| <<<Back 1 day (to 2019/11/18) | Fwd 1 day (to 2019/11/20) >>> | 20191119 |
Juliandel | Good afternoon gents | 12:06.35 |
| I came to ask what this error means: "Unrecoverable error: typecheck in .putdeviceprops" | 12:06.46 |
kens | It means that there was a problem configuring the device | 12:07.23 |
Juliandel | I am using this: "C:\Program Files\gs\gs9.50\bin\gswin64c.exe" -I "C:\Program Files\gs\gs9.50\Resource\Init" -sDEVICE=pdfwrite -dPDFFitPage -dAutoRotatePages -o TEST123.PDF "C:\temp\input.PS" | 12:07.27 |
kens | and the error was so serious, it cannot be recovered from | 12:07.35 |
Juliandel | configuring which device? | 12:07.46 |
kens | YOK so you have not specified an OutptuFile | 12:07.50 |
Juliandel | -o | 12:08.03 |
kens | THe pdfwrite device, the one specified using -sDEVICE= | 12:08.04 |
Juliandel | is output? | 12:08.06 |
kens | Yeah sorry Imissed that | 12:08.13 |
Juliandel | So if it's PDFWRITE and has input/output.. what's the hangup? | 12:08.30 |
kens | Ah you have put AutoRotatePages but not supplied a value | 12:08.31 |
| That's not (IIRC) a -d switch | 12:08.48 |
| Oh actually eys it is | 12:08.54 |
| but you need to supply a name value | 12:09.01 |
| You need to supply /None, /PageByPage, or emm something else | 12:09.26 |
Juliandel | is that new? | 12:09.32 |
kens | No | 12:09.35 |
Juliandel | I have this _EXACT_ system set up with gs9.26 | 12:09.41 |
| and I don't get this error | 12:09.45 |
| I am testing 9.50 on a totally new system | 12:09.54 |
kens | You know, it would help if you mentioned these facts | 12:09.58 |
Juliandel | and using the exact same commands | 12:09.59 |
| Oh, I'm sorry! I'm not sure everything you need.. | 12:10.09 |
kens | To save us guesing so much.... | 12:10.10 |
| Well if that worked before then I can't see why it wouldn't now | 12:10.32 |
Juliandel | I'm not meaning to make you go blind. I just don't have enough experience to know which is relevant or not. | 12:10.32 |
| exactly.. | 12:10.38 |
kens | Try removing stuff one at a time till the problem goes awat | 12:10.53 |
| away* | 12:10.57 |
| I'd start by removing the -I | 12:11.02 |
| The the PDFFitPage | 12:11.19 |
| Oh and what's the entire back channel transcript, nt just the error ? | 12:11.54 |
| There might be useful stuff before that | 12:12.00 |
Juliandel | Where can I see that? I'm running it from a bat file atm. | 12:14.11 |
kens | It'll be in the terminal, sicne you are running the command line version.Same place you found the error | 12:14.52 |
Juliandel | Ok, so I redid the same command just to show you | 12:21.03 |
| https://i.imgur.com/1JTufKJ.png | 12:21.04 |
kens | OK so nothign useful | 12:21.31 |
Juliandel | I have custom entries in Fontmap.GS | 12:21.40 |
kens | As I said, start removing bits from teh command line, I'd start by removing the -I | 12:21.43 |
| Lets worry about your fonts after we get GS actually to work | 12:22.04 |
Juliandel | Ok, removed the -I, and the subsequent path to the init folder. Same error. Will try removing next one and so on. | 12:22.42 |
kens | Yes please | 12:22.49 |
| If its not thart, tehn it may be the content of the PostScript file, and I'd need to see it in that case | 12:23.06 |
Juliandel | after removing PDFWRITE | 12:23.22 |
| it did briefly display the project did a bunch of things and closed. | 12:23.33 |
kens | Umm, I wouldn't remove the -sDEVICE | 12:23.35 |
Juliandel | I do see lots of bitching about fonts. xD | 12:23.52 |
kens | Yeah that would not be surprising | 12:24.04 |
| leave the DEVICE alone but removve PDFFitPage | 12:24.16 |
| and then remove -dAutoRotatePages | 12:24.25 |
| That should leave a very simple command line | 12:24.42 |
| If that doesn't work I think I'd need to try it myself with your PostScript file | 12:24.57 |
chrisl | It's AutoRotatePages | 12:28.20 |
Juliandel | Oh? | 12:28.31 |
| Yeah, seems after removing that it's doing something | 12:28.38 |
kens | Well, that's what I expected | 12:28.40 |
| So try setting -dAutoRotatePages=/None | 12:28.53 |
| To turn it off | 12:28.56 |
Juliandel | Ok | 12:29.02 |
kens | or /PageByPage to turn it on | 12:29.03 |
