| <<<Back 1 day (to 2015/09/29) | 20150930 |
kens | Something really weird happening with cluster tests. I cluster p[ushed my latest commit and it was clean, I pushed it and the test result comes back with 7750 files producing errors. My change was in pdfwrite, most of the errors are produced in straight interpreting. Same thing happened to me on a cluster test yesterday too :-( | 12:03.33 |
Robin_Watts | rebooting. brb. | 12:04.13 |
kens | Hmm all the failing results are on peeves.... | 12:08.06 |
| Yes, every test on peeves returned '1', all the other nodes were fone. | 12:09.44 |
| In every case peeves' error messdage was "GPL Ghostscript 9.19: Can't find initialization file gs_init.ps." | 12:10.17 |
marcosw1 | kens: odd that peeves fails in this way. I haven't made any changes to the cluster in weeks. I'll look into it. | 12:52.22 |
kens | marcosw1 I'm not at all sure what the problem is, beyond it not being my code changes!¬ | 12:58.10 |
marcosw1 | kens: when I manually run one of the jobs from the logs it works fine. trying a clusterpush now | 12:58.59 |
kens | marcosw1 I did another cluster push and it was OK, it seems to be highly intermittent | 12:59.23 |
| I'm unsure what would cause the problem too..... | 12:59.40 |
marcosw1 | kens: my clusterpush worked as well. I'll take peeves offline until I'm more awake and can think. | 13:00.12 |
kens | There's no rush | 13:00.49 |
| I had the same problem yesterday on a test and just reran it, then thought no more about it, the only reason I looked at ths one is because it was from a commit (I'm very glad I cluster tested before the commit). | 13:01.30 |
| Hopw peeves is unable to find gs_init.ps baffles me though | 13:01.49 |
Robin_Watts | ok, electrical testing rescheduled for tomorrow. | 14:36.23 |
rayjj | kens: your problem was with peeves, not peeved, right ? | 14:51.13 |
kens | yes correct | 14:51.24 |
Robin_Watts | trusts that rays next machine will be "irate". | 14:52.13 |
rayjj | or ticked | 14:52.31 |
| marcosw1: I notice that peeved is disabled. Is it still misbehaving ? | 14:53.31 |
| Logging into it no longer has delays for me. | 14:54.17 |
kens | chrisl I assigned that isntaller bug to you but I'm inlcined to close it as 'wontfix' | 15:09.28 |
chrisl | kens: I agree with your comment. The only thing I *might* do is refuse to install over a previous install | 15:10.07 |
kens | That makes sense | 15:10.48 |
chrisl | Of course, I've no idea *how* I would do that...... | 15:11.08 |
kens | For the next release :-) | 15:11.09 |
chrisl | I'd *really* like to get the revisions to the installer finished at some point | 15:11.34 |
kens | could check for the existence of bin/gswin32 or bin/gswin64 | 15:11.35 |
chrisl | Well, yes, but how to mallet nsis into doing that is another story | 15:12.05 |
kens | cannot help with that | 15:12.20 |
chrisl | I'll spend some time on the installer in the next couple of months - I think I had it working quite well | 15:12.59 |
marcosw1 | rayjj: yes, peeved still takes a long time to compile ghostscript. | 15:50.28 |
| chrisl: can I ask you stupid questions about compiling ghostscript under windows? | 15:50.57 |
Robin_Watts | marcosw1: You can ask stupid questions, and if chrisl is unavailable, and I can answer them, I will. | 15:52.01 |
| Though I would point out that asking stupid questions is generally my area of responsibility. | 15:52.44 |
chrisl | I'm here - just catching up with some e-mails..... | 15:52.56 |
marcosw1 | Robin_Watts: you haven't heard my questions :-) | 15:53.10 |
kens | GMail has been spam binning gs-bugs for me again :-( | 15:53.15 |
marcosw1 | I'm trying to automate regression testing under windows, so am building Ghostscript using the command line with nmake. | 15:53.20 |
| I have Visual Studio 2013 installed in Windows 7.1. | 15:53.49 |
