IRC Logs

Log of #ghostscript at irc.freenode.net.

Search:
 <<<Back 1 day (to 2014/09/03)2014/09/04 
rayjj i recall a snake oil salesman pitching a "super image compression" algorithm to my previous company. They were upset when I suggested that compressing a 640x480 256 color image to 16 bits could only be based on a hash of "canned" images looking up images.00:00.24 
rayjj wonders if they got any suckers on that one00:00.44 
kens Yet more setpagedevice insanity :-(08:03.55 
chrisl Is this a new thing, or a regression?08:08.00 
kens Its a pre-existing 'problem'08:08.12 
  Not a regression of any kind.08:08.28 
  If you set LockDistillerParams, then gsave, tehn setpagedevice, then grestore, it fails08:08.51 
  Because installpagedevice insists that any keys which asre not processed in the dictionary is an error08:09.11 
  And if LockDistillerParams is set we don't process the distillerparams, so they aren't read08:09.32 
  Hence, error....08:09.45 
chrisl So, it should "just" be a case of processing the distillerparams, but discarding anything they do?08:10.09 
kens The only solution is to have pdfwrite process all the distillerparams, and throw them away if ... Yes, that's what I was saying08:10.19 
  But of course, distillerparams are *very* complex, and all the code at the moment writes the data directly into the device structure08:10.46 
chrisl With the usual caveat of all hell breaking loose because it's pdfwrite.......08:10.59 
kens Its more a case of having to write loads and loads of code08:11.14 
  Instead of each parameter writing into the device structure, I'll have to write it to an intermediate variable, then if LockDistillerParams isn't true, copy that into the device structure08:12.02 
  WHich will be slower than it is now08:12.18 
  (perhaps I should say *even* slower....)08:12.28 
chrisl Couldn't you just swap between the real device (normal) and a dummy device structure (LockDistillerParams)?08:13.23 
kens Its a pretty horrendous structure we're talking about.....08:13.47 
chrisl That way you wouldn't have to copy stuff around08:13.55 
kens Have you looked at the pdfwrite device structure recently ? :-(08:14.26 
chrisl So?08:15.30 
  It's a bit late to be worried about memory efficiency in pdfwrite!!08:15.56 
kens I don't fancy trying to duplicate it.08:15.59 
  Reaidng the values to intermediate variables looks like a better bet to me.08:16.27 
  One or two are already done that way too (CompatibilityLevel for example)08:16.48 
chrisl Seems a shame to slow down normal running just to handle a relatively rare situation08:17.11 
kens Its more just the whole nonsense of setpagedevice which annoys me. Our implementation is downright broken08:17.13 
chrisl Well, you know my thoughts on that.....08:17.37 
kens I *could* fix it by having .installpagedevice not worry about unprocessed keys08:17.38 
chrisl That would be a major change in behaviour, which would worry me a bit08:18.41 
kens Well that's what worries me about it too.08:21.40 
  Though I've never seen it happen in practice08:21.57 
  Grr Bugzilla not talking to me again.... Not having a good day here08:23.42 
jogux kens: it's not just you, bugzilla seems pretty unhappy :(08:40.32 
kens TBH that makes me happier.....08:40.44 
  I did manage to pull some stuff off it earlier, but now i can't connect to it at all08:41.13 
jogux ditto08:41.24 
  traceroute gets to somewhere in comcast08:41.45 
kens could be a DDOD or the 512k routes problem, or TWC, or well 'something'08:42.24 
jogux :-)08:42.56 
chrisl Is bugzilla still hosted on casper?08:43.22 
jogux doesn't seem to be08:43.30 
chrisl I have a vague memory of marcosw intending to move it elsewhere, but wasn't sure08:44.44 
kens Well, I guess I'll ignore the setpagedevice for now and go back to forms. Coffee first though08:45.29 
tor8 Robin_Watts: re your latest patch, does that mean we never store 1-bit images compressed?09:07.16 
kens chrisl when are you thinking of doing a release candidate ? I have a good grip on the forms problem and I'd like to get the fix in, but it may take me the rest of the day.09:09.08 
chrisl kens: Next Mondat09:09.32 
  Monday09:09.36 