chrisl | I expect we're now correctly handling an error condition we ignored before | 12:29.24 |
kens | Oh the third option is /All | 12:29.43 |
Juliandel | Yeah, I see more information now | 12:29.55 |
| I have specified in the Fontmap.GS: " /Courier (c:/Windows/Fonts/cour.ttf) ; " | 12:30.40 |
| And it complained like so: https://i.imgur.com/agkiooF.png | 12:31.14 |
| Maybe I have some slashes in backward or something | 12:31.40 |
| the file itself does exist. | 12:31.45 |
kens | forward slashes are acceptable | 12:31.50 |
Juliandel | Hmm then. | 12:32.14 |
| Would you like to see the full output? | 12:32.39 |
kens | Not especially | 12:32.46 |
Juliandel | I chose 'PageByPage'. | 12:32.51 |
kens | Won't affect fonts | 12:32.57 |
| d: | 12:33.06 |
| <sigh> wrong window | 12:33.19 |
Juliandel | I wonder why it says it can't find the font files then | 12:33.19 |
| I confirmed the location | 12:33.22 |
kens | Well obviously there's a problem | 12:33.30 |
| if you give me a minute I'll try it here | 12:33.37 |
Juliandel | I'm... not allowed to give the project out. | 12:33.51 |
| However, I can tell you that, this specific project file works just fine with 9.26 GS | 12:34.07 |
| With these same fonts and specifications | 12:34.16 |
chrisl | try adding --permit-file-read="c:/Windows/Fonts/*" to the start of the command line | 12:34.18 |
kens | Might make it pretty difficutlt to help you then | 12:34.19 |
Juliandel | I know kens, and I don't omit it from you to be rude. | 12:34.35 |
| trying chrisl | 12:34.47 |
kens | I'd have to guess that its the file permissions, as Chrisl suggests | 12:34.56 |
Juliandel | that very well could be. I am using an admin CMD prompt though | 12:35.15 |
chrisl | gs file permissions, not OS ones | 12:35.30 |
kens | Its GS's own file security | 12:35.32 |
| Yeah invalidfont in findfont | 12:36.08 |
Juliandel | oh! wow | 12:36.23 |
| So I used my command with that permit | 12:36.32 |
| and it worked | 12:36.35 |
kens | Then that's the problem | 12:36.41 |
| Ghostscript 9.50 is a *lot* more secure about openming files | 12:36.52 |
| Though I *thought* we permitted anything actually in fontmap.GS | 12:37.04 |
Juliandel | That is good since we use it on a server. | 12:37.07 |
kens | In fact I remember testing it :-( | 12:37.13 |
Juliandel | Ok, this is interesting though. | 12:37.21 |
| It still doesn't work with just the plain autorotate, but I don't mind using /all or what have you. As long as it does it | 12:37.43 |
chrisl | That's odd, it *shoud* work. It's possible the string getting added to the list automatically isn't quite right | 12:37.57 |
| Oh, it might be a "/" vs "\" thing | 12:39.37 |
kens | ah.... | 12:39.44 |
| Well the file permissions is all new | 12:39.54 |
Juliandel | I will play with the / vs \ | 12:40.15 |
| but will I need to start using that permit option on fonts folder now? | 12:40.26 |
kens | Seems that way, for this release at least | 12:40.57 |
kens | lunches | 12:41.06 |
Juliandel | Ah. So I started this in the summer. got everything working.. | 12:41.10 |
| Now they wanted me to document how I did it. So I wrote it all up. Then tested it using the latest GS | 12:41.24 |
| and was surprised to bump into this issue | 12:41.31 |
kens | Well the new file security is in the release notes | 12:41.42 |
| Its the major change and the reason this release is not 9.28 | 12:41.51 |
| It was significant enough a change that we bumped the release number up | 12:42.06 |
Juliandel | Ok | 12:42.48 |
| I will read the release notes then | 12:42.54 |
| Thank you both for the help | 12:42.58 |
chrisl | Hmm, yeh, it looks like the patch is lacking the trailing slash it should have :-( | 12:44.50 |
Juliandel | ? | 12:44.58 |
| What do I change? | 12:45.04 |
chrisl | You don't, I do.... | 12:45.12 |
Juliandel | Oh, for next release or something? | 12:45.20 |
chrisl | Yes | 12:45.42 |
Juliandel | Cool :) | 12:45.50 |
| Thank you guys for your help today kens and chrisl | 13:20.55 |
| When do you think next release is? | 13:21.16 |
kens | Apologies, was at lunch, the next release is aimed for next March | 13:23.23 |