kens | This permits you to test a new version before switching for example.You need the visual studio command propt too, or you need to adjust the paths in your shell using vcvars.bat | 15:54.12 |
| Oops, paste error | 15:54.21 |
Robin_Watts | The supplied VS projects will not work with 2013 without being 'imported'. | 15:54.38 |
| You would do better to use VS2005. | 15:54.49 |
| or VS2008. | 15:55.00 |
chrisl | Not relevant when using nmake directly | 15:55.05 |
kens | I htnk 2008 imports them too | 15:55.06 |
| But chrisl is correct. | 15:55.15 |
Robin_Watts | 2010 imports, 2008 does not. | 15:55.19 |
| Fair enough, yes. | 15:55.23 |
marcosw1 | First of all, it's taking 15 minutes to build, that seems like a long time. | 15:55.30 |
Robin_Watts | sorry, I'll shut up. | 15:55.33 |
chrisl | marcosw1: is that in a VM? | 15:55.43 |
kens | Robin_Watts : I'm on 2008 team stusio, and it insists on importing them | 15:55.50 |
marcosw1 | chrisl: yes. | 15:56.00 |
Robin_Watts | marcosw1: nmake doesn't do parallel builds. | 15:56.10 |
marcosw1 | Robin_Watts: yeah, I figured that out. | 15:56.20 |
chrisl | Then that's about the same as I get in a VM | 15:56.22 |
Robin_Watts | jom does, but last I checked, it broke with gs. There was a bug open. | 15:56.32 |
marcosw1 | Robin_Watts: It doesn't really matter, once I get it working the compile time is not an issue. it's only while I'm testing that it's annoying. | 15:57.03 |
chrisl | So, I haven't actually seen a question..... | 15:57.35 |
marcosw1 | anyway, I've managed to get it to build in the past with this configuration, but now am seeing missing ws2_32.lib issues. | 15:58.00 |
| is that something I screwed up? | 15:58.10 |
kens | Are you using the visual studio command prompt, did you run vcvars.bat ? | 15:58.34 |
marcosw1 | I am not using the visual studio command prompt, but am running "(long path)\Common7\Tools\VsDevCmd.bat" | 15:59.23 |
kens | I am not certain that's the right thing | 15:59.41 |
chrisl | IIRC, ws2_32.lib is for the Windows printing stuff | 15:59.54 |
kens | Possibly you don;t have the environment variables set up correctly. However, I am not running VS 2013, so that could be correct | 16:00.08 |
| I htought it was the sockets library | 16:00.40 |
| In fact it is | 16:00.53 |
chrisl | Yes, I thought it was a dependency for one of the print queue libs, but seems not. Seems it's there for cups..... | 16:01.50 |
kens | Hmm, well I admit I don't know why it would bethere. | 16:02.11 |
marcosw1 | chrisl: right. forgot to mention that I'm building with cups support. | 16:02.13 |
kens | Since we dropped cups from WIndows, maybe we don't need ths ? | 16:02.21 |
marcosw1 | i.e. WITH_CUPS=1 | 16:02.22 |
kens | ah | 16:02.28 |
marcosw1 | we dropped cups from windows? | 16:02.32 |
| how am I going to regression test it? :-) | 16:02.42 |
kens | From the default windows build, I thnk, yes | 16:02.47 |
| How were you going to run CUPS on Windows ? | 16:03.00 |
chrisl | The cluster just deals with cups rasters | 16:03.16 |
kens | I know, I know, you weren't | 16:03.19 |
marcosw1 | kens: the same way i run it on linux. generate a raster. | 16:03.21 |
chrisl | So, either VS2013 doesn't come with ws2_32.lib, or something strange is happening..... | 16:04.20 |
kens | According ot the quesions on SO, it is in VS 2013 | 16:04.36 |
marcosw1 | the thing is, I'm 100% sure this used to work and I haven't made any changes to anything. I'm still testing with some ancient ghostpdl source tree and haven't touched the VM since I set it up two years ago. | 16:05.19 |
kens | Can you find the .lib file on the VM ? | 16:05.40 |
marcosw1 | I'm using .bat files to launch nmake, so I'm sure the options are correct. | 16:05.45 |