kens Good enough, I should have it done by then, thanks09:09.46 
chrisl So, for the first time in *years* I've ended up buying spinning disc HDD that wasn't IBM/Hitachi/HGST - now I have two brand new Seagate drives on my desk, clicking away merrily. This is why I kept buying HGST drives :-(09:14.43 
kens Why did you buy Seagate ?09:15.20 
tor8 aren't seagate infamous for being noisy?09:15.38 
chrisl This isn't noisy, this is proper, pending disaster clicking......09:16.04 
tor8 eep!09:16.21 
chrisl The only two options seemed to be WD and Seagate, and the only drives I've had more problems with than Seagate are WD!!09:16.41 
tor8 I haven't bought a spinning platter device in years... I'm a bit out of touch. even my latest external USB drive has an SSD in it...09:16.54 
Robin_Watts chrisl: Samsung make drives too.09:16.58 
chrisl tor8: this is for storing videos and such - I didn't really want to spring for 2TB of SSD09:17.46 
  Robin_Watts: I looked at Samsung, but couldn't find the capacity I wanted09:18.10 
tor8 chrisl: yeah, my external SSD is more for quick transfers between computers with crappy operating systems that can't run proper ssh software...09:18.19 
  chrisl: I do have a set of 2TB disks for video and backups, but they're pretty old09:18.50 
Robin_Watts http://www.ebuyer.com/656732-samsung-4tb-d3-station-external-hard-drive-hx-d401tdb-g?utm_source=google&utm_medium=products&gclid=Cj0KEQjwm6CgBRC0zOmrydrqmosBEiQA_xoLRv6LErdgTa4aDvWyexA96Pa9VcY7cxG1CVhTfeYrIboaAi_Z8P8HAQ09:18.53 
kens Hmm, one of those would be nice for my backups09:19.29 
chrisl wanted internal....09:19.33 
Robin_Watts I bought one of these, and shelled it to get an internal: http://www.ebuyer.com/339426-hgst-4tb-touro-dx3-desktop-hard-drive-0s0340009:20.13 
  Though when I bought it it was 109 quid or something.09:20.22 
chrisl There seems a lot more choice in 2.5" drives just now, too......09:21.33 
jogux I like to use https://www.amazon.co.uk/Samsung-Slimline-Portable-Hard-Drive/dp/B008PABFX8/ref=as_sl_pc_ss_til?tag=h9l-21&linkCode=w01&linkId=&creativeASIN=B008PABFX8 for backups10:12.49 
  I have a couple, ones always attached to the machine, and every few weeks I switch the one on the machine for one from the shelf10:16.19 
Mario__ Hello Team, is there any update regarding this bug? http://bugs.ghostscript.com/show_bug.cgi?id=695459 It is quite a problem for our us11:56.02 
Robin_Watts Mario__: Are you a customer?11:56.42 
Mario__ Not yet11:56.56 
  we are looking in to it11:57.01 
Robin_Watts ray is looking into it. I believe he's found what is going wrong.11:57.12 
  and he's figuring out how to fix it now.11:57.30 
  (we trip over a problem when reading from the clist, but the problem actually occurs because the write to the clist went wrong)11:57.48 
  Ray is in california, so won't be here for a few hours.11:58.10 
Mario__ We have ancountered this issue with several files till now11:58.28 
  this could lead to losing a future client, as their files don't work with our demo12:01.41 
  I will try to contact Ray later on12:03.20 
  But we have an other issue: http://bugs.ghostscript.com/show_bug.cgi?id=69546412:03.34 
  which is even worse, as it isn't producing an error message12:03.51 
kens Mario__ : free user bugs get fixed 'as and when' we have time.12:04.23 
Mario__ Yes, I know, and therefore we are looking into a commercial license12:05.24 
Robin_Watts And that bug is a whopping 2 days old now. You need to give us some time.12:05.44 
Mario__ I understand. But also I wanted to ask, did you had similar issues in the past?12:07.55 
kens Well you can search our bug database12:08.17 
Robin_Watts we've had problems with the clist in the past, but then we've fixed all those problems.12:15.55 
  Each time we hit such a problem, we add the file to our test suite.12:16.09 
  Every time we commit a change to gs, it is checked with a combination of over 75000 files/resolutions/devices etc.12:17.06 
  so we can be fairly sure we don't regress.12:17.19 
  The tiff problem looks to be related to overprinting. That would probably need a different engineer to look into it. We'll check in with him later.12:18.25 
Mario__ Thank you12:34.23 
