| <<<Back 1 day (to 2015/04/22) | 20150423 |
kens | rayjj I have a copy of both the draft and final versions of the JPEG2000 specifications, if you need one at some point. | 08:07.32 |
| The spec says that an ICC profile (in a JP2k file) mauy contain a white point, but ths is ignored by the 'restricted ICC processing' of a JP2K file. As far as I can see there is no other way to specify a white point in JP2K | 08:12.49 |
Robin_Watts | I just renewed the godaddy hosting for the gs downloads. | 14:35.45 |
henrys | Robin_Watts: okay so I'll strike the agenda item. | 14:37.29 |
Robin_Watts | henrys: It's only $107, so if we want to swap somewhere else cost is not a massive block. | 14:38.03 |
| But I don't know that we could reasonably swap anywhere else and get a noticeably better service. | 14:38.29 |
| We run through terrabytes of downloads a month, IIRC, so a small fraction of those are bound to be broken. | 14:38.54 |
henrys | mvrhel_laptop: I never looked at this new license in gsview. | 15:06.20 |
| what the hell does it mean? ;-) it sounds like I *may* have a license agreement. | 15:07.41 |
mvrhel_laptop | henrys: it was from miles | 15:12.03 |
henrys | either we had an argument about the licensing of gsview and I lost and forgot or Miles snuck this past me. | 15:12.21 |
mvrhel_laptop | well one does want some sort of EULA in place | 15:12.52 |
henrys | mvrhel_laptop: well sure I wanted it to be AGPL I don't understand what it means to say "we may offer" it is as such. | 15:14.27 |
mvrhel_laptop | another item for work flowy | 15:15.10 |
| need to restart brb | 15:15.26 |
henrys | mvrhel_laptop: on a related note is there a "source" release we are going to add to gsview.com or no plan for that? I'm not sure what we should do about that. | 15:18.30 |
mvrhel_laptop | I am not sure either | 15:18.47 |
henrys | fair enough | 15:18.56 |
mvrhel_laptop | This needs to be talked about | 15:18.59 |
henrys | right agenda items... | 15:19.23 |
mvrhel_laptop | I suspect this is not going to be the quickest of meetings in June | 15:19.43 |
Robin_Watts | henrys: The source is available for gsview in the git repo. | 15:21.04 |
henrys | I'm fine solving that right now. Is there a technical issue with putting a source zip on the site. Will it be confusing to build with all the dependencies? | 15:21.32 |
Robin_Watts | There is no 'release' as such, and I can't see that that's a particular problem until it stabilises at least, right? | 15:21.55 |
mvrhel_laptop | I agree | 15:22.03 |
Robin_Watts | What license is gsview under? | 15:22.07 |
mvrhel_laptop | I am not real comfortable with this right now. I think it does need to be talked about | 15:22.22 |
kens | We're only beta now, so source is not a huge issue | 15:22.23 |
henrys | Robin_Watts: read it for yourself and if you understand it explain it to me ;-) | 15:22.32 |
kens | Robin_Watts : the licence is what henrys and mvrhel_laptop were talking about | 15:22.37 |
Robin_Watts | OK, so GSView is under a homebrew license. | 15:22.43 |
kens | I had read it, but shrugged | 15:22.47 |
Robin_Watts | As such, it can't be using Ghostscript or MuPDF under the GNU AGPL. | 15:23.13 |
| That's fine for us, cos it's our code. | 15:23.22 |
mvrhel_laptop | we pay ourselves | 15:23.27 |
kens | The licencing does need discussed though | 15:23.47 |
Robin_Watts | But it means that anyone that downloads gsview and builds their own copy can only do so under the AGPL. Which means they can't redistribute. | 15:23.56 |
kens | And if we go for a 'pro' paid for version we need to think about how we separate the sources | 15:24.06 |
mvrhel_laptop | right. I don't really see a whole lot of advantage to having the source code available to be honest with you | 15:24.37 |
Robin_Watts | kens: Indeed. | 15:24.42 |
| mvrhel_laptop: I agree. | 15:24.47 |
mvrhel_laptop | its not like it would be used as a library like mupdf or gs | 15:25.02 |
Robin_Watts | Well, by having the source code available we might get bug reports in with patches. | 15:25.10 |
kens | If we aren't going to make the source available, we should probably make it non-public in Git asap | 15:25.13 |
mvrhel_laptop | kens: yes | 15:25.22 |
Robin_Watts | If it was me, I'd just move the repo from /home/git to /home/git-private | 15:25.44 |
henrys | it totally depends on the license if we agree on AGPL we should put the code up. | 15:25.44 |
Robin_Watts | henrys: If we agree on AGPL we *have* to put the code up. | 15:26.02 |
kens | Maybe we should move it private untli we ahve the licence ironed out | 15:26.15 |
mvrhel_laptop | lets move it to /home/git-private now and then we will discuss further at the meeting | 15:26.24 |
Robin_Watts | But given that it has this homebrew license, it's just confusing having the source there. | 15:26.29 |
| mvrhel_laptop: That sounds like a good plan to me. | 15:26.35 |
henrys | Robin_Watts: no I think the git repo is good enough I'm asking if there should be a zip on mupdf.com? | 15:26.35 |
Robin_Watts | henrys: I get whiplash from that statement. | 15:26.58 |
henrys | okay until the meeting... | 15:26.59 |