kens | There is a suggesiotn here that yo uneed the PLatform SDK, I'm not sure if that's true, but that's why I suggest you check to see if its present on your VM | 16:06.42 |
marcosw1 | there is a ws2_32.dll in \windows\system32 | 16:06.54 |
kens | No, you need the .lib | 16:07.03 |
marcosw1 | also in \Windows\SysWow64 | 16:07.08 |
kens | You cna't statically link against a dynamic link library | 16:07.28 |
| You need the .lib file | 16:07.34 |
| Its apparently in c:\program files\Microsoft Visaul Studio 8\VC\PlatformSDK\Lib | 16:08.16 |
| YMMV | 16:08.19 |
marcosw1 | nevermind, I have four copies of ws2_32.lib (apparently the finder thingy doesn't show .lib extensions (but does show .dll, so that's fun]. | 16:10.09 |
kens | You can change that (I have because I hate it). WHere are they ? I have 14 copies myself :-) | 16:10.52 |
| So I'd have to guess none of those .lib files is in the linker path used by nmake | 16:11.25 |
marcosw1 | kens: not sure, can't figure out how to change the search results column widiths. | 16:11.26 |
kens | grab the little | between columns and pull it around | 16:11.43 |
marcosw1 | kens: nope, no | between columns | 16:11.56 |
kens | Hmm what version of WIndows is ths ? | 16:12.06 |
marcosw1 | 7.1 | 16:12.11 |
kens | Weird, that's what I'm using. Oh well | 16:12.22 |
pedro_mac | I think the newer VS installs use platform SDKs fro Program Files (x86)/Windows Kits/<ver> | 16:12.52 |
kens | Anyway, I would guess that the paths are the problem somewhere, the shell you are using and the nmake invocation aren';t finding those libraries | 16:13.01 |
marcosw1 | kens: I know. | 16:13.13 |
kens | pedro_mac : yes that's *one* of the places I have the libs :-) | 16:13.16 |
marcosw1 | but why? | 16:13.18 |
kens | That I'm less sure of. Especially without knowing where they are to be found (you haven't said) | 16:13.52 |
| It doesn't help that I'm not using VS 2013 of course | 16:14.23 |
marcosw1 | kens: I don't think it matters where they are. Someone where in c:\Program Files (x86)\ | 16:14.33 |
| they are wherever the installer put them. | 16:14.48 |
kens | Right, but are they installed as part of VS 2013, or as part of something else | 16:15.01 |
marcosw1 | they had to be installed with vs2013. this machine doesn't have anything else installed on it. | 16:15.20 |
kens | These thngs can be installed multiple ways (which is why I have 14 copies) | 16:15.21 |
| OK well that helps | 16:15.26 |
marcosw1 | i have a clean copy of win7.1 that I clone whenver I do anything with windows. | 16:15.41 |
| (possibly a technical violation of license terms but since I delete the vm immediately after I think I'm okay). | 16:16.26 |
kens | won't tell anyone | 16:16.46 |
marcosw1 | good thing we don't log irc. | 16:17.01 |
| speaking of licenses. I was going to buy another 7.1 license after I got this working since the regression vm will need to run 24/7 but Microsoft apparently stopped selling retail 7.1 last year. You can still buy oem versions of 7.1 but it's very unclear whether that's okay to use | 16:18.18 |
kens | Umm it seems MS changed the name ofthe batch file for 2013, so I'm kind of at a loss to help you since I don;t have it installed, anyone else using it ? | 16:18.59 |
| I guess hte other thing to do is to try it in the solution and see if that works | 16:19.41 |
marcosw1 | http://www.zdnet.com/article/is-it-ok-to-use-oem-windows-on-your-own-pc-dont-ask-microsoft/ | 16:19.56 |
mvrhel_laptop | ok lunch set with MQ for Friday | 16:20.45 |
marcosw1 | i'm happy to try a different version of Visual Studio. I tried 2015 first but it doesn't run on Windows 7.1 (installs without warning or error, but just doens't work). | 16:21.17 |
kens | I thnk that needs windows 8 or better | 16:21.58 |
marcosw1 | I don't think I want to use VS2005 or even VS2008. Anything made this decade okay to use? | 16:22.07 |