chrisl Someone else who wants us to fix bugs by telepathy.....12:50.38 
kens Yepo :-(12:50.54 
leena Can anyone suggest how to read "layers" from an EPS created in Adobe Illustrator using Ghostscript?12:59.00 
chrisl leena: such layers are not part of the Postscript spec, thus Ghostscript does not handle them13:01.01 
leena chrisl: is there any way we can simulate with Ghostscript?13:04.17 
chrisl leena: Not that I'm aware of, and as they are not part of the spec, but proprietary to Adobe, there'd no guarantee of what we did being being correct13:05.53 
leena ok. 13:06.23 
kens As far as I remember, the layters are 'implemented' in Illustrator EPS files as comments, so you could extract them with comment processing13:06.37 
  But as Chris rightlys says, there is no specification, so you're on your own figuring out what the comments mean13:07.25 
richud hi, i have cross compiled ghostscript, I can run gs --help and get the help, but doing anything else gives no output at all13:20.23 
  if i compile with make debug I get one line of output with -DDEBUG, but nothing with any combination of -Z13:21.14 
chrisl Sounds like there's something wrong with your cross-compilation.....13:22.07 
richud anyone got any ideas? I am using 9.07 (newest one with cups support) cross compiled with uclibc (i386) 13:22.13 
kens I'd be inclined to agree. WHt's the line of output from -DDEBUG13:22.23 
  richud : the current version is 9.14, and 9.15 is due any time soon. All have the same level of CUPS support13:22.44 
chrisl Did you check the exit code?13:22.54 
richud START 0 1431892 136812 1309856 27892 true 540 3 <0>13:22.59 
kens WHat's your command line ?13:23.16 
richud gs -DDEBUG13:24.29 
  well it does the same except last 540 is then 53813:24.47 
kens Well, sounds broken to me, time to break out the debugger I'd say13:24.51 
chrisl richud: oh, and what platforms are you compiling on, and for?13:25.04 
richud if I run it through strace it all looks ok13:25.38 
  ken: I need gsraster , which was removed after 9.07 , sorrty13:26.03 
kens Well you aren't getting a GS> promtp. so its clearly not OK13:26.11 
richud chris: ubuntu 14.04 x64 to uclibc i386 target13:26.21 
  (its a terminal session)13:26.35 
chrisl Well, it should be very easy, then13:26.57 
richud yes, I am stumped13:27.16 
chrisl As kens says, you'll need to debug it13:27.34 
richud have to go, back later, ill read back if anyone has any ideas!13:27.39 
  thanks for your responses13:27.51 
kens Good grief!13:45.46 
  http://www.reuters.com/article/2014/09/04/us-usa-guncontrol-idaho-idUSKBN0GY2E62014090413:45.46 
chrisl Well, you know what the problem there was? His foot didn't have a gun......13:47.39 
kens If he was teching computing some of my .sigs might have come true.....13:48.05 
  Actually, I suspect the problem was the students didn't have guns.13:49.03 
chrisl No, I still think the foot needed to have some kind deterrent available, then it would never have happened13:50.12 
kens I was amused to see that they don't allow firearms where nuclear research is going on. Cos, you know, nuclear stuff is so much more vulnerable to bullets than say glass jars of chemicals13:50.23 
  I was thinking that if ther students had guns then i wouldn't happen again :-)13:50.46 
chrisl It makes you wonder why guns don't come with some kind of "catch", you know, for "safety"......13:52.21 
kens I'm sure the NRA would regard that as a bad idea13:52.40 
chrisl Well, that's just triggered the surveillance bots......13:53.18 
kens I'm sure we were under NSA surveillance already, as pinko commie socialist non-Americans13:53.59 
chrisl If not them, then GCHQ.....13:54.41 
kens I htink GCHQ just watches everybody13:54.54 
scinfu hi14:51.13 
ghostbot Welcome to #ghostscript, the channel for Ghostscript and MuPDF. If you have a question, please ask it, don't ask to ask it. Do be prepared to wait for a reply as devs will check the logs and reply when they come on line.14:51.13 
scinfu DO you think to create a Cocoapods project for MuPDF iOS ?14:53.11 
rayjj scinfu: I think paulgardiner had a recent discussion on that.14:53.51 
paulgardiner That may have been jogux 14:55.08 