mvrhel_laptop | Robin_Watts is it a simple move process or is there some git witchcraft that has to occur | 15:27.16 |
Robin_Watts | mvrhel_laptop: A simple mv should do it :) | 15:27.26 |
| want me to? | 15:27.55 |
mvrhel_laptop | sure | 15:27.59 |
| thanks | 15:28.00 |
henrys | I was just saying a git repo pointer satisfies the GPL we don't have to have a "tar.gz" or "zip" of the source. | 15:28.38 |
| but we are going private for now so it isn't important. | 15:29.23 |
Robin_Watts | moved. | 15:29.29 |
mvrhel_laptop | Thanks Robin_Watts | 15:29.42 |
Robin_Watts | mvrhel_laptop: You'll need to adjust the origin pointers from /home/git/gsview.git to /home/git-private/gsview.git | 15:30.01 |
mvrhel_laptop | ok thanks | 15:30.07 |
Robin_Watts | which is possibly easiest achieved by editing .git/config on your local machine. | 15:30.42 |
henrys | agenda updated! that'll heat things up in San Fran | 15:31.08 |
kens | Yeah, as Michael said, going to be a lengthy meetng ths time I expect. | 15:31.28 |
mvrhel_laptop | Robin_Watts: ok made the switch here | 15:32.10 |
| oh Robin_Watts I probably need to move my repos on gs | 15:33.19 |
Robin_Watts | on casper? yes. | 15:33.41 |
mvrhel_laptop | yes | 15:33.45 |
| thats what I mean | 15:33.47 |
Robin_Watts | just make a private-repos dir in your homedir and move the dir there. | 15:34.14 |
| I can do that if you want. | 15:34.19 |
mvrhel_laptop | I have one already for SOT | 15:34.27 |
Robin_Watts | of course. | 15:34.34 |
mvrhel_laptop | ok. I moved it | 15:36.46 |
henrys | mvrhel_laptop: we can make the meeting much less painful if engineering comes to consensus on the license status before the meeting. If it is open source I can convince miles in advance and we're done. If there is not consensus amongst engineering I think we should just plop the thing in Miles' lap and let him handle it. | 15:51.21 |
| or decide | 15:51.28 |
mvrhel_laptop | As I said earlier, I don't see a whole lot of advantage to making it open source at this time | 15:51.53 |
| So that is my opinion | 15:52.01 |
Robin_Watts | If we hope to charge for it, then it should not be open source. The hassle involved in forking 2 versions is too great. | 15:52.06 |
mvrhel_laptop | yes | 15:52.17 |
Robin_Watts | I can't believe we'll get meaningful patches from people for this app. | 15:52.32 |
mvrhel_laptop | I agree | 15:52.37 |
Robin_Watts | I think it should be closed source, at least initially. | 15:52.41 |
chrisl | OTOH, closed source will *never* get much penetration in the Linux world - we might as well give up on that now...... | 15:53.43 |
Robin_Watts | chrisl: That's freds project - not necessarily the same deal there... | 15:55.44 |
mvrhel_laptop | other topic for the agenda henrys... | 15:56.02 |
kens | Did anyone read the article about 'Device Guard' ? If we plan to charge for Windows, we will *need* to get a certificate in order to do signing | 15:56.31 |
chrisl | Robin_Watts: I tend to regard GSView as one project, with two parts..... I think we either open source both, or close both. | 15:56.53 |
henrys | I'm firmly in favor of open source but it ain't worth an argument to me. | 15:56.57 |
kens | Article about device guard: | 15:57.30 |
| http://www.theregister.co.uk/2015/04/23/microsoft_windows_10_device_guard/ | 15:57.30 |
mvrhel_laptop | I looked at the prices for certification yesterday. We really need this if we want to avoid the whole warning business during install | 15:57.48 |
henrys | kens: yeah I got this really obnoxious thing from chrome that was very confusing when I downloaded gsview and I accidentally deleted the download. | 15:57.54 |
mvrhel_laptop | oh I had not seen this for windows 10 though | 15:57.55 |
kens | Yeah, its obviously going to be ENterprise level, but we already have people whinging that our .inf file isn';t signed.... | 15:58.28 |
Robin_Watts | "It's designed to be a very tiny, tightly controlled secure execution environment." | 15:58.58 |
kens | Butits still WIndows :-) | 15:59.10 |
Robin_Watts | i.e. the first person to get a rootkit in there is going to be a millionaire. | 15:59.11 |
mvrhel_laptop | its like $1200 for a three year certificate | 15:59.30 |
Robin_Watts | mvrhel_laptop: That's cheaper than I'd seen before. | 15:59.43 |
kens | If we want to charge on WIndows we'll need it. | 16:00.00 |
henrys | was russell doing anything for that? | 16:00.00 |
mvrhel_laptop | yes. I thought it was not too bad | 16:00.01 |
kens | henrys I don't believe so | 16:00.08 |
Robin_Watts | I think it's got to the stage where Artifex having its own certificate is a no-brainer. | 16:01.19 |
henrys | on os X I guess we'd have to look into the "Gatekeeprer" business. | 16:01.34 |
kens | Robin_Watts: q we should use it for MuPDF on Windows also of course | 16:01.46 |
kens | knows nothgin about OS/X | 16:02.01 |
henrys | https://support.apple.com/en-us/HT202491 | 16:02.47 |
kens | Looks similar to the MS certfication thing | 16:03.21 |
henrys | I was hoping to get gsview in through "homebrew" or "ports" but without open source that won't work. | 16:04.03 |
| more for the agenda ugh | 16:05.24 |
mvrhel_laptop | have run to airport to pick up my father. bbiaw | 16:16.39 |