kens | I don't know, I'musing 2008 :-) | 16:22.23 |
| Michael must be using somethign more recent | 16:22.35 |
marcosw1 | kens: I have no idea what to say about the fact that you are using 7 year old development tools. Are you running on Vista or XP? | 16:23.30 |
kens | Neither Widows 7 | 16:23.46 |
| Just bought a new laptop fdor Melanie now that it comes with Windows 10 | 16:24.20 |
marcosw1 | but Windows 7 wasn't released until 2009. VS2008 was (presumably) released in 2008. | 16:24.22 |
chrisl | marcosw1: the binaries we ship are built with VS2005 | 16:24.23 |
marcosw1 | chrisl: on paper tape or punch cards? | 16:25.08 |
| can I ask why kens using VS2008 and chrisl VS2005? Are the newer versions that bad? | 16:25.50 |
chrisl | VS2005 for maximum compatibility | 16:26.08 |
kens | This permits you to test a new version before switching for example.I don't want to spend money on a newer version I don;t need, and I utterly detest the newer user interfaces | 16:26.18 |
chrisl | kens: you *really* need to sort out that copy'n'paste thing! | 16:26.55 |
kens | I know, I know :-( | 16:27.02 |
marcosw1 | kens: right, forgot that Microsoft charges for their development kits (and operating systems). | 16:27.05 |
| I remember the yelling when Sun started charging for their Sparc C compiler. Lot's of yelling on the internets. Wonder how Sun is doing... | 16:28.15 |
kens | Its defineitly holding down shift whihch pastes it, but I have that turned off in Windows, so I don't know where that's coming fomr :-( | 16:28.29 |
chrisl | So, looks like: Program Files (x86)\Windows Kits\8.x\Lib\winv6.3\um\x86 (or x64)\WS2_32.lib | 16:29.05 |
| WTF have the M$ people been smoking?!?! | 16:29.25 |
marcosw1 | chrisl: that sort of matches one of my locations: ...\Windows Kits\8.0\Lib\win8\u... | 16:30.44 |
| also ...\Microsoft SDKs\WIndows\v7... | 16:30.56 |
| the latter have a timestamp of 10/1/2012 and th former 8/2/2013. | 16:31.23 |
kens | TIme for me to go, good luck with bulding marcosw1 | 16:31.33 |
chrisl | marcosw1: what was the "issue" you saw with the library? | 16:31.51 |
marcosw1 | the linker doesn't find it. | 16:31.58 |
Robin_Watts | marcosw1: VS2005 will build for windows XP SP1. | 16:32.22 |
| Sorry. | 16:32.31 |
marcosw1 | btw, just to confuse things further, I'm using Windows 7.1 x64. | 16:32.38 |
Robin_Watts | VS2008 will build for XP SP1 I think. | 16:32.45 |
| VS2005 builds for earlier versions too. | 16:32.56 |
marcosw1 | and I do seem to have both 32 and 64 bit versions of ws2_32.lib installed | 16:32.59 |
chrisl | Hrm, it really should find it under Microsoft SDKs\Windows\v7.1A...... | 16:33.07 |
Robin_Watts | hence we use VS2005 to ensure maximum compatibility. | 16:33.12 |
marcosw1 | I have one there. | 16:33.44 |
| ...SDKs\Windows\v7.1A\Lib\x64 | 16:34.02 |
| and one in V7.1A\Lib | 16:34.38 |
chrisl | I'd think that should work. Do you get the same error for 32 bit and 64 bit builds? | 16:35.37 |
marcosw1 | here's the thing. the other lib specified in libc32.tr (e.g. winspool.lib) are in the same directory but are found no problem. | 16:35.43 |
| chrisl: I think so. Stopped trying 64 bit builds last night because it was taking so long. | 16:36.07 |
| (btw, there are more libs in libc32.tr, winspool.lib was just an example). | 16:36.29 |
chrisl | So, VS2012 was still using vcvars32.bat and vcvarsx86_amd64.bat to setup the command prompt environment.... did they really change the name in VS2013? | 16:37.48 |
marcosw1 | http://stackoverflow.com/questions/13774375/difference-between-vsdevcmd-bat-vcvarsall-bat-in-vs2012 | 16:38.21 |
| i can't recall why I'm not using vcvars32.bat. The compile .bat file is 2 years old. | 16:39.21 |
chrisl | Can you post the nmake command line? | 16:39.47 |
marcosw1 | did someone say that ws2_32.lib is only needed for cups? | 16:39.50 |