rayjj paulgardiner: it was, sorry. scinfu please see http://ghostscript.com/irclogs/2014/08/25.html14:55.28 
  scinfu: some discussion toward the top14:55.53 
  scinfu: BTW, the "search" box on the irclogs is quite handy (upper right)14:56.46 
jogux scinfu: I think the basis conclusion was it was probably a good idea to create a cocoapod, but unfortunately everyone is busy on other projects and hasn't had time to look at it just yet.15:02.18 
scinfu rayjj: jogux: thank you 15:03.16 
  I will try to do it alone but I doubt that I will have a good outcome15:04.03 
Robin_Watts scinfu: If you want to try, we'll happily answer any questions you may have.15:04.06 
  And we'd be interested in seeing what you come up with.15:04.17 
  The MuPDF source code has no complex requirements.15:04.35 
scinfu I have already created a library with MUPDF, I will try to replicate it with pods15:05.46 
Robin_Watts rayjj: Previous timing was 14 minutes.15:10.16 
richud re15:13.41 
Robin_Watts hello richud15:14.02 
richud btw i was compiling with make -j4 CCAUX=gcc (after patching to fix the endian issue)15:14.15 
chrisl What endian issue?15:14.49 
richud in 9.06/9.07 i take it there arent any other problems I am missing? (it compiled fine, everyhting that should be 32bit looks right with file)15:14.50 
  9.06/7 fail cross compiling as the endian isnt detected15:15.06 
chrisl But both your host and target are little endian.....15:15.41 
richud im using those versions as they were last ones to build gsraster for a cups filter (which is what i am ultimately trying to get working)15:15.46 
  i know, it fails if it detects you are cross compiling15:16.11 
Robin_Watts richud: What are your host/target ?15:16.29 
kens See earlier in the logs15:16.39 
Robin_Watts I have cross compiled on x86 targetted at arm before now. both little endian.15:16.56 
  From x86 to arm register sizes are the same too, so I just configure on one, and change the compiler.15:17.40 
  From x64 to x86, register sizes potentially change, so that's more to configure.15:18.03 
richud the configure piece to patch is15:18.40 
  if test "$cross_compiling" = yes; then : { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error $? "cannot run test program while cross compiling See \`config.log' for more details" "$LINENO" 5; }15:18.43 
  anyway it got fixed in newer versions, but I am using old for gsraster15:19.13 
  can anyone hazard a guess as to why it would work enough to display the help fine but the actual program have no output?15:20.26 
  strace doesnt show it trying to open anything15:20.41 
chrisl There are too many possibilities to even start to guess.....15:21.03 
Robin_Watts richud: No. As people have said, use a debugger and step through.15:21.04 
  chrisl: Is there a configure option he can use to get cups support in later versions?15:21.30 
  --with_local_cups=yes or something?15:21.45 
chrisl We have cups support15:21.52 
  What did gsraster do, anyway? I forget.....15:22.17 
Robin_Watts That would be a linux question, so I refer you to yourself :)15:22.35 
kens Oh no, recursion!15:22.54 
chrisl Actually, that's a cups question, and thus I deny all knowledge.... recursion prevented15:23.16 
  I'm just wondering if a simple shell script would work around the fact that gsraster went away15:24.01 
richud it is very odd15:24.29 
kens Without knowing what it did, its hard to tell. I can't find nay reference ot it in the HTML history15:24.31 
  Or a trivial search in Git15:24.54 
richud at the moment i have disabled cups just to try and make it work in tiself15:25.00 
chrisl it's gstoraster.....15:25.12 
  and I *think* it's just a wrapper around gs that converts PS/PDF to cups raster15:25.20 
richud ill stick my compile blurb in a wiki page if you care to look, one moment15:25.35 
kens Ah, ,well that makes more sense, I might be able to find ot now....15:25.37 
  Right, it just outputs a CUPS raster from GS15:26.06 
richud http://www.richud.com/wiki/Ghostscript15:26.44 
  ive tried lots of other variates to no avail15:27.02 
Robin_Watts tor8: For the log. Updated version. I fixed a bug that meant we were looking for alphas, and I made it check that we have 8 bit streams.15:27.56 
  It now takes 70 seconds.15:28.23 