rayjj | btw, on certificates, I had looked previously and there are separate prices for drivers and apps. At the time I looked a cert for a driver (i.e. printer installer, even if it is just an inf) was about 3x the cost for a cert for an app | 16:31.10 |
Robin_Watts | You'd kinda have hoped that we could buy a single cert that would cover us for everything. | 16:31.49 |
| Our website, our apps, our drivers, etc... | 16:31.58 |
henrys | Robin_Watts: don't we have an apple cert for artifex that could be used for the gsview dmg? | 16:34.44 |
| fredross-perry: we're talking about a project for you on skype. | 16:35.13 |
Robin_Watts | henrys: I don't profess to understand much about certs in general - on MacOS even less so. | 16:35.53 |
fredross-perry | hello. | 16:35.56 |
henrys | Robin_Watts: I'll look into it. | 16:36.16 |
Robin_Watts | We're ios developers, and so we get a signing key for our apps as part of that. | 16:36.20 |
rayjj | Robin_Watts: I think if we have a driver cert, we can use it for apps | 16:36.22 |
Robin_Watts | I suspect that if we want a signing key for our apps on macos, we need to be mac developers. | 16:36.49 |
henrys | Robin_Watts: I was wondering if it is ios and os X | 16:36.51 |
fredross-perry | I think certs for iOS and OSX are separate. | 16:37.19 |
Robin_Watts | https://developer.apple.com/programs/which-program/ | 16:37.21 |
rayjj | I was going to use gsview as my default pdf viewer, but the need to use the mouse to zoom is a killer (IMHO) | 16:38.00 |
Robin_Watts | So I'm thinking that we need to be both ios and mac developers. At $99 a year, that ain't no big thing. | 16:38.11 |
| rayjj: ctrl->mousewheel? | 16:38.24 |
| rayjj: How would you like to be able to do it? | 16:38.33 |
henrys | ctrl-+ and - is pretty much expected. | 16:38.52 |
Robin_Watts | OK. I've never used ctrl+ and - personally. | 16:41.22 |
| ctrl + mousewheel is easier than clicking and typing at the top. | 16:41.59 |
rayjj | Robin_Watts: I want hot-keys to do it, as henrys says, <ctrl>+ and <ctrl>- are commonly used, although some apps also accept just + and - | 16:42.31 |
fredross-perry | I use ctrl + and - all the time. gsview mac and linux have this. I can look at mouse zooming if many think itâs important. I personally do not. | 16:42.40 |
Robin_Watts | rayjj: I cannot argue with that. | 16:42.43 |
henrys | touching the mouse is like a context switch ;-) | 16:42.45 |
Robin_Watts | fredross-perry: I think we should take ctrl-mousewheel too - browsers and editors etc generally accept it. | 16:43.23 |
rayjj | and the gsview "zoom control" at the bottom is a pain, given that the rest of the controls are along the top | 16:43.25 |
Robin_Watts | rayjj: At the bottom? | 16:43.49 |
| on windows? | 16:43.51 |
rayjj | Robin_Watts: yes, on Windows | 16:44.00 |
henrys | rayjj: can you do an enhancement request for gsview so we don't forget? | 16:44.15 |
rayjj | henrys: OK | 16:44.28 |
Robin_Watts | rayjj: The zoom controls are all along the top for me, and always have been. | 16:44.39 |
| I'm on windows. | 16:44.48 |
| Windows 7 in case that makes a difference. | 16:45.33 |
rayjj | Robin_Watts: there's the Options/Zoom Control item. I didn't know what it did, but it adds another control along the bottom with a zoom slider. Pretty worthless, IMHO | 16:46.08 |
| Robin_Watts: I'm on 7 as well | 16:46.24 |
fredross-perry | as far as mac-ness is concerned, in Preview the scroll wheel scrolls, not zooms. The trackpad, on the other hand, does pinch-zooming which is nice. We could support that I suppose. | 16:46.39 |
Robin_Watts | Options => Output Intents/Show Annotations/Antialias - that's all for me. | 16:47.00 |
| fredross-perry: ctrl + scroll wheel? | 16:47.21 |
rayjj | fredross-perry: I agree that if the track pad has pinch (or the touchscreen), that's a good way to zoom | 16:47.38 |
Robin_Watts | scroll wheel scrolls on windows. ctrl + scroll wheel zooms. | 16:47.54 |
fredross-perry | I canât get Preview to zoom using <any-modifier>-scroll wheel. | 16:48.14 |
Robin_Watts | fredross-perry: How about (say) firefox or chrome ? | 16:48.34 |
henrys | fredross-perry: really I get a scroll | 16:48.47 |
rayjj | Robin_Watts: I just downloaded two days ago. My version has Output Intents/ Zoom Control/Show Annotations/AntiAlias | 16:48.54 |
fredross-perry | Adobe Reader zooms with option-wheel | 16:49.05 |
| henrys: ?? | 16:49.38 |
| Preview DOES scroll with the wheel. It just doesnât zoom using the wheel. | 16:50.15 |
rayjj | Adobe Reader also zooms with <ctrl>+ and <ctrl>- on Windows -- maybe some option key with + and - will do that on Mac ? | 16:50.35 |
henrys | fredross-perry: sorry I misread "can't" as "can" my bad. | 16:50.59 |
fredross-perry | np | 16:51.07 |
Robin_Watts | cmd+/- zoom on acrobat on macos. | 16:51.15 |
henrys | shift scroll scrolls horizontally | 16:51.17 |
fredross-perry | Safari/mac does not zoom with the wheel (it would seem) | 16:51.24 |
| nor does Chrome | 16:51.40 |
| Firefox does. | 16:52.06 |
rayjj | Chrome on Windows zooms with <ctrl>wheel | 16:52.10 |
| and on Windows <shift>wheel does horiz scrolling | 16:53.09 |
