| | <<<Back 1 day (to 2014/11/09) | 20141110 |
| mahtab | Hello | 09:45.45 |
| 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. | 09:45.45 |
| mahtab | I have a query regarding Mupdf android sample application | 09:46.33 |
| | could anybody please help me | 09:46.39 |
| | ? | 09:46.44 |
| | how can I reduce or remove gap between pages in the reader view | 09:47.20 |
| | ? | 09:47.21 |
| jogux | mahtab: probably Robin or Paul can say something helpful about that; they'll probably be around in the next few hours | 09:49.46 |
| mahtab | okk thanks | 09:50.07 |
| henrys | helmut is becoming a "regular" | 15:05.13 |
| kens | I htink its all the same problem, he doesn't understand how the font code searching works | 15:05.34 |
| | (not surprising, since its complicated) | 15:05.43 |
| Robin_Watts | kens: Is there a "Here is how Ghostscript font searching works" document ? | 15:16.05 |
| | One suspects that writing such a thing would save time in the long run. | 15:16.18 |
| kens | Nothing as simple as that, unfortunately | 15:16.19 |
| | Part of the problem is that it truly is complicated | 15:16.33 |
| Robin_Watts | Kens: I understand that; but if we were to write such a document, we could update it each time a question came in that it did not cover. | 15:17.09 |
| kens | THere are several differnt kinds of searches, and people tend to get confused because Fonts are not the same as CIDFonts, and TrueType fonts are 'none of the above', yet can be used as either fotns or CIDFonts depending how you define them | 15:17.22 |
| Robin_Watts | The problem is that without it, people will continue to misunderstand and only get more and more confused. | 15:17.37 |
| kens | TBH it doesn't matter to most people, Helmut's problem is, at least in part, that he is dealing with installations which already have Ghostscript installed on them | 15:18.12 |
| | Which goes back to what Chris was pointing out before, that we only have one environment variable for GS, no matter whic hversion it is..... | 15:18.33 |
| chrisl | There is a section on "how Ghostscript finds fonts" in the docs - I don't know how accurate it is, nor whether it properly covers CIDFonts | 15:18.48 |
| kens | Most people are unaware that CIDFonts are not Fonts.... | 15:19.11 |
| Robin_Watts | "TBH it doesn't matter to most people" - but for those people for whom it matters, it really matters. | 15:19.31 |
| chrisl | Robin_Watts: Reviewing the docs is on my to-do list | 15:20.04 |
| kens | THe fonts.htm doc is quite basic | 15:20.12 |
| | The 'platform fonts' section seems to have disappeared, though the link at the top of the page is still there, which is puzzling | 15:20.51 |
| | And that document doesn't mention CIDFonts at all | 15:21.12 |
| chrisl | Oh, I removed it since we no longer support it - I thought I'd removed the link, too | 15:21.19 |
| kens | Its in the bullet list at the top, but if yo uclick it, nothgin happens | 15:21.40 |
| | That does look like the right file to put a revised set of documentation in though | 15:22.10 |
| | Use.htm should probably then reference that file, rather than containing informaiton itself. Having the info in 2 places can't be good | 15:22.51 |
| chrisl | The trouble is there is overlap there | 15:23.51 |
| kens | Looks like there are 6 different ways to specify search paths for fonts, is it surprising people are confused ? | 15:24.00 |
| | Oh, and CIDFonts are differnt again :-) | 15:24.32 |
| chrisl | IIRC, we never "search" for CIDFonts | 15:25.00 |
| kens | THe docs just ssay 'are fgound using a different method' | 15:25.22 |
| chrisl | Well, that'll be the method defined in the PLRM, I expect | 15:26.11 |
| kens | Seems we only look in /Resource/CIDFont (but also presumably anyhting directly specified in cidfmap) | 15:26.15 |
| henrys | the last sentence in 8.3 CID fonts ... why is that not sufficient? | 15:26.33 |
| kens | sufficient for what ? | 15:26.50 |
| henrys | sufficient for documentation? | 15:27.13 |