  Also, I made our jpeg handling avoid printing "Application didn't transfer enough scanlines" every time we bale out early.15:28.47 
richud ken/chris would you kindly cast your eyes on it, its from my makefile, anything obviosuly wrong?15:29.07 
chrisl richud: Nothing obviously wrong, no. Like as not, if something was obviously wrong, you wouldn't get as far as building it15:30.06 
kens As far as I can see we haven't removed gstoraster, its been moved to cups-filters. So, nothing to do with us at all. I don't believe that using gs 9.07 is any help at all in this, it shoudl work or not just as well with 9.1415:30.12 
richud chris: thanks for looking, yes it is very odd15:30.59 
chrisl kens: If the cups version is contemporaneous with 9.07 then it won't have gstoraster in it15:31.06 
richud ken:yes, there are ton of dependaencies to build for cups-filters, hence using old ghostscript with it in15:31.26 
chrisl richud: did you check the exit code? I can't remember if you answered that.....15:31.53 
kens Like I said, not our problem. Building an old version of GS doesn't seem like a good idea, but its up to you15:32.02 
kens still says: start debugging15:32.46 
richud chris: yes 015:33.02 
  ken: yes :(15:33.13 
chrisl Hmmm, so we don't even realise there's been an error - hmmmm15:33.34 
richud yeah , very strange :)15:35.03 
  i spent 2 weeks messing with cups to begin with before i relaised it was gs problem , spent a week fiddling with it now15:35.35 
kens I'd hardly classify it as a GS problem.15:35.52 
richud ok, well related to gs15:36.05 
  as in not cups15:36.16 
kens THe CPS people changed what thjey do and how they distribute it. Not really anything to do with us15:36.36 
richud anyway ill try latest version of GS tomorrow and see if that at least runs15:36.54 
  thanks for everyones thoughts/ideas/help15:37.02 
chrisl Well, I cross compiled 9.14 for i386 with uclibc a couple of months ago, and I got it working without too much pain15:37.09 
richud chris: ok ill come back with results tomrrow15:37.52 
  l8r15:37.53 
chrisl I wonder how often he's going to ask before realising we can't help..........15:38.57 
kens Night all16:18.21 
pedro_mac g’nite kens16:20.04 
chrisl Well, I have a working x86 gs 9.07, statically linked with uclibc, built on my x86_64 machine, so it is possible.......16:20.05 
  But I can't build gstoraster because I ain't trying to cross compile cups!!16:20.38 
  Robin_Watts: did you do an upgrade for the Chicago flight(s)?16:22.14 
Robin_Watts I upgraded coming home, I think.16:24.01 
chrisl Did you keep a note of the cost of the economy ticket before the upgrade?16:24.41 
Robin_Watts no, but I bet paulgardiner has it.16:26.34 
paulgardiner 764 possibly16:28.15 
jogux hm. did I forget to ask what class of flight artifex pays for? :)16:28.37 
Robin_Watts economy.16:29.03 
jogux ah, right, good.16:29.14 
  though I did get the extra legroom option I think :)16:29.26 
  are you guys actually paying for your own upgrades then? or is this some cash+points deal?16:30.04 
Robin_Watts upgrading using miles.16:30.51 
  That's "airmiles", not "Miles" :)16:31.02 
chrisl I think I did miles + cash....16:31.02 
jogux ah, cunning. and on the way back only so you don't get humped for UK business flight tax :)16:31.16 
Robin_Watts on the way back cos it's overnight, and it was cheaper :)16:32.59 
jogux :)16:33.05 
Robin_Watts It was 825.41+miles, so yes, looks like cash+miles.16:35.03 
chrisl paulgardiner: thanks, that close enough to what I have here, I'm fairly sure I won't be conning Miles16:35.37 
t4nk856 hi18:27.05 
ghostbot Welcome to #ghostscript, the channel for Ghostscript and MuPDF. If you have a question, please ask it, don't ask to ask it. Do be prepared to wait for a reply as devs will check the logs and reply when they come on line.18:27.05 
t4nk856 I tried MuPDF for Android, it is really good. But, for our requirement, ReaderView should be vertical ( as in Acrobat ), I tried to make it Vertical. It is working but not working perfectly. Could you please share if you have already done a vertical ReaderView. If yes, please send it to kparunjyothis@gmail.com. I have sent an email regarding this to android@ghostscript.com18:31.41 