fredross-perry | I think we can do what we want here, as long as allversions of gsview do the same thing. | 16:53.58 |
Robin_Watts | We *should* support ctrl-wheel for zooming and shift-wheel for horiz scrolling, because on at least some of our platforms it's an expected thing. | 16:56.00 |
| and on other platforms it doesn't hurt. | 16:56.09 |
| rayjj: I downloaded gsview on the 21st. 25730619 bytes. | 16:57.29 |
rayjj | There's an outfit that claims to have cheap code signing keys $180/yr. https://www.comodo.com/landing/ssl-certificate/cheap-code-signing/?gclid=CLW7iIPvjMUCFciIfgodplwAXQ and includes Apple on the list of "Compatible with" as well as Win 7 and 8 | 16:57.31 |
| Robin_Watts: yep, that's what I downloaded. | 16:58.13 |
| now the next question is: Is that what I'm running ? (because the About is pretty vague) | 16:59.09 |
| I'm going to uninstall and reinstall | 16:59.25 |
Robin_Watts | http://codesigning.ksoftware.net/ | 16:59.25 |
rayjj | strange. The control panel "uninstall" had the install date as July 2014. And I know I installed it just this week (or tried to). | 17:01.26 |
| Robin_Watts: well, I did that (un/re install) and now the 'zoom control' option is gone (and it says the Ghostscript is 9.18) -- BTW the Copyright on the About page still says 2014 | 17:04.00 |
| it would be nice to make sure that we start incrementing "minor" version numbers from here on (it's currently stuck at 6.0.0.0) It's not like we have a shortage of digits to play with :-) | 17:07.44 |
henrys | yes fredross-perry | 17:07.51 |
fredross-perry | all righty. BTW, whatâs the version of the current GS? | 17:08.47 |
chrisl | 9.16 AGPL or 9.17 commercial release | 17:09.15 |
rayjj | Robin_Watts: that site specifically mentions Windows Driver signing and also says "As of late August 2013, all valid (not expired, not revoked) Comodo Code Signing Certificates can be used for Kernel-Mode Code Signing!!! (For Windows Vista and greater)" which is good news! | 17:09.53 |
chrisl | I think you'd want to use the 9.16 release | 17:09.53 |
| fredross-perry: ^^ | 17:09.58 |
Robin_Watts | rajj: Indeed. | 17:10.08 |
| Supposedly tucows are even cheaper... $75 a year | 17:10.18 |
rayjj | chrisl: yes, I mentioned that the other day -- IMHO, gsview should ship with a release of mupdf and gs (not just whatever was built in on the day it was built) | 17:10.49 |
chrisl | rayjj: that would be my preference, too..... | 17:11.13 |
rayjj | Robin_Watts: but the ksoftware site doesn't mention Apple compatibility, while the comodo site does. Confusing | 17:11.25 |
Robin_Watts | rayjj: ksoftware are comodo resellers, so it's the same thing regardless. | 17:11.42 |
rayjj | iirc, Robin_Watts felt that as long as it was built with whatever was tagged, that was OK, but I thought that was somewhat risky since it hadn't gone through the more thorough pre-release test cycle | 17:12.28 |
Robin_Watts | rayjj: Indeed. It's entirely possible we'll make tweaks to mupdf to suit gsview as gsview develops. | 17:13.16 |
| We are not doing new releases of mupdf every time that happens :) | 17:13.26 |
| If we're closed source, then it's less of an issue. | 17:13.51 |
| If we're open source and only releasing via the git repo, it's also less of an issue. | 17:14.21 |
rayjj | Robin_Watts: good point. Same for rebuilding an updated version with a patched gs | 17:14.38 |
chrisl | I think potentially getting different results from mupdf as mupdf, and gs as gs compared to using them as part of GSView is undesirable | 17:14.46 |
Robin_Watts | chrisl: So, you want to be releasing new ghostscripts every time it's fixed for gsview? | 17:15.24 |
rayjj | chrisl: the app is more than just a way to invoke gs or mupdf, and if it doesn't link outside of what we ship, then being able to release important patches as incremental updates would be nice (and more like what our customers do) | 17:16.27 |
chrisl | Well, I've said my piece...... | 17:17.05 |
rayjj | as you can tell, I'm waffling ;-) | 17:17.24 |
chrisl | I just wonder why we bother with releases at all, if even we don't bother to use them ourselves.... | 17:17.58 |
Robin_Watts | chrisl: Releases are what we give to other people and say "use this, and we'll support it" | 17:18.25 |
chrisl | And GSView is "other people" as far as mupdf and gs is concerned | 17:18.57 |
Robin_Watts | chrisl: Not really. | 17:19.31 |
henrys | chrisl: I don't know about this different release number business, obviously this table is wrong: http://www.ghostscript.com/download/gsdnld.html. the ghostscript site say 9.15, I thought this would get confusing. | 17:19.51 |
chrisl | henrys: what site says 9.15? | 17:20.28 |
henrys | ghostscript.com/Releases.html | 17:20.54 |
Robin_Watts | That just needs updating. | 17:21.16 |
chrisl | Erm, I fairly sure I updated that - I wonder if it got messed up yesterday...... | 17:21.34 |
Robin_Watts | It wasn't me. I didn't do it. You can't prove anything. | 17:21.48 |
henrys | check the logs I didn't even log into casper .... ;-) | 17:22.19 |
chrisl | No, I was faffing around trying to get the gs, mupdf and gsview sites more consistent | 17:22.47 |