| kens | Bear in mind that /Resource is a ROM file system *unless* you specify -I | 15:27.15 |
| chrisl | I think that is sufficient - the deficiency is/may be in describing CIDFont substitutions | 15:27.27 |
| kens | Its also important that people are aware that using -I can override the location of /Resource | 15:27.57 |
| chrisl | Only under certain conditions..... | 15:28.19 |
| kens | Yes, and I'm not sure that's clearly documented either | 15:28.36 |
| chrisl | It is not, as far as I'm aware | 15:28.57 |
| henrys | it looks to me like eveything is documented somewhat, if we don't simplify the process folks are always going to trip over something. | 15:29.01 |
| Robin_Watts | Twiki page! | 15:29.09 |
| kens | prefers documentation to be supplied with the application | 15:29.41 |
| Robin_Watts | That way it can easily be tweaked/expanded etc, and when it stabilises we can roll it back into the release. | 15:29.51 |
| kens | I kind of doubt anyone would alter it, but I'm not averse in principle | 15:30.22 |
| | Up to CHris though, its his area | 15:30.30 |
| henrys | chrisl: The last sentence if 8.3 seems wrong, well at least you should say something about -I causing the interpreter to look in a "different" /Resource directory. | 15:31.57 |
| chrisl | henrys: the -I does not *always* have that effect | 15:32.22 |
| kens | It only does that if you direct it to a directory which includes the init files | 15:32.27 |
| | IIRC | 15:32.47 |
| chrisl | No, that's not correct - it only does it if a search finds the string "Resource" (or similar) in the path | 15:33.07 |
| | Which is just insane....... | 15:33.32 |
| kens | Hmm that seems a little prone to error :-) | 15:33.35 |
| chrisl | Well, that's why I would never have allowed that code out the door | 15:33.52 |
| henrys | chrisl: we've had several of these conversations about -I and friends... I wonder how all that will work with mooscript? | 15:34.01 |
| chrisl | henrys: we'll have to make it work the same | 15:34.19 |
| kens | At the moment, not at all | 15:34.21 |
| chrisl | With muscript, it'll only be fonts we need to worry about, though | 15:34.43 |
| | or mooscript | 15:34.50 |
| kens | Might need CMaps too, but that's kind of fonts | 15:35.10 |
| henrys | chrisl: fonts are a good amount of worry ;-) | 15:35.11 |
| kens | Hmm, and ICC profiles ? | 15:35.34 |
| | Oh they aren't in Resource | 15:36.01 |
| chrisl | henrys: the problem is, over the years, lots of people have added these options, and special cases to these options, and no one has kept an eye on things to keep them consistent and sane | 15:36.27 |
| henrys | chrisl: I don't know how we got in this mess but a reading of 8.2 Font lookup paragraph 1, it's clear, no amount of documentation is going to make this easy to use. | 15:42.49 |
| kens | THat's pretty much what I was trying to get across earlier. | 15:43.15 |
| | We can certainly improve what documentation we have, and it wold be excellent to do so, but... its complicated. | 15:43.37 |
| chrisl | henrys: I agree entirely - like I said, there are just too many things that influence the behaviour in unexpected (and often subtle) ways | 15:43.59 |
| kens | Helmut, like may people is placing *way* too much trust in the Acrobat preflight. | 15:44.28 |
| chrisl | Also, we haven't changed *any* of this since 9.10, to the best of my knowledge, anyway | 15:45.47 |
| kens | And of the files he has sent, one does not display properly in Acrobat, and the other does not contsain the fonts | 15:45.59 |
| chrisl | kens: it may display correctly if you have the appropriate font(s) installed | 15:46.24 |
| kens | Hmm, I doubt it, but let me look closer | 15:46.45 |
| Robin_Watts | Might I suggest we document the current mess as best we can, and give a 'best practice' section on how it should be done. | 15:47.13 |
| kens | The missing fonts are CIDFonts | 15:47.14 |
| chrisl | Robin_Watts: When I get time, I will - if you feel strongly about it, have at it | 15:47.34 |