Robin_Watts | tor8: Had you reviewed the 3 commits on robin/master? | 16:40.13 |
chrisl | It was added for cups | 16:40.19 |
marcosw1 | nmake -f psi\msvc.mak BUILD_SYSTEM=64 MSVC_VERSION=12 | 16:40.23 |
| that's without cups obviously. trying that know. | 16:40.38 |
| s/know/now | 16:40.45 |
| I know I have notes somewhere that lists what I went through when figuring this out two years ago, just cant' find them. | 16:41.39 |
chrisl | I would suggest dropping the BUILD_SYSTEM=64 | 16:42.18 |
marcosw1 | really wishes it didn't take 15 minutes to compile. Now looking for windows ccache... | 16:42.29 |
| I can try that, but I suspect it's there for a reason. | 16:42.51 |
chrisl | Use real hardware | 16:42.52 |
Robin_Watts | marcosw1: With ccache, you'd just end up debugging ccache too. | 16:43.20 |
chrisl | marcosw1: it *was* there for a reason, that reason no longer exists - and I think is what's causing you problems | 16:43.34 |
marcosw1 | chrisl: nothing has changed, i'm compiling gs source from November 2013. | 16:44.28 |
| wish I had my notes... | 16:44.45 |
chrisl | Well, then I would suggest using more up to date sources! | 16:45.34 |
marcosw1 | thinks it's ironic that it's been suggested I use earlier development systems AND new source | 16:46.27 |
chrisl | Well, take your pick - use old source and old VS, or newer source and newer VS..... | 16:46.56 |
marcosw1 | seems like vs2013 and source from 2013 shouldn't be a problem... | 16:47.36 |
| anyway, I was just trying to ask if this is a known issue. I'll keep working on it. | 16:48.48 |
chrisl | VS2013 was released mid-October 2013 | 16:48.50 |
henrys | sent official word to URW to start the project ;-) | 16:49.13 |
marcosw1 | make sense. apparently I worked on this in November 2013. | 16:49.22 |
henrys | should have been :-) | 16:49.37 |
Robin_Watts | henrys: Excellent. | 16:49.49 |
chrisl | Cool - I wonder how long..... | 16:50.01 |
| marcosw1: I'd really urge you to use the current code (or at least the RC) | 16:50.39 |
henrys | chrisl: hopefully before GG discovers what they are doing ... but seriously it was an 8 week estimate | 16:51.28 |
Robin_Watts | If marcosw1 is envisaging using this machine as something to build old versions of the source on, then using an older toolchain might make sense. | 16:51.55 |
chrisl | henrys: I gotta think GG will want this too, so if we pay for it, they get it free...... | 16:52.07 |
Robin_Watts | let's not mention that to Miles, eh? :) | 16:52.49 |
chrisl | It's not a new thing | 16:53.33 |
henrys | chrisl: sure but you know some asshole manager gets hears about it and starts making noises... who knows | 16:53.46 |
| s/hears/word | 16:53.57 |
chrisl | Hopefully, the worst of the asshole managers aren't there any more | 16:54.13 |
| I would prefer marcosw1 to use VS2005 at least as a baseline, since it's what we use for our binaries | 16:54.45 |
henrys | we spent a good deal of time with kendall and his collegues in chicago, great folks I thougth. | 16:54.57 |
| but we are competitors. | 16:55.22 |
chrisl | I'll be impressed (and relieved) if they get it done in 8 weeks - bearing in mind the last set we had done.... | 16:56.13 |
henrys | I forgot to bring up yesterday I've approved with Miles increasing the bounty what do ya'll think of that? I'm thinking it should be doubled. | 16:56.28 |
rayjj | chrisl: I thought there was a problem building 64 bit on VS 2005 | 16:57.47 |
tor8 | Robin_Watts: yes, 3 commits on robin/master are good to go | 16:58.16 |
rayjj | (maybe it's just my VS 2005 doesn't) | 16:58.18 |
Robin_Watts | tor8: Thanks. | 16:58.23 |
| rayjj: There are 64bit compilers you can install for VS2005. | 16:58.38 |
| IIRC. | 16:58.41 |
tor8 | Robin_Watts: 3 epub commits on tor/master | 16:59.07 |
| and I think we should be able to push the tor/glut branch to master as well | 16:59.25 |