fredross-perry | âgsview should ship with a release of mupdf and gsâ - agreed. | 17:22.52 |
chrisl | henrys: I think it might be nice to reduce the number of places where we "announce" the new releases on the website | 17:24.04 |
Robin_Watts | fredross-perry: So, next time you find a bug in mupdf that stops gsview behaving, you'll be happy to wait 6 months until you can ship a fix? | 17:24.25 |
henrys | chrisl: right there should be 1 place. | 17:24.45 |
chrisl | Well, how often are we planning to release gsview?? | 17:24.53 |
fredross-perry | No. So, I officially rescind the thing I just said. | 17:24.59 |
Robin_Watts | chrisl: At the moment? quite often. | 17:25.12 |
chrisl | Right, but once we get passed beta, I assume it will fall into the 6 monthly cycle with the other products | 17:25.38 |
Robin_Watts | chrisl: And at that point we can reevaluate the issue. | 17:26.01 |
fredross-perry | Thereâs a fairly robust automated testing regimen for gs. mupdf too? if so, then releasing gsview with well-tested gs and mupdf would be ok. | 17:26.11 |
Robin_Watts | fredross-perry: Yes. Both gs and mupdf are tested on the cluster the same way. | 17:26.29 |
henrys | chrisl: I guess the table is okay for this release. | 17:27.32 |
fredross-perry | so thatâs great. If âreleaseâ means âwe support itâ, then we decide if weâll support out-of-cycle versions of gs and mupdf for the purpose of releasing gsview out-of-cycle. And if we feel comfortable, then we do it. | 17:28.16 |
chrisl | I'm fine with pulling in "current" mupdf and gs releases whilst gsview is pre-release | 17:29.02 |
Robin_Watts | fredross-perry: exactly. | 17:29.09 |
chrisl | I mean current mupdf and gs code.... | 17:29.14 |
henrys | be back in a while | 17:41.30 |
Robin_Watts | rayjj: I've just made an account for author.tucows.com, and they do indeed sell certs. | 17:43.21 |
| $75 for one year, $195 for 3. | 17:43.33 |
rayjj | Robin_Watts: cool | 17:48.14 |
Robin_Watts | That's pretty much a reasonable price. | 17:48.41 |
| (I mean, ultimately it's a couple of prime numbers multiplied and modded, but...) | 17:49.02 |
rayjj | chrisl_away: I /think/ that QueryPerformanceCounter _is_ available on windows 95 (and derivatives) via the win32 thingy that we need anyway (we no longer support 16-bit segmented builds). Do you have a Windows 95 VM set up that you can try it, or do I need to do one ? | 17:50.11 |
| I _do_ still have Win 95 distribution CD's laying around | 17:50.48 |
| of course VMWare may laugh at me trying to install it :-) | 17:51.15 |
Robin_Watts | https://msdn.microsoft.com/en-us/library/windows/desktop/ms644904%28v=vs.85%29.aspx | 17:51.23 |
| Minimum supported client Windows 2000 Professional | 17:51.41 |
| Minimum supported server Windows 2000 Server | 17:51.55 |
| http://www.drdobbs.com/windows/win32-performance-measurement-options/184416651 | 17:52.33 |
| That suggests it is there on Windows 9x things | 17:53.31 |
rayjj | Robin_Watts: yes, I think that's what I've seen -- MS claims "minimum supported" levels, but it seems to be part of the win32 layer on Windows 95 | 17:55.15 |
| The DrDobbs seems to indicate that only WinCE doesn't have it, although there is the "(given hardware support)" caveat -- which can probably be ignored | 17:57.01 |
| chrisl_away: I think I'll commit the change "as-is" and later (maybe never) check if it works on Win 95, unless you know of someone that does still run Win 95/98/ME that can try the build | 17:58.37 |
Robin_Watts | rayjj: The Doctor Dobbs thing says that in practice they've never met a computer that doesn't have it in. | 17:59.20 |
fredross-perry | I thought WinME was illegal in some states | 18:06.50 |
rayjj | Robin_Watts: right, and the code I had checked for a 0 return from QueryPerformaceFrequency and used the old method if the hardware didn't support it. I think it's OK, and replied to chrisl_away. | 18:09.00 |
Robin_Watts | rayjj:That sounds ideal. | 18:09.19 |
rayjj | Robin_Watts: thanks for that DrDobbs link. That seems much more authoritative | 18:09.20 |
| and since it seems to be a standard part of win32, it shouldn't cause a build problem | 18:10.24 |
| at least I know the push won't cause a build problem on the cluster :-) | 18:12.19 |
mac- | hey, how can I quickly go to some page number in MoPDF ? | 18:22.56 |
| eem, I mean MuPDF | 18:23.24 |
rayjj | mac-: probably "instantly" since PDF's are random access, but how long will it take to render the bitmap so you see the page can vary | 18:23.58 |
mac- | no, I mean how can I type in page number to jump to it | 18:24.31 |
| :0 | 18:24.32 |
| :) | 18:24.33 |
rayjj | mac-: there is no need for mupdf to read through a bunch of other pages in order to jump to a page | 18:24.44 |
mac- | I can see that now after change page it is 'animated' | 18:26.49 |
| can I just turn it off ? | 18:26.54 |
| to see it quicker ? | 18:26.59 |
| can I in same way see on what page I am ? | 18:32.37 |
| usually pages written in the PDF document are not correct | 18:32.53 |
rayjj | mac-: #g takes you to a specific page # | 18:33.12 |
| mac-: the page numbers showing on the page do not have any relation to which page you are on in the PDF, as you realize. There is supposed to be a way for a PDF creator to give the relation, but I don't think anyone bothers | 18:39.25 |
| things like title pages, table of contents, etc. all happen in "sections" before the document "logical" page 1 (the one that the document creator wanted to be labeled as "1" instead of i, ii, or nothing) | 18:40.42 |
| chrisl_away: (for the logs) I committed the change. Since then I've found that the linux 'times' is guaranteed to be 1 microsecond or better. There is a clock_gettime that is supposed to be similar to QueryPerformanceCounter, but I'm not going to change *n*x builds for now. | 18:42.37 |
mac- | yeah I know but if there is any way to get know on which page actually am I ? | 18:42.55 |
rayjj | chrisl_away: mainly I wanted the higher resolution timers to give us something to measure on Windows. Linux profile tools actually work | 18:43.40 |
| mac-: not in mupdf. If you wait until we do a beta release of our 'gsview' app (that uses mupdf) it has the page number displayed | 18:44.21 |
mac- | uhm | 18:44.45 |
| sure | 18:44.46 |
rayjj | mac-: or use the free Adobe reader product (some of which use mupdf internally, I think) | 18:45.20 |
mac- | I have 64bit Linux | 18:45.35 |
| no Adobe for it | 18:45.40 |
| this company simply sucks | 18:45.49 |
rayjj | mac-: really ? I thought it was available | 18:45.56 |
mac- | and their products are weak tbh | 18:45.57 |
| heavy and not reliable | 18:46.09 |
rayjj | but I agree that Adobe sucks (probably why they use mupdf to reduce the bloat) | 18:46.33 |
| but then, I shouldn't talk badly about a customer | 18:47.13 |
mac- | :> | 18:47.19 |
rayjj | if only they used our product more, maybe they'd have a better product | 18:47.34 |
mac- | then you're saying that Adobe PDF Reader for Linux is based on MuPDF ? | 18:48.07 |
rayjj | mac-: but we are getting very close to a beta release of our 'gsview' that uses mupdf for PDF, TIFF, XPS, CBZ, PNG, JPG viewing and relies on Ghostscript for PS and EPS | 18:48.40 |
| mac-: I don't know which Adobe Readers use mupdf, but I m pretty sure it is not on linux | 18:49.17 |
mac- | how far you are from release ? | 18:49.17 |
| I see that last release was in 2012 | 18:50.42 |
| ? | 18:50.44 |
rayjj | mac-: dotting the i's and crossing the t's on license issues internally. Technically, we are at 'beta' now in that there are improvements to be made, but it looks pretty nice and is cross platform. mvrhel_laptop and fredross-perry are the main developers for it | 18:50.44 |
mac- | uhm | 18:51.05 |
rayjj | mac-: that was an old ghostscript based version that Ghostgum developed, but it had good name recognition, so Artifex bought the name and is doing a modern version | 18:51.29 |
mac- | I didn't knnow | 18:52.05 |
rayjj | gsview < 6 is Ghostgum's -- GSview 6+ is Artifex's | 18:52.17 |
mac- | it will be commercial only or GNU ? | 18:53.55 |
rayjj | 6.0 beta is what is coming, and uses the newer technology to get better interactivity (via mupdf) and better printing (using GS) and let's us use the best of the two (or more) Artifex technologies to improve the user interaction | 18:54.00 |
mac- | what is licensing model ? | 18:54.01 |
rayjj | mac-: that's what is being discussed. It (most likely) will *NOT* be GPL, but since we own mupdf and gs, it doesn't have to be. There will definitely be a free version, but it may not be open source. There probably will also be "pro" version features for $$$ | 18:55.49 |
| mac-: there's nothing stopping anyone from doing an AGPL viewer based on mupdf and/or ghostscript, but since we own both, we can do a proproetary (non GPL) version. Even if we open source it, the viewer layer can be licensed differently to the GPL (the old gsview was AFPL -- Aladdin Free Public License) | 18:58.38 |
| mac-: the Aladdin license prohibited selling it, but still required open source, virally, as the GPL does | 18:59.45 |
mac- | uhm | 19:00.12 |
rayjj | mac-: "uhm" is not useful -- just takes up a line on my IRC log | 19:00.48 |
mac- | well, if it will be free of charge for private use its okay for me | 19:04.09 |
| those who will need much more probably won't have any problem to pay for pro version | 19:04.35 |
rayjj | mac-: in any case, since mupdf (platform/x11/pdfapp.c) is open source, you are welcome to modify it to show the page #, either always or in response to some key. If you do, feel free to send a patch to us and we might include it (if you give us the rights to the patch) | 19:05.25 |
mac- | well, it is not my area but as long as it is written in C (and it is as I see from file extension) I can do sth to improve the programme when I fight now with C and will go further in system programming | 19:06.40 |
rayjj | mac-: that's how many of the features in mupdf got there -- we intended it to be a 'demo' app, but it has sort of grown with lots of screwy options (like rotations at other than 90 degreees :-/ ) | 19:06.43 |
mac- | I use Linux and now going into OpenBSD | 19:07.07 |
| yeah I can see some interesting features :p | 19:07.35 |
| the 90 deg rotating was one of surprising :p | 19:07.50 |
rayjj | mac-: I started on OS's you would not even recognize, but then unix (mostly Sun) and finally got dragged into using Windows and some Windows development. Thank goodness for tools like msys (msysgit) that make my Windows system more like a real system (unix like) | 19:08.53 |
mac- | :> | 19:09.12 |
rayjj | mac-: BTW, I don't know if it is intentional, but the http://gsview.com/downloads.html site _does_ have a linux x86 and x86_64 version available. | 19:12.18 |
| mac-: use at your own risk. Feedback (here or via email) appreciated | 19:12.49 |
| mac-: that's a "pre-beta" version, and no source | 19:13.41 |
mac- | understand | 19:13.57 |
| will install on sunday when I do sth else now | 19:14.08 |
rayjj | mac-: I don't know how up to date it is (fredross-perry would) but I am trying it | 19:15.23 |
| (the x86_64 version) | 19:15.35 |
fredross-perry | that version was built on April 16. So very recent. | 19:21.44 |
rayjj | hmm.. well, it wouldn't install on peeves (missing some library) /lib/libc.so.6: version GLIBC_2.11 not found (required by GSView-6.0-Installer) | 19:24.09 |
| fredross-perry: trying on a more recent linux (peeved) | 19:24.37 |
| fredross-perry: well, it worked on peeved (#42-precise1-Ubuntu Aug 14, 2013) | 19:27.32 |
fredross-perry | all right. The installer package is not one that Iâve built. | 19:28.09 |
rayjj | fredross-perry: peeves is running an ancient '#75-Ubuntu Mar 2011' | 19:28.44 |
| fredross-perry: who's doing the installer packages ? | 19:29.11 |
fredross-perry | Itâs from the Qt folks (I think). It consumes our files and some config data and produces an executable. | 19:30.40 |
| We could be better off with a commercial install-making program, but $$ were somehing of an issue when last we looked. | 19:31.34 |
| or making packages that play nice with rpm and the like. | 19:34.03 |
| stepping out for a bit | 19:34.48 |
rayjj | fredross-perry: it completed the install, but I can't find it anywhere | 19:34.55 |
fredross-perry | there should be an icon on your desktop. Otherwise, itâs under âgsviewâ in your home directory | 19:35.28 |
| can you Skype me from the device? So I can see your screen? | 19:36.10 |
rayjj | fredross-perry: oh, I think I found it. It put it in my /home/ray/gsview directory | 19:37.02 |
fredross-perry | ok. launch âgsviewâ, not âgsview.exeâ btw. | 19:37.29 |
rayjj | strangely, there's a gsview.exe in it and a MASSIVE "uninstall" file | 19:37.59 |
henrys | maybe I'm being to fussy but I think calling 9.17, 9.16 here http://www.ghostscript.com/download/gsdnld.html is bad, can we at least hack the web page if not update the releases? | 19:38.10 |
| for the logs chrisl_away | 19:38.27 |
fredross-perry | I donât know why the uninstall is so huge. Again, thatâs built by the install system I am using. | 19:39.02 |
| gsview is a script to insure that gsview.exe is launched from the gsview folder. | 19:39.53 |
rayjj | of course, it crashes immediately when trying to display an ATS file. But at least it's able to bring up annots.pdf | 19:41.04 |
fredross-perry | please send me he file itâs crashing on. Thanks. | 19:43.44 |
rayjj | the display is somewhat disappointing -- the area on the window outside the page (in fit page or zoomed out mode -- anything but fit width) is just white like the page. I expect 'gray' or a border or something | 19:44.52 |
| fredross-perry: If you can get any or the tests_private/pdf/PDF_1.7_ATS/*pdf file to open, let me know. I tried 5 | 19:45.40 |
fredross-perry | it actually is gray. Maybe just not dark enough. | 19:45.40 |
| internal path please? | 19:46.36 |
rayjj | strange. Now it's not crashing. | 19:49.16 |
| fredross-perry: what do you mean by "internal path" ? | 19:49.45 |
| the page scaling on /home/marcos/cluster/tests_private/pdf/JEITA/J10_acrobat.pdf is quite "funky". Also, somewhat disconcerting. I single click on the file in the "file Open" dialog and before I hit "Open" it opens and the dialog closes | 19:55.57 |
| on J10 p 2, I hit "fit width" and there is a LOT of white space to the right of the page image | 19:57.32 |
| nm. That appears to be how p2 is defined | 20:00.42 |
fredross-perry | I am not sure how to access tests_private or marcos/cluster. If you experience sustained issues with any files please send me access instructions, thanks. | 20:03.30 |
rayjj | fredross-perry: do you have ssh into any of the cluster machines ? | 20:04.10 |
| (or casper) ? | 20:04.16 |
fredross-perry | II think so. | 20:05.59 |
henrys | rayjj, fredross-perry there is a test file fetcher on the dashboard | 20:06.21 |
rayjj | henrys: oh, thanks. | 20:06.37 |
| I forgot about that since I have most local, and I use pscp for the others | 20:07.01 |
henrys | at the prompt enter tests_private/pdf/JEITA/J10_acrobat.pdf | 20:07.14 |
fredross-perry | go it. | 20:07.49 |
| got it | 20:07.54 |
henrys | I got carded getting compressed air to clean my computers... apparently kids are snorting that now... geez sounds like a really bad idea. | 20:08.31 |
rayjj | henrys: a 'browse' option would *really* be nice. getting all the CaPs right is a pain | 20:08.55 |
| snorting air???? and why would they card you ???? | 20:09.28 |
fredross-perry | browse = +1 | 20:09.59 |