Juliandel | No apologies needed, asked it only 3 minutes ago | 13:23.49 |
| Ok, so a ways out still. Good to know. Thank you again for your help today | 13:24.00 |
kens | NP hope it works out for you | 13:24.10 |
Juliandel | me too | 13:24.14 |
kens | Adding the path explicitly should be safe into hte future, even if it becomes redundant | 13:24.25 |
Juliandel | the fonts path? | 13:25.39 |
kens | Yes. | 13:25.48 |
Juliandel | Ok, that does help | 13:25.54 |
kens | As chrisl said (I htink) you shouldn't have to do that, | 13:26.00 |
| For now you will, the next release shoudl fix it, but adding it explicitly won't cause a probelm with that release | 13:26.20 |
Juliandel | That is very good to know | 13:26.53 |
| So did something change with the pdfrotate? | 13:31.18 |
| In 9.26, it does it exactly like we want it to. | 13:31.28 |
kens | Not that I'm aware of, but its been 12 months | 13:31.38 |
Juliandel | I take the same exact project file with 9.50, and a couple pages are not rotated right | 13:31.47 |
kens | As chrisl says, very possibly we now flag an error that was previously ignored | 13:31.55 |
Juliandel | Again, same scripts essentially. Just diff GS versions | 13:31.58 |
| I tried /pagebypage and also /all | 13:32.09 |
kens | Try /None, or simply omitting it | 13:32.24 |
Juliandel | Ok, let me test | 13:32.32 |
kens | I certainly haven't knowingly changed its behaviour in years | 13:32.46 |
Juliandel | If I remove it, or use any of the variables, that page will still be vertical instead of horizontal. | 13:34.59 |
kens | Again, kind of hard to comment without seeing the file | 13:35.17 |
Juliandel | 9.26 we have only the flag "-dAutoRotatePage" with nothing else, and it does it horizontally | 13:35.19 |
| Yeah I know. I'm just confused why the same project acts differently. | 13:35.31 |
kens | Well as I said initially -dAutoRotatePages with no arguments is an error | 13:35.36 |
| I'd have expected that to stick with the default setting, so dropping the switch should behave the same | 13:36.10 |
Juliandel | Yeah removing it does not change the behavior | 13:36.25 |
kens | But since the action is totally dependent on the page content, I can't really say | 13:36.26 |
Juliandel | yeah | 13:36.31 |
| Maybe we should stick to 9.26 for now | 13:36.41 |
kens | Then you are vulnerable to known and publicly disclosed security vulnerabilities | 13:37.04 |
Juliandel | It would be nice to have the newer security enhancements, but it's more important for us to have accurate schematics | 13:37.11 |
| Kens | 13:40.48 |
kens | ? | 13:40.57 |
Juliandel | Assuming I got permission.. and provided a .PS file | 13:41.05 |
| do you offer any sort of privacy promise, etc? | 13:41.17 |
kens | We keep everything private unless people post it to our bug tracker | 13:41.35 |
| Adn will delete on request when finished | 13:41.46 |
| Obviously if you put it on our public bug tracker we kind of assume its OK to be public :-) | 13:42.05 |
Juliandel | Ok. I will create a meeting with my supervisors and discuss the issues here, and see if I can get the OK with some project that does the problem. | 13:42.08 |
| So we can at least mitigate. | 13:42.18 |
kens | Fair enough | 13:42.22 |
Juliandel | I want the security of 9.50 | 13:42.26 |
kens | Maybe someone can come up with a non-sensitive file instead | 13:42.31 |
Juliandel | but need my rotations to work like they do already :) | 13:42.36 |
| yeah, the trick will be finding one that is not sensitive and has the same kinds of conversion problems | 13:42.51 |
kens | Yup, not an easy task probably | 13:43.02 |
Juliandel | exactly | 13:43.06 |
kens | Note that I can't promise I'll change anything | 13:44.13 |
| Free users don't qualify for support so its at our discretion | 13:44.28 |
Juliandel | Ah, I understand | 13:50.44 |
| well, I will bring it up to the bosses and we will see what happens. | 13:50.58 |
kens | OK you can always contact us here | 13:51.08 |
Juliandel | Thanks kens | 13:59.41 |
| Take care, and good night | 13:59.46 |
kens | NP goodnight to you also | 13:59.53 |
| <<<Back 1 day (to 2019/11/18) | Forward 1 day (to 2019/11/20)>>> | |