| there's a commit on mujs tor/master if you want to review that as well, for the setjmp fixes | 17:00.13 |
| which we should pull into mupdf | 17:00.17 |
marcosw1 | chrisl: if I remove ws2_32.lib from libc32.tr and link manually it works. | 17:00.46 |
| chrisl: i'm happy to use vs2005, but can't figure out how to download the express edition. still looking... | 17:02.08 |
chrisl | I don't there was an express of 2005 | 17:03.16 |
marcosw1 | https://www.microsoft.com/en-us/download/details.aspx?id=804 | 17:03.29 |
| there is a service pack | 17:03.36 |
| chrisl: how do you propose I get a copy? Do we have a spare license? | 17:04.30 |
| (not for the express edition, but for the paid for one). | 17:04.53 |
Robin_Watts | tor8: /* TODO: reference count outline */ would seem important? | 17:06.04 |
| tor8: epub changes seem reasonable, other than that TODO. | 17:12.31 |
| and that todo may be ignorable if we have similar limitations elsewhere? | 17:13.22 |
tor8 | oh yeah, I should probably fix that before we push | 17:14.33 |
| the fz_link_dest structure is a bit inflexible and very pdf oriented :( | 17:15.04 |
| I might give that a remake when bashing more on forms after the release | 17:15.21 |
marcosw1 | rayjj: I just tested peeved and it does seem to be okay. I've re-enabled it and peeves (peeves was having odd issue before with ken's clusterpush, but I can't recreate the problem so am ignoring it for now). | 17:17.12 |
chrisl | marcosw1: I have to finish now. Good luck! | 17:25.11 |
marcosw1 | chrisl_away: thx | 17:28.16 |
| rayjj: never mind, there is still something strange going on with peeved. I'll write you an email describing the symptoms. | 17:28.49 |
| who needs fibre? I'm getting 10.9 MB/s sustained downloads from casper :-) I know, fibre is ~100MB/s. | 17:32.00 |
henrys | marcosw1: my friends are getting about 400 | 17:32.38 |
| marcosw1: I'm still waiting ... impatently | 17:33.51 |
Robin_Watts | I checked on fibre here yesterday. | 17:36.20 |
| Every house in my street is listed as being in an "Enabled Area". Except mine. | 17:36.38 |
| I think when I moved house 13 years ago, and brought the phone number with me, they didn't update the details of which telephone pole I'm connected to. | 17:37.21 |
| Hence they think my house is 3 miles away. | 17:37.31 |
| That's gonna be fun to sort out with BT :( | 17:37.40 |
henrys | marcosw: I've seen that gs_init.ps thing to but I cleaned my directory and couldn't reproduce it. | 17:43.42 |
| marcosw: not during a cluster push | 17:44.00 |
marcosw1 | Robin_Watts: probably easier if you just move back to the old house. | 17:44.24 |
| henrys: my least favourite words: "could not reproduce it" | 17:44.53 |
Robin_Watts | Well, I buy broadband through a reseller of BT, eclipse. Eclipse have good customer service. Eclipse can sort it out with BT :) | 17:45.25 |
henrys | Robin_Watts: so you need ethernet cable to your neighbors ? | 17:45.29 |
Robin_Watts | henrys: Hell no. One of my neighbours is a direct customer of BT, the other is a Talk Talk customer. I'll live with ADSL rather than touch either of those :) | 17:46.24 |
henrys | I forget what you're ADSL speeds are but I recall they are a little low. | 17:47.13 |
Robin_Watts | 18 down, 1 up. | 17:48.33 |
marcosw1 | are those MB or Mb per sec? | 17:50.18 |
Robin_Watts | Mb | 17:50.23 |
marcosw1 | ouch, I have 150Mb/s down (and I think 20Mb/s up). | 17:50.55 |
henrys | Robin_Watts: do you game online? I guess latency is more a factor with that. | 17:51.52 |
Robin_Watts | ADSL2 tops out at 24Mb/s down, 1Mb/s up. | 17:52.09 |
| henrys: I don't generally play online games, but even if I did, I think my speeds would be fine. | 17:52.29 |
| Fibre is not critical for me, but it would certainly be nice. | 17:53.03 |