Robin_Watts t4nk856: Hi. Did you email me earlier about this?18:32.01 
t4nk856 Yes 18:32.09 
Robin_Watts And I replied...18:32.12 
t4nk856 Yesterday night18:32.13 
  got your reply18:32.19 
  I just replied to that email 5 min ago18:32.32 
Robin_Watts Right. The guy you need to speak to is paulgardiner, but he's probably gone for the night.18:33.33 
  Try again in about 16 hours time.18:33.59 
  You say "our requirement"...18:34.20 
  what are you trying to do with MuPDF?18:34.29 
t4nk856 integrate MuPDF library & open pdf inside the project18:35.02 
Robin_Watts t4nk856: So you're trying to use MuPDF within your own app.18:35.22 
t4nk856 download pdf from back end service & open in the app18:35.34 
Robin_Watts Will this app be released under the GNU AGPL too?18:35.35 
t4nk856 Yes using MuPDF inside the app18:35.45 
Robin_Watts (which will require you giving away full sources etc)18:36.15 
t4nk856 This is for a well known financial organization in US18:36.19 
  Our product team is connecting with your marketing team @ Texas18:36.36 
  for commercial licensing to use inside the app18:36.51 
Robin_Watts Ok, so you know about the requirement for a commercial license. Cool.18:36.52 
t4nk856 Yes18:36.59 
Robin_Watts I just didn't want that to be a nasty surprise for you.18:37.05 
t4nk856 We can use only if we get license18:37.14 
Robin_Watts Well, I believe the changes to the source shouldn't be huge, but paulgardiner is the man to talk to here.18:37.28 
t4nk856 I am a developer18:37.29 
Robin_Watts What timezone are you in?18:37.34 
t4nk856 PST - San Francisco18:37.55 
  GMT - 818:38.02 
Robin_Watts OK, so you either need to be up early or up late to talk to paulgardiner.18:38.23 
  I will forward your mail to him.18:38.35 
t4nk856 thats fine I normally awake till 12 PM PST18:38.51 
  Thank you for your support18:39.13 
Robin_Watts That's 4am here, I think.18:39.20 
t4nk856 :) but your response was really quick18:39.44 
  what would be the best time to contact Paul18:40.02 
  in PST 18:40.08 
Robin_Watts between 9:00 and 17:00 UK time, so that's... 01:00 and 11:00 PST?18:40.52 
  but on friday's he starts about 2.5 hours later than that.18:41.07 
t4nk856 Ok np...18:41.36 
  If he is able to reply to my email, it would be great18:42.00 
Robin_Watts He's really busy on a different project at the moment, but I hope he'll be able to give you some pointers.18:42.26 
t4nk856 Great18:42.42 
  Thank you Robin. I am going to leave this chat, I am connected to a wifi for accessing irc, irc is not allowed inside our intranet18:44.51 
Robin_Watts np.18:44.58 
t4nk856 bye...I will try to get him on chat if any more doubts after his email reply18:45.44 
mattapperson any MuPDF devs around?19:15.50 
  looking for pricing info, and under a heavy time constraint 19:16.10 
Robin_Watts mattapperson: Hi19:24.28 
  I'm a MuPDF dev.19:24.35 
mattapperson Hey Robin_Watts19:24.36 
  :) 19:24.38 
  any chance you can tell me the pricing?19:24.48 
Robin_Watts but the devs are not generally involved in commercial sales.19:24.48 
  we have a sales guy to handle that.19:25.01 
mattapperson any rough ideas? sales has not responded for several days and I have a client that needs an answer ASAP19:25.27 
Robin_Watts Forward the mail to me (robin.watts at artifex.com) and I will ensure he gets it.19:25.50 
  Had he replied to you at all?19:27.00 
mattapperson not at all19:27.04 
  maybe it went to spam?19:27.10 
  idk :)19:27.12 
Robin_Watts Normally as soon as he gets mail, he sends back a big list of questions.19:27.21 
  I just spoke to scott, he's got your mail and should get back to you in a mo.19:33.45 
mattapperson ty :) 19:34.25 
Robin_Watts You should have got an email from me too (to you and scott). If you want to reply to that giving as many details as possible (like how many you expect to ship, what sort of price they go for, what proportion of the apps functionality you expect to get from MuPDF) that will save time.19:35.10 
  Basically the more details you can give, the more scott can tailor the license proposal to your needs.19:35.37 
 Forward 1 day (to 2014/09/05)>>> 
ghostscript.com
Search: