| | <<<Back 1 day (to 2012/08/09) | 2012/08/10 |
| alexcher | marcosw: How can I see bmpcmp results for 32 vs 64 bit builds? | 00:01.43 |
| Robin_Watts | alexcher: You can't using the cluster | 00:13.48 |
| | but in gs/toolbin/ there is a file htmldiff.pl that you can use to run 2 different versions of gs and give you something very bmpcmp like. | 00:14.18 |
| | (It's the script that bmpcmp was developed from) | 00:14.37 |
| alexcher | OK, I'll try. | 00:15.42 |
| Robin_Watts | LCMS 2.4 is out. Clearly my piddling contribution was so great he's listed me twice in the AUTHORS file :) | 00:19.39 |
| marcosw | alexcher: at the moment I don't think you can. I'll modify the cluster code so that bmpcmp can take the 32 bit build option. | 01:55.36 |
| | henrys: I just emailed out the distiller vs gs pdf file size conversion. I haven't looked at the generated PDF files yet, but I've uploaded a .tar file with all the files to casper. | 01:56.32 |
| mvrhel_laptop | henrys: I just sent you and email about the dUsePDFX3Profile | 02:27.03 |
| | hopefully I am not too late | 02:28.33 |
| furrywolf | any chance some version of ghostscript I don't have can deal with the forms pdf files produced by adobe livecycle? | 04:23.12 |
| | I don't need the actual interactive forms bit, just to print it and fill it out by hand... but gv can't render it at all. | 04:24.42 |
| | file in question can be found at http://www.chp.ca.gov/publications/pdf/chp362.pdf | 04:27.52 |
| coffeesam | hi | 04:32.13 |
| ghostbot | hi, coffeesam | 04:32.13 |
| coffeesam | anyone can explain when will pdftops will produce a postscript file that is almost entirely black? | 04:32.40 |
| | I opened the pdf file in OS X preview and it looks fine to me | 04:32.51 |
| | I can use gs to produce a png that is ok too | 04:33.07 |
| | gs sDEVICE=pswrite produces too big a Postscript file | 04:33.37 |
| | so I thought of trying out pdftops | 04:34.03 |
| | anyone? | 04:34.06 |
| alexcher | coffeesam: If pdftops produces a file that is different from other devices, it's a bug. | 04:49.05 |
| coffeesam | hmmm | 04:49.17 |
| | ok | 04:49.20 |
| alexcher | Please file a bug report. Don't forget to attach your file. | 04:49.29 |
| coffeesam | it gives me a pixelated black file | 04:49.35 |
| | :( | 04:49.36 |
| | ok | 04:49.37 |
| | for now i am using gs | 04:49.43 |
| | where should i report it to? | 04:49.51 |
| | i downloaded it from xpdftools | 04:50.00 |
| alexcher | bugs.ghostscript.com | 04:50.04 |
| coffeesam | ok cool | 04:50.08 |
| | thanks | 04:50.09 |
| | :) | 04:50.10 |
| furrywolf | is "gv can't render evil files" bug-report worthy? heh | 05:00.58 |
| | all I need to do is print out a couple pages, but it can't even pretend to render it. :( | 05:01.57 |
| alexcher | furrywolf: The latest gs cannot render this file. Please file a bug report. | 05:11.01 |
| | furrywolf: The The file needs working JavaScript. Ghostscript has no JavaScript and won't have it in the near future. This is a known problem. | 05:16.20 |
| furrywolf | That's what I figured... are there any free-software viewers that do? | 05:17.32 |
| | _nothing_ seems to render those files, at least not that I can find. even google's search just shows the "get acroread" blurb. | 05:17.57 |
| | the files seem intentionally designed to be as non-compatible as possible. | 05:20.21 |
| alexcher | furrywolf: I don't know. You can run AR on Linux. | 05:23.00 |
| furrywolf | I'll probably have to head to the local chp office tomorrow and get someone there to print one for me... I really hate dealing with those people. heh. | 05:23.16 |
| | acroread is huge bloated crap with spyware, and apparantly it doesn't even render those files on linux without finding some plugin for it, from what I found googling. | 05:23.54 |
| | oh well, I'll drive there tomorrow... at dialup speeds, the 30 mile round trip is still faster than downloading acroread. | 05:30.18 |
| | thanks anyway! | 05:30.22 |
| ray_laptop | hi, kens | 07:18.37 |
| | looks like you are up and working (I'm about to stop for the day) | 07:19.04 |
| kens | Hi ray_laptop yes, up and started work ;-) | 07:19.17 |
| | trynig out the code fro Robin last night | 07:19.27 |
| | Looks like I wasted a day for both of us chasing a stupid memory error | 07:19.42 |
| | Lol I like the mail from Guillaume's vustomer. He suggests using an XML file for a bunch of settings. So we'd need to add an XML interpreter to the PCL intperpreter just to read its configuration....... | 07:21.39 |
| | Maybe we should have a PostScript file to control the XML interpreter settings... | 07:22.12 |
| chrisl | Well, to be fair, including an XML interpreter would be much less invasive than including the PS interpreter | 07:22.42 |
| ray_laptop | I didn't read it, but adding command line options that don't require the PS interp seems like a better way to control things -- that or PJL (it's how HP does settings) | 07:58.56 |
| kens | The PS interpreter was a joke folks.... | 07:59.09 |
| ray_laptop | kens: sorry, I missed the winky | 07:59.33 |
| kens | I was just amused that they wanted to add an XML configuraion file, which would mean adding an XML parser... | 08:00.09 |
| chrisl | Actually, PJL sounds like a winner - IIRC, it has scope for vender specific settings | 08:01.01 |
| ray_laptop | I guess I've been "away" and haven't noticed for a while now. :-O | 08:45.00 |
| | I think I won't bother with it anymore -- it's too much trouble | 08:45.36 |
| kens | :-) | 08:45.47 |
| ray_laptop | if anyone from Artifex wants me they can just SMS me (like Robin did this AM) | 08:46.00 |
| | and henrys does to remind me of the Tue meeting :-) | 08:46.15 |
| | time for bed now... | 08:48.43 |
| kens | goodnight ray_laptop | 08:48.51 |
| ray_laptop | back in a while (hopefully 7+ hrs) | 08:49.05 |
| | g'nite, kens | 08:49.21 |
| | at least this late night work is good for reviewing some OLD bugs and closing them. One is _still_ valid (which is why I'll let it percolate while I sleep). The clist WRITER fillpage routine is being called during the clist playback !!! | 08:52.05 |
| kens | THat sounds terminal | 08:52.23 |
| ray_laptop | somehow the 'plane_extract' target device is honked up. :-( | 08:52.45 |
| | luckily it's just the bmpsep1 device which nobody really cares about, but I suspect we will want to revamp the plane extraction to REQUIRE the planar mode mem devices | 08:54.11 |
| | enough for now... | 08:54.41 |
| kens | Indeed... | 08:54.53 |
| Robin_Watts | kens, chrisl, tor8: Did you see my mumblings about the ESTA stuff last night? | 09:22.27 |
| chrisl | Robin_Watts: yes, but mine's up to date - thanks | 09:22.48 |
| kens | Robin_Watts: , sorry was off for coffee | 09:27.43 |
| | yes, I saw, but I had to pay for my ESTA last time, I still have some left on it | 09:27.59 |
| | BTW the memory stuff works fine now. I think I know what was going on.... | 09:28.19 |
| | The actual bug you found with Memento (Offsets table) was the *real* problem, but initially it was manifesting as a problem with the ResourceUsage, probably because it wrote over the memory. | 09:29.09 |