henrys | marcosw1: is that business class or something I get around 60 down 6 up with cable. | 17:56.22 |
| ? | 17:56.23 |
Robin_Watts | marcosw: Are you DSL/Cable/Fibre? | 17:57.50 |
| tor8: Are you here? | 18:30.24 |
marcosw1 | henrys: no, it's consumer. What comcast calls Blast. | 19:08.27 |
| Robin_Watts: cable. I get internet/tv/phone through one coax. | 19:09.09 |
mvrhel_laptop | fixed a pile of issues with gsview. now the 3001 page stress document works pretty reasonable. as soon as I add a couple more things to the forms capability (listbox and signatures) I will make a new release | 19:11.29 |
marcosw1 | henrys: looks like I might have Blast 150 | 19:11.37 |
mvrhel_laptop | lunch time now | 19:11.46 |
fredross-perry | mvrhel_laptop: check your email for something from me regarding gsview. Thanks. | 19:12.15 |
mvrhel_laptop | fredross-perry: ok thanks | 19:12.26 |
| fredross-perry: I don't see anything... | 19:13.08 |
fredross-perry | feh. hang on⦠| 19:13.19 |
mvrhel_laptop | oh | 19:13.44 |
| went in the trash.... | 19:13.49 |
| sorry about htat | 19:14.05 |
fredross-perry | ha | 19:14.10 |
| just sent it again. | 19:14.16 |
mvrhel_laptop | fredross-perry: so let me add in some error checking that you can copy to check that the profiles are appropriate | 19:15.18 |
rayjj | marcosw: is peeved still a problem ? It is still "disabled", but I think it should be OK now | 19:15.19 |
fredross-perry | that sound fine. Thanks. | 19:15.41 |
mvrhel_laptop | fredross-perry: I will also send you a couple profiles to use for testing | 19:15.46 |
| which should give more saturated looking colors | 19:15.56 |
| but it may not be until later this week | 19:16.08 |
fredross-perry | good as well. OK | 19:16.12 |
mvrhel_laptop | bbiaw | 19:16.34 |
tor8 | Robin_Watts: I'm here now. | 19:30.49 |
Robin_Watts | tor8: I am confused by something in pdf_repair.c | 19:31.13 |
| In pdf_repair_obj, we call pdf_parse_dict and we pass a NULL doc. | 19:31.51 |
| The comment says "Send NULL xref so we don't try to resolve references" | 19:32.10 |
| How would we ever try to resolve refs when parsing a dict? | 19:32.22 |
tor8 | if it's the encryption dictionary, we read from it in the next if-block | 19:34.16 |
| and then we read the /Length | 19:34.25 |
| etc. | 19:34.26 |
rayjj | henrys: et. al, it seems that the company M board I have is dead. When I plugged it in, I heard a 'pop' and the lights never went on. I put my DVM on it and the 5v to the board is OK, but when I connect the jumper to pass it to the board, it goes to 0.00v, so it is somehow toasted | 20:10.20 |
| mvrhel: henrys: I don't think it is worth it to request the older board, but if you want to, fine and I will replicate the previous tests with the current code | 20:11.33 |
| henrys: in the meantime, I will fire up my Pi and run the tests there -- assuming I get similar results, then we can consider performance analysis on the Pi (if we start with results that are close, profiling should be valid) | 20:12.51 |
| henrys: NB: I'll run the Pi with the same params, and at 800 MHz which is what the M board ran | 20:13.37 |
rayjj | did what I did to trace the board, but without schematics and layout, it's really hard on a mutilayer PCB with mostly surface mount compents. No Joy :-( | 20:14.56 |
rayjj | suspects a voltage spke that fried the voltage regulator (but I don't even know if there is just one, or several) | 20:15.52 |
| could be as simple as a shorted cap near the input power, but SM components are hard to deal with | 20:16.48 |
mvrhel_laptop | rayjj: the old part is still in production and used | 21:04.08 |
| It would probably be worthwhile to get another one as this chip is a different market price point I believe. I will talk to MQ about it on Friday | 21:04.57 |
| Forward 1 day (to 2015/10/01)>>> | |