henrys | rayjj: they inhale this stuff and makes them high, crazy whatever... I had no idea. | 20:10.06 |
rayjj | or is it really compressed Nitrogen ? You know that they are using Nitrogen to execute people in Oklahoma (presumably through anoxia) | 20:10.48 |
fredross-perry | can you nme that ATS file? | 20:11.15 |
| *name* | 20:11.20 |
rayjj | if the compressed stuff somehow doesn't have enough O2, then it's the same as several other ill advised (deadly) methods of getting "high" (dizzy due to anoxia) | 20:12.02 |
henrys | the lady at home depot said some kid took some in the bathroom, inhaled and went bat shit crazy, they had to call the cops. | 20:13.35 |
| the can doesn't say what gas is used. | 20:16.25 |
rayjj | fredross-perry: i can't reproduce the crash now. Strange. I'll try later. | 20:16.53 |
| henrys: if it's nitrogen, but not very pure, (also used to fill tires) then it can form nitrous-oxides (laughing gas) | 20:18.07 |
| but I wouldn't be suprised if it just didn't contain O2 (so it couldn't accelerate fires -- a "safety" measure) and anoxia is the culprit | 20:19.15 |
henrys | I can't imagine inhaling it, it's extremely cold coming out of the can | 20:20.11 |
astrodog | Usually, the canned air stuff is various volatile organics and hydrocarbons. | 20:20.16 |
rayjj | fredross-perry: what do you mean: "browse = +1" | 20:20.21 |
astrodog | It's... sort of an intoxicant. | 20:20.27 |
| That or refrigerant, which isn't. | 20:20.36 |
fredross-perry | âa 'browse' option would *really* be nice. getting all the CaPs right is a painâ | 20:20.49 |
| upvoting that | 20:20.54 |
henrys | fredross-perry: Robin_Watts maintains it... | 20:21.01 |
rayjj | oh, I would have said: browse++ | 20:21.09 |
fredross-perry | sounds like a compiler variant | 20:21.28 |
rayjj | or browse += 1 ;-) | 20:21.29 |
henrys | astrodog: interesting ... | 20:21.41 |
astrodog | henries: It's pretty close to what you'd get inhaling a mix of boiling propane and R134a. | 20:21.43 |
rayjj | *YUCK* | 20:21.58 |
fredross-perry | Iâll hve what astrodog is having. | 20:22.07 |
astrodog | Ha. | 20:22.11 |
rayjj | I bet you have a HELL of a hangover | 20:22.30 |
astrodog | You can use some brands of canned air to test explosive gas sensors. | 20:22.32 |
| rayjj: I would imagine so. The organics will bind along the same lines as ethanol. No clue what the refrigerant does. | 20:23.03 |
rayjj | and here I thought having O2 in compressed air might be a safety risk | 20:23.04 |
astrodog | ray: The reason they make it that way is pressure. Light hydrocarbons have nice low vapor pressures, so you don't need a really thick can to contain them, and it'll recharge itself by boiling them. | 20:23.50 |
henrys | I'm just going to clean my computer with it ;-) | 20:24.04 |
rayjj | henrys: yeah, right | 20:24.14 |
astrodog | If you used straight nitrogen, you'd need the can to withstand a few thousand PSI to have any real quantity of gas in there. | 20:24.16 |
astrodog | throws the "The More You Know" rainbow across the channel, and goes back to lurking. | 20:24.38 |
| :P | 20:24.39 |
rayjj | astrodog: so that's why it creates so much "frost" when I spray my KB. The stuff is evaporating/boiling and cooling off more than just air would | 20:25.28 |
| I wonder how they get by with calling it "compressed air" | 20:26.12 |
| it sounds as bad as the contact cleaner I used to use when repairing TV tuners (back when they had switches and capacitors and stuff -- not just Phase locked loops) | 20:27.33 |
| time for lunch. bbiaw | 20:28.56 |
mvrhel_laptop | ok I will add ctrl +/- for zoom | 23:06.15 |
Robin_Watts | fredross-perry, rayjj: Most of the time you want a file, it's because it's gone wrong in a cluster test. Just click the link in the report to download it. | 23:09.29 |
| Adding a browse option - will ponder that. | 23:09.52 |
fredross-perry | there are probably times when you want to go get some files that you remember are interesting in some way, but you donât exactly remember their names. | 23:10.59 |
mvrhel_laptop | fredross-perry: you still there? | 23:13.42 |
fredross-perry | yesâm | 23:13.49 |
mvrhel_laptop | I saw ray's comment about white around the edges | 23:14.07 |
| I had the issue on linux too | 23:14.11 |
| my son had the monitor contrast maxed out | 23:14.19 |
| once I fixed that it was fine | 23:14.27 |
| ;) | 23:14.29 |
| but you might want to darken it a bit | 23:14.45 |
| I wonder how ray had a zoom control on the bottm | 23:14.54 |
| he must have had an old version | 23:15.00 |
| Robin_Watts: you don't have a zoom control on the bottom to you? | 23:15.11 |
| s/to/do/ | 23:15.17 |
| It is interesting to see how every one uses different controls do to the same operations (e.g. zooming) | 23:16.05 |
| I had always just clicked the zoom control on the ui which is why that was in there first. then Robin_Watts pointed out the need for the ctrl mouse wheel so I added that, now ray wants ctrl +/-. Anything else? | 23:17.21 |
Robin_Watts | mvrhel_laptop: I do not have such a control. | 23:23.02 |
mvrhel_laptop | ok good :) | 23:23.13 |
| Forward 1 day (to 2015/04/24)>>> | |