| Robin_Watts | We can then mark the other sections as deprecated, and look to remove them/fix them in future. | 15:47.41 |
| kens | OK its probably because I don;'t have a CenturyGothc font available | 15:47.54 |
| Robin_Watts | chrisl: Ordinarily I would. But a) I don't do fonts, and b) I'm kinda busy :( | 15:48.20 |
| kens | As are we all :-( | 15:48.32 |
| chrisl | Robin_Watts: we're *all* kinda busy | 15:48.33 |
| Robin_Watts | I appreciate that we are all busy though (you beat me too it) | 15:48.46 |
| | but this sounds like something that should be on a to do list somewhere. | 15:49.00 |
| kens | Chris did say it was on his | 15:49.09 |
| chrisl | Robin_Watts: Besides, I don't think documenting it would help much, bearing in mind he keeps saying 9.10 worked and 9.15 doesn't - and as I said, we haven't changed this stuff | 15:49.34 |
| kens | I suspect he is confused as to what he has tested and on which setup | 15:50.06 |
| | Its very easy to get confused by all this | 15:50.25 |
| Robin_Watts | I suspect that in order to shut Helmut up (sorry, "solve his problem"), we're going to need to explain a sizeable proportion of what would be in such a doc. | 15:50.50 |
| henrys | the documentation isn't lacking we need to go straight to the deprecation! | 15:50.54 |
| chrisl | Robin_Watts: Can you remember: did we have a bug that stopped the registry key retrieval working properly with a Unicode enabled build? | 15:51.17 |
| kens | Deprecating and removing some of this, and rationalising the rest would make me happy, but how many people would scream ? | 15:51.25 |
| chrisl | Every one that some random bit of it was added to help out | 15:51.56 |
| Robin_Watts | kens: How many people that we care about? :) | 15:52.02 |
| kens | WHo knows ? THa'ts part of the problem..... | 15:52.21 |
| | bizarre apostrophe placement r me | 15:52.43 |
| Robin_Watts | chrisl: I can remember no such bug. But then I have my name tattooed on my right thumb so I can remember it... | 15:53.12 |
| henrys | I think this is so extreme it needs to be fixed (and broken for some) ... sometimes you have to do it the Apple way ;-) | 15:53.24 |
| kens | There's a couple of prospects when I do a search, but nothing leaps out | 15:53.42 |
| chrisl | henrys: that will p*ss off a lot more people than changing the device API.... | 15:53.59 |
| kens | Which most people never see :-) | 15:54.14 |
| chrisl | Exactly...... | 15:54.21 |
| | Robin_Watts: I have a *very* vague memory of something related to either environment variables or the registry keys not working correctly - but I try to forget everything Windows related as quickly as I can | 15:55.14 |
| henrys | chrisl: we can get a sense of open source usage by searching SO for the parameters right? | 15:56.00 |
| chrisl | henrys: maybe - I'm more concerned about customers | 15:56.28 |
| kens | suspects most of those will have my name attached :-) | 15:56.35 |
| chrisl | Anyway, I will get around to reviewing the docs, and writing a "do it this way" section at some point | 15:57.39 |
| | In the meantime, I have to go to the post office, and then squash....... | 15:58.03 |
| henrys | chrisl: my instinct is open source users are challenging this part of the code more than customers but I don't know for sure. | 15:58.03 |
| | chrisl: okay | 15:58.14 |
| chrisl | henrys: you may be right, but I, for one, don't have a clear picture | 15:58.48 |
| | TTFN | 15:58.52 |
| rayjj | chrisl_away (for the logs): kens: Note that you can use -sGenericResourceDir=... to set the Resource path independent of -I... That's what I usually recommend to customers | 17:01.59 |
| kens | True, but that just adds to the complexity of interacting path definitions | 17:02.29 |
| rayjj | rather than relying on the 'automagic' wonky method that "tries" to find one using the -I path | 17:02.52 |
| kens | heads off for the night. | 17:04.07 |
| | Forward 1 day (to 2014/11/11)>>> | |
IRC Logs