| | I changed the gc pointer tracking, mistakenly thinking I was taking it out of GC, which provoked a new and different problem, which merely happened to look pretty much the same as the previous one :-( | 09:30.00 |
| | But anyway, its all fixed now, thanks for the help! | 09:30.28 |
| | I've just discovered that my neat trick for finding the last reource used by page 1 doesn't work, because of the stupid way we write out fonts :-( | 09:31.10 |
| Robin_Watts | kens: Yay and Boo, then. | 09:32.22 |
| kens | The yay is much bigger than the boo.... | 09:36.45 |
| | It just means I have to do some more work on renumbering the objects which I thought I could avoid. | 09:37.03 |
| Fandekasp | hi there. Isn't it strange that man mupdf doesn't tell us about Shift+P to show the number of pages ? Is there a lot of other features that I can only find on google ? | 09:45.02 |
| Robin_Watts | Pssh. So it's not "Metro", it's "Windows 8". | 10:15.29 |
| kens | and was always going to be :-) | 10:15.45 |
| | And nothing at all to do with anyone having trademarks on the name.... | 10:16.58 |
| Robin_Watts | My sisters shop used to be called "metro" | 10:17.12 |
| kens | I'm sure that's fine as long as she wasn't trading in Germany :-) | 10:17.29 |
| Robin_Watts | src_size = penum->rect.w - 1; | 11:20.35 |
| | remove the -1 ? | 11:20.40 |
| | I'm getting a pixel of crap on the right hand side with my test file, but it's closer. | 11:21.20 |
| | mvrhel_laptop: ^ | 11:21.26 |
| | have a good flight, I'll keep bashing at it. | 11:21.34 |
| mvrhel_laptop | Robin_Watts: ok | 11:22.05 |
| | cool | 11:22.06 |
| | Robin_Watts: I was thinking that we needed the -1 on both the src and des lengths | 11:22.50 |
| | did get a proper scale amount | 11:22.56 |
| | the issue seemed to me to be the calculation of the destination length | 11:23.10 |
| Robin_Watts | The scale_factor isn't really used. | 11:23.13 |
| mvrhel_laptop | oh | 11:23.17 |
| Robin_Watts | We spot the identity and half cases, but that's all. | 11:23.31 |
| | Ok, I have something that looks good with my test case. | 11:23.51 |
| mvrhel_laptop | oh | 11:23.58 |
| | you are correct sir | 11:24.01 |
| Robin_Watts | and for yours. | 11:24.50 |
| | change: | 11:25.17 |
| | @@ -813,7 +813,7 @@ image_render_mono_ht(gx_image_enum * penum_orig, const byte * buffer, int data_x | 11:25.37 |
| | flush_buff = true; | 11:25.39 |
| | } | 11:25.41 |
| | } | 11:25.43 |
| | - src_size = penum->rect.w - 1; | 11:25.44 |
| | + src_size = penum->rect.w; | 11:25.46 |
| | switch (posture) { | 11:25.48 |
| | case image_portrait: | 11:25.50 |
| | and: | 11:25.52 |
| | @@ -906,7 +906,7 @@ image_render_mono_ht(gx_image_enum * penum_orig, const byte * buffer, int data_x | 11:25.55 |
| | } | 11:25.57 |
| | if (flush_buff) goto flush; /* All done */ | 11:25.59 |
| | /* Set up the dda. We could move this out but the cost is pretty small */ | 11:26.01 |
| | - dda_init(dda_ht, 0, src_size, data_length-1); | 11:26.02 |
| | + dda_init(dda_ht, 0, src_size, data_length); | 11:26.03 |
| | devc_contone_gray = devc_contone[0]; | 11:26.05 |
| | if (penum->color_cache == NULL) { | 11:26.07 |
| | /* No look-up in the cache to fill the source buffer. Still need to | 11:26.08 |
| | but my fix is just by experimentation - it could easily be wrong. | 11:27.17 |
| mvrhel_laptop | ok. let me give that a try | 11:27.48 |
| Robin_Watts | I have a more cutdown turkey.pdf that makes it more obvious. Let me upload that for you. | 11:28.26 |
| mvrhel_laptop | oh cool ok | 11:28.33 |
| Robin_Watts | I shrank the page and made the turkey exactly fit it. | 11:28.54 |
| | http://ghostscript.com/~robin/turkey.pdf | 11:29.19 |
| | The right hand column of turkey pixels are all empty. | 11:29.30 |
| | and the png rendering correctly shows an empty column. | 11:29.41 |
| | The fast dithered rendering doesn't render those pixels at all. | 11:30.00 |
| mvrhel_laptop | ok that is helpful | 11:31.34 |
| | and yes, it is the proper length now. | 11:37.20 |
| | let me beat on some other images with this and I think I will do a cluster push too | 11:37.38 |
| | then I have to head off | 11:37.58 |
| Robin_Watts | I'm cluster pushing it at the moment. | 11:38.01 |
| mvrhel_laptop | oh | 11:38.05 |
| | ok | 11:38.07 |
| | thanks | 11:38.10 |
| Robin_Watts | In fact the clusterpush has run. | 11:38.44 |
| | just the bmpcmp to go. | 11:38.46 |
| | OK. Must fix that bug (specifying filter= with bmpcmp confuses it) | 11:40.09 |
| mvrhel_laptop | I was wondering about that | 11:41.27 |
| Robin_Watts | marcosw: Apparently my camera is within 20 miles of you :) | 11:57.08 |
| marcosw | Robin_Watts: If the camera is within 20 miles of my house it's closer to you than it is to me, I'm currently 5500 miles rom my house. | 12:00.05 |
| Robin_Watts | ha! | 12:00.15 |
| | lunchtime. | 12:06.47 |
| mvrhel_laptop | bbiaw | 12:11.55 |
| marcosw | kens: is it worth it trying to bisect to find the change re. <http://bugs.ghostscript.com/show_bug.cgi?id=693266> or is it unlikely to be a change that's patchable back to 8.71? | 13:35.25 |
| Robin_Watts | marcosw: One of these days, you are going to have to send a patch to someone from 8.71 to HEAD :) | 13:36.56 |
| kens | marcosw there's only one likely fix listed in details.htm | 13:37.37 |
| | I htink its all PostScript (in the PDF interpreter) so its likely back-portable | 13:38.24 |
| chrisl | It might need two fixes | 13:38.44 |
| kens | Perhaps | 13:38.49 |
| Robin_Watts | marcosw: Hmm. I may need your help with some perl here... | 13:38.58 |
| | Currently in the cluster if I clusterpush: gs bmpcmp filter=plank | 13:39.21 |
| chrisl | http://git.ghostscript.com/?p=ghostpdl.git;a=commitdiff;h=1225753b and http://git.ghostscript.com/?p=ghostpdl.git;a=commitdiff;h=c5cc5db0 | 13:39.36 |
| Robin_Watts | it splits that into 2 jobs: gs filter=plank and bmpcmp filter=plank | 13:39.43 |
| | That's arguably correct. | 13:39.50 |
| marcosw | kens: okay I'll test the fix listed in details.htm and if it works send it to the customer. | 13:39.54 |
| | chrisl: thanks. | 13:40.10 |
| chrisl | marcosw: of course, 8.71 is unsupported | 13:40.21 |
| kens | marcosw, OK its not my area, so I could be mistaken | 13:40.36 |
| Robin_Watts | but "bmpcmp filter=plank" currently doesn't work, cos it sends filter=plank to bmpcmp as an arg, and we get no (useful) results. | 13:40.37 |
| | The problem appears to be in build.pl,at line 55ish | 13:40.59 |
| | oops. line 49ish | 13:41.21 |
| | where we assume that any options after the filename are bmpcmpOptions. | 13:42.00 |
| marcosw | Robin_Watts: the bmpcmp clusterpush shouldn't need the filter=plank line; bmpcmp reads the users latest clusterpush results and builds the commands based on the files listed. I'm not sure what the easiest way to fix this is. | 13:43.34 |
| Robin_Watts | marcosw: Right, but I'd like the ability to say: bmpcmp only the plank files that failed. | 13:44.42 |
| | Removing the filter= lines from bmpcmp commands is easy to do in clustermaster, but that's limiting I reckon. | 13:45.36 |
| marcosw | oh, I misunderstood. That's a different issue but probably easy to fix. Give me a minute... | 13:45.37 |
| | I'm going to reenable henrysx6. If it generates another compiler error and I don't notice please email me. | 13:59.29 |
| Robin_Watts | mvrhel_laptop: Hi. | 14:51.56 |
| | You there now, or still travelling ? | 14:52.09 |
| mvrhel_laptop | Robin_Watts: at airport now | 14:52.38 |
| Robin_Watts | I have some cases for which the new code still doesn't match the old code. | 14:52.59 |
| mvrhel_laptop | ok | 14:53.12 |
| Robin_Watts | but it's closer with my formulation than it was with the old new formulation. | 14:53.14 |
| | (if you follow me) | 14:53.21 |
| | http://ghostscript.com/~robin/arm2.pdf | 14:53.42 |
| mvrhel_laptop | yes I understand | 14:53.50 |
| Robin_Watts | Draw that at 300dpi with tiffg4 and compare it to the png16m | 14:54.04 |
| | blinking back and forth between the two, we shift slightly to the right in the fast code. | 14:54.24 |
| | Compared to USE_FAST_CODE = 0 with the png, it doesn't shift at all. | 14:54.57 |
| | I think we need to move the dda by half a step. | 14:55.18 |
| | but I can't immediately see a macro to do that (which is embarassing, as I rewrote all these macros a while ago) | 14:55.44 |
| henrys | I knew this HPGL/2 thing was going to explode into my full time occupation - sigh... | 14:55.53 |
| mvrhel_laptop | Robin_Watts: So are you fine digging further into this and I will look at the issue with the level and phase | 14:59.06 |
| Robin_Watts | mvrhel_laptop: Sure. | 14:59.47 |
| mvrhel_laptop | ok thanks | 14:59.50 |
| Robin_Watts | Following on from my idea of putting an SSD in my laptop, I think I've decided to go for a hybrid SSD disc. 750Gig for 120quid. Not as fast as a raw SSD, but much cheaper/more storage, and apparently gives much of the same feel. | 15:01.15 |
| kens | I'll be interested to hear what you think | 15:01.58 |
| Robin_Watts | I'll order it next week, so will (hopefully) know by the staff meeting. | 15:02.31 |
| henrys | Robin_Watts:rotating silicon? | 15:02.49 |
| Robin_Watts | 7200rpm disc + 8gig of SSD cache. | 15:03.08 |
| mvrhel_laptop | henrys: was the little blurb about pdfx3 output intent what you needed? | 15:03.10 |
| henrys | yes I sent everything to scott, you guys get to review it when the newsletter is published ;-) | 15:03.48 |
| Robin_Watts | http://www.ebuyer.com/321969-seagate-750gb-momentus-xt-ssd-st750lx003 | 15:03.53 |
| mvrhel_laptop | plane boarding. bbiaw | 15:04.10 |
| | sideways rain here in michigan | 15:04.22 |
| Robin_Watts | The timings in that video are very impressive. | 15:07.19 |
| | For a while now (15 months), we've had 2 formulations of the dda stuff in gxdda.h. Both equivalent, except the one that's enabled is faster. | 15:24.42 |
| | Anyone object to me removing the slow one speak now... | 15:24.53 |
| | no one? great. | 15:25.00 |
| | Morning ray_laptop | 16:28.33 |
| ray_laptop | Hi, Ro | 16:30.43 |
| | oops. missed the tab | 16:30.55 |
| | Hi Robin_Watts | 16:30.59 |
| Robin_Watts | How was the beach? | 16:31.18 |
| ray_laptop | great ! | 16:31.25 |
| | it was 77 F and the water was the warmest in the last 10 or so trips, so we stayed in the water for 2+ hrs playing in the somewhat rough waves | 16:32.38 |
| | we all got tumbled a few times on our boogie boards, but no injuries -- just tired muscles | 16:33.17 |
| | got a few really good runs on the boards when I caught the wave just right | 16:34.09 |
| | (the tumbles were when I was trying for a big one and missed) | 16:34.38 |
| Robin_Watts | marcosw: Here's a case in point. I just run a test that had 800 failures. | 16:38.48 |
| | and pretty much every file that failed failed in pbmraw.300.0 pbmraw.300.1 and pbmraw.72.0 | 16:39.18 |
| | I'd like to be able to: git bmpcmp filter=pbmraw,72 to save time. | 16:39.34 |
| ray_laptop | Robin_Watts: oh, so this would do the run and automatically follow with the bmpcmp ? | 16:41.25 |
| Robin_Watts | ray_laptop: You can already do that, by doing: git cluster bmpcmp gs say | 16:42.06 |
| ray_laptop | Robin_Watts: so what is it that you'd like to be able to, but can't, do | 16:42.49 |
| Robin_Watts | but 1) you can't currently do: git bmpcmp filter=blah because the cluster passes filter=blah as an argument to the bmpcmp exe rather than filtering the jobs list. | 16:43.10 |
| | and hence 2) you can't do: git cluster bmpcmp gs filter=blah | 16:43.34 |
| | Currently bmpcmp does "bitmap compares of the jobs that failed in the last test" | 16:44.11 |
| | I'd like to be able to make it do "bitmap compares of a filtered set of the jobs that failed in the last test" | 16:44.30 |
| ray_laptop | wow, stars must be in alignment. Two PDFFitPage bugs in the same day ! (one against 8.71 and one for 9.05). The one from cust 531 is 9.05, but they are fitting to -g10200x66000 and reversing the page order | 16:52.13 |
| marcosw | Robin_Watts: I understand what you want and agree it would be useful; I just wish the cluster code didn't look like it had been written by a large number of monkeys. | 16:52.34 |
| | :q | 16:52.36 |
| Robin_Watts | ook. | 16:52.41 |
| marcosw | wrong window :-) | 16:52.42 |
| ray_laptop | interesting pagesize. I thought most people would be using PDFFitPage to produce thumbnails. That's a BIG thumb ! | 16:53.05 |
| | alexcher: I recommend that you try cust 130's bug with 9.06 and tell them to upgrade (of course they probably won't) | 16:54.24 |
| | off to get coffee... | 16:54.45 |
| henrys | marcosw:let me know if you want a meeting today, didn't look like we needed on yet. | 17:02.48 |
| | s/on/one | 17:04.47 |
| kens | Heading off for the night, bye all.... | 17:14.09 |
| chrisl | is also off for the evening..... | 17:16.37 |
| Robin_Watts | hmm. My macbook seems to have lost 'make' :( | 17:19.17 |
| ray_work | Robin_Watts: that's a strange one | 17:45.18 |
| Robin_Watts | It's a side effect of upgrading to Mountain Lion. I need to reinstall Xcode. | 17:45.51 |
| GhostShelly | henrys:anyone heard from Hin-Tak recently? I am waiting for a reply on a bug. | 18:21.11 |
| Robin_Watts | GhostShelly: Hi. | 18:21.34 |
| GhostShelly | Robin_Watts: hey robin good to finally catch you, although isn't it time to go enjoy the weekend? | 18:22.16 |
| Robin_Watts | GhostShelly: I'm about to go enjoy a holiday for a week. | 18:22.43 |
| GhostShelly | Robin_Watts: have a great one, how will you cope without GS for a week? | 18:23.21 |
| Robin_Watts | I'm taking a laptop :) | 18:23.33 |
| GhostShelly | Robin_Watts: at least your family will still have a good time :-P | 18:24.34 |
| henrys | GhostShelly:no haven't heard anything, I'll get to your email shortly, usually I won't be able to respond right away but it goes on my todo list and I get to it eventually | 18:26.29 |
| GhostShelly | henrys: no problem, just wanted to make sure the shelly spam filter did not kick in. | 18:27.13 |
| marcosw | Robin_Watts: I've added the ability use filter= with clusterpush.pl bmpcmp. It's barely tested so I'd be surprised if it works. | 18:27.16 |
| | henrys: sorry, lost track of time, again. No, I don't think we need a meeting. | 18:27.52 |
| GhostShelly | good night guys, time to go enjoy the weekend. | 18:28.19 |
| henrys | bye GhostShelly | 18:28.28 |
| | okay marcosw | 18:28.32 |
| Robin_Watts | you too | 18:28.34 |
| ray_laptop | yuck. The plane_extract device is SEVERELY bit-rotted. | 18:43.15 |
| Robin_Watts | Why are you needing the plane_extract device again? | 18:43.36 |
| ray_laptop | Robin_Watts: bmpsep1 uses it, and it hadn't been updated to use the planar memory buffers | 18:44.55 |
| | arguably we could just fix bmpsep1 and rip out the plane_extract device, I suppose | 18:45.45 |
| ray_laptop | checking to see if anyone else uses it... | 18:46.01 |
| | Robin_Watts: hmm... the 'pksmraw' and 'pamcmyk32' devices also use it | 18:48.25 |
| | at least I don't find any 'contrib' devices devices | 18:51.10 |
| | oops -- was in cust 532's directory -- there _is_ a contrib device | 18:51.31 |
| | contrib/gdevbjc_.c | 18:51.44 |
| | The main thing wrong is that it hadn't been updated with any of the newer procs that use HL color (such as fillpage) | 18:52.17 |
| | I think I'll just fix it. | 18:52.44 |
| | Robin_Watts: can I request a regression run against a device that doesn't usually run ? | 18:53.16 |
| Robin_Watts | ray_laptop: No. | 19:24.21 |
| ray_laptop | Robin_Watts: thanks. | 19:24.34 |
| Robin_Watts | What would it have to compare against? | 19:24.43 |
| ray_laptop | OK, bug 690963 works -- now to run lots more files through. | 20:09.53 |
| | at least I can break anything, since we don't test it and it never was close to working. | 20:10.43 |
| henrys | Robin_Watts, marcosw:seems reasonable that a device known to ghostscript but not to the cluster should produce a checksum database and then run the test. I think we'd get into the habit of checking more devices if we had that feature. | 20:23.32 |
| Robin_Watts | henrys: We could do a thing like bmpcmp, where it ran the current head binary to produce an md5 for a file, then ran the new version. | 20:26.46 |
| | henrys: When you install the new Xcode, you also need to go to the behaviours thing within it, and install the command line tools separately. | 20:29.29 |
| | oh, marvellous. Now it's failing to find the X headers :( | 20:30.09 |
| henrys | I dread the Xcode install... | 20:36.00 |
| Robin_Watts | Apple have dropped X11 for mountain lion. | 20:37.07 |
| | they are relying on XQuartz. | 20:37.14 |
| | Apple are missing a trick by not offering a "downgrade to Snow Leopard" on the app store. | 20:37.55 |
| | ok. reinstalling Xquartz has solved it. | 20:39.46 |
| henrys | it sounds like apple is going to keep X11 working though... http://support.apple.com/kb/HT5293 | 20:40.25 |
| Robin_Watts | I hope so! | 20:40.50 |
| | I've pushed a patch to my gs repo on casper that I hope will solve the scaling issues properly. It tests out with lots of diffs, but they all look like the right sort of diffs | 20:41.43 |
| | (at least the first 10 pages worth did, then I got bored). | 20:41.54 |
| | I'd like mvrhel to double check it before I commit it though. | 20:42.08 |
| | I'll send him an email in case he doesn't read this. | 20:42.22 |
| mvrhel_laptop | Robin_Watts: you there? | 23:26.31 |
| | I think this looks fine | 23:30.17 |
| ray_laptop | Hi, mvrhel_laptop | 23:35.48 |
| | sort of late for Robin_Watts, but you never know | 23:35.58 |
| sebras | Robin_Watts: I updated tor8 on the ESTA, so its in his calendar. | 23:49.38 |
| mvrhel_laptop | hi ray_laptop | 23:55.46 |
| | yes | 23:55.48 |
| | I will try to catch him this weekend | 23:55.56 |
| | Forward 1 day (to 2012/08/11)>>> | |
IRC Logs