs in a few | 17:18.32 |
sebras | Robin_Watts: I just read the logs and your fix... | 19:05.07 |
| I'm not sure that the fix is correct. | 19:05.21 |
| from what I saw it was nt the pdf_dict_gets() that failed, but rather later in pdf_to_int(). | 19:05.46 |
| the reason is that resolving references is delayed until they are used. | 19:06.05 |
| therefore it wouldn't help to release the lock while doing the pdf_dict_gets() as far as I understand. | 19:06.28 |
metan | robin_watts_mac: http://bugs.ghostscript.com/show_bug.cgi?id=693009 | 19:07.16 |
sebras | robin_watts_mac: oh... .so, now what? ;) | 19:08.00 |
robin_watts_mac | sebras: You may be right. | 19:42.39 |
| Do you have a test file that actually exercises this code ? | 19:42.53 |
sebras | I synthesized a few, yes. | 19:43.18 |
| I'll send them to you. | 19:43.27 |
| robin_watts_mac: sent. sorry, I though I explained the problem well enough. | 19:49.59 |
| tor8: hey! you arrived safely in .uk? | 19:50.20 |
tor8 | sebras: yup. | 19:50.34 |
| bumpy descent though, almost got a bit motion sickness :/ | 19:51.05 |
robin_watts_mac | sebras: Yes, you explained the problem nicely, but nothing beats actually testing it :) | 19:51.23 |
sebras | tor8: as bad as the descent at cph.dk last time? | 19:51.58 |
tor8 | different, this one was all wobbly sideways | 19:52.19 |
sebras | tor8: I'm still happy I slept last time when landing in cph | 19:56.35 |
| tor8 robin_watts_mac: there are a few patches over at sebras/master. have a look at them. | 19:57.34 |
robin_watts_mac | sebras: So, if I move the pdf_to_int into the unlocked section, we should be OK, right? | 19:58.26 |
sebras | robin_watts_mac: I think so. | 19:58.38 |
| robin_watts_mac: also we could make sure that the object is not indirect before continuing. | 19:59.23 |
| in the case of /Size and /Prev I believe that it is required by spec that those are not indirect (and if either is of another type we interpret it as 0 and something will break anyway). | 20:00.08 |
| in the case of /XRefStm I can not find that the spec requires it _not_ to be indirect... strangely enough. | 20:00.56 |
robin_watts_mac | ok, testing fix now. | 20:13.00 |
| new.pdf segvs. | 20:27.17 |
| new2.pdf works | 20:27.20 |
sebras | new2.pdf has problems with the locking for me. | 20:29.24 |
robin_watts_mac | OK, so I've fixed that one then. Thanks. | 20:29.41 |
| pushed. | 20:29.57 |
| and a fix pushed for new.pdf as well. | 20:32.57 |
sebras | robin_watts_mac: btw, did you intend to do a merge? | 20:34.38 |
| the tree looks a bit strange now. | 20:34.46 |
robin_watts_mac | bum. | 20:35.21 |
sebras | we got mupdfposter app and the 2 removed variables as well. | 20:35.25 |
robin_watts_mac | Those were intentional. | 20:35.42 |
sebras | I wasn't sure you started out on the correct local branch or something like that. | 20:36.26 |
| anyway it is less chaotic than other git trees I have seen with 100s of branches (none of them transient!). | 20:37.02 |
robin_watts_mac | Fix pushed. It was never wrong. It was always like this. These are not the droids you're looking for. | 20:37.45 |
sebras | these are not the droid's I'm looking for. | 20:38.39 |
| you can go about your business. | 20:38.45 |
| move along, move along. | 20:38.50 |
robin_watts_mac | Damn. I forgot to bring a USB lead to charge my phone. I don't suppose any of you guys have a spare micro USB lead I could use? | 21:10.24 |
| Forward 1 day (to 2012/05/01)>>> | |