| <<<Back 1 day (to 2017/10/05) | 20171006 |
rIMpossible | Hello :) | 09:49.08 |
| I am using gs 9.07 with OpenBSD 6.1-stable. | 09:49.34 |
kens | Then I'd suggest you upgrade | 09:49.43 |
rIMpossible | Does pswrite cover all the other ps* devices, means autodetect if gray, rgb or b/w? | 09:50.15 |
| kens: OpenBSD 6.2 is on it's way with new version | 09:50.29 |
kens | You can always build from source, I very much doubt that a new version of OpenBSD will haev anything like an up to date version of Ghostscript. | 09:51.05 |
| But I don't understand your question. | 09:51.12 |
| pswrite emits (very bad) PostScritp. | 09:51.20 |
| You would be better off using ps2write in any event | 09:51.42 |
rIMpossible | kens: better option? | 09:51.54 |
| ah ok | 09:52.04 |
| kens: /usr/local/bin/gs -q -dNOPAUSE -dSAFER -sDEVICE=psrgb -sOutputFile=~/pOut/toPrinter.ps fromProg.pdf produces good output, device=ps2write produces one empty page (same params as before), device=pswrite produces increases time for printing, but result like psrgb, anything wrong? | 10:03.25 |
| s/increases/increased | 10:03.56 |
kens | Apart from using a 4 year old version of Ghostscript you mean ? | 10:03.58 |
rIMpossible | Yes :D | 10:04.31 |
kens | the pswrite device renders a load of content to bitmaps, whch means you get crappy output, the ps2write device does that much less. | 10:04.35 |
| To the point where we no longer support pswrite or any of its variants | 10:04.56 |
rIMpossible | ... but only in newer version of gs, right? | 10:04.56 |
kens | Well obviously there are 4+ years of bug fixes in the current code | 10:05.16 |
| Without seeing your input file there's not a lot I can say | 10:05.34 |
rIMpossible | kens: Ok, I will try to compile it myself | 10:06.26 |
kens | Should be easy enough provided you have autotools installed | 10:06.54 |
rIMpossible | autoconf, automake, m4 and that tools, yes? | 10:08.44 |
kens | I'm not a Unix expert, but that sounds right. autconf at least is required | 10:09.05 |
| In the ghostpdl directory you do ./configure | 10:09.23 |
| That should run the donfigure script | 10:09.30 |
| That ought to create makefile so then you just do make | 10:09.43 |
rIMpossible | kens: ok, thank you, just installing all the tools | 10:12.39 |
kens | Unfortunately our build expert is out today...... | 10:12.59 |
rIMpossible | No excuses :P | 10:14.42 |
| Kidding ... :) | 10:14.47 |
| I will get it done, hopefully there is no interference with pledge | 10:15.56 |
kens | No idea what pledge is | 10:16.22 |
rIMpossible | A security model based on trust in OpenBSD | 10:16.56 |
kens | No idea then, don't know how it works | 10:17.17 |
| Well, GS doesn't call pledge() | 10:18.39 |
rIMpossible | Current version of gs is 9.22 from here? https://github.com/ArtifexSoftware/ghostpdl-downloads/releases/download/gs922/ghostscript-9.22.tar.gz | 10:18.42 |
kens | Yes | 10:18.49 |
| Or you can grab the absolute bleeding edge from Git | 10:19.01 |
| Our Git repository I should say | 10:19.11 |
rIMpossible | I'm 45, I've still bleeded enough in the long past :D | 10:19.35 |
kens | Well the release is literally 2 days old, so not a lot has changed | 10:19.54 |
rIMpossible | I will report back ... | 10:20.22 |
kens | ok | 10:20.26 |
rIMpossible | Thanks for the moment ... | 10:20.34 |
kens | NP | 10:20.37 |
rIMpossible | configure: WARNING: Unable to include opvp/oprp driver due to missing or disabled prerequisites... | 10:55.46 |
| configure: WARNING: Unable to include opvp/oprp driver due to missing iconv/libiconv... | 10:55.53 |
| libiconv 1.14p3 is installed | 10:56.06 |
kens | Shuoldn't be a problem, unless ou need the OpenPriting device | 10:56.11 |
rIMpossible | *** Parse error in /home/sk/downloads/ghostscript-9.22: Need an operator in 'directories' (./contrib/contrib.mak:1201) | 10:56.43 |
| *** Parse error: Need an operator in 'directories' (./contrib/contrib.mak:1205) | 10:56.51 |
| *** Parse error: Need an operator in 'directories' (./contrib/contrib.mak:1209) | 10:56.54 |
kens | Hmmm | 10:57.00 |
rIMpossible | first was configure /second make | 10:57.02 |
kens | Well, opvp is in contrib, so that would suggest its the problem | 10:57.37 |
| No idea what configure is whinging about though | 10:57.46 |
| Possibly it wants a newer version of libiconv | 10:58.05 |
rIMpossible | which version of automake / autoconf / bison / m4 is needed? | 10:58.08 |
kens | I'm afraid I don't know | 10:58.19 |
| But it doesn't look like that's the problem | 10:58.26 |
rIMpossible | I am compiling as user - not as root | 10:58.53 |
kens | Doesn't matter | 10:59.00 |
rIMpossible | make clean sucks, too | 10:59.20 |
kens | Are you using gnu make ? | 10:59.37 |
rIMpossible | same as with make line 1205/1209 | 10:59.41 |
kens | THe makefile is generated by autoconf, so an error configure is probably what's causing your problem | 11:00.09 |
| Which I'd have to guess is due to the missing dependency for opvp | 11:00.22 |
rIMpossible | automake-1.15p0 GNU Standards-compliant Makefile generator | 11:00.38 |
kens | libiconv appears to be up to 1.15 but I would have thought 1.14 would be sufficient | 11:00.59 |
rIMpossible | autoconf-2.69p2 automatically configure source code on many Un*x platforms | 11:01.03 |
chrisl | You don't need automake or any autotools to build ghostscript | 11:01.16 |
kens | Let me start up Linux and see what I'm using there | 11:01.24 |
| Oh hi chrisl, thought you were out | 11:01.32 |
chrisl | I was.... | 11:02.05 |
rIMpossible | chrisl: any idea on how to build on OpenBSD 6.1-stable, then I remove autoconf/automake/m4/bison | 11:02.10 |
| amd64 version | 11:02.27 |
chrisl | Are you using GNU make or BSD make? | 11:02.34 |
| I'm guessing by the error you posted that it's not GNU make | 11:03.09 |
rIMpossible | GNU make 1.15p0 | 11:03.09 |
| sk | 0 | ~ | $ pkg_info | grep make | 11:03.28 |
| automake-1.15p0 GNU Standards-compliant Makefile generator | 11:03.28 |
chrisl | Ummm, I have GNU Make 4.1 | 11:03.31 |
| That automake, *not* make | 11:03.44 |
rIMpossible | chrisl: How to find out version of make? cmd line param? | 11:04.31 |
chrisl | Possibly make --version | 11:04.40 |
| Or make -v | 11:04.49 |
| It very depends on the implementation | 11:05.05 |
rIMpossible | all 6 versions dont show version | 11:05.10 |
| -v -V --v --V --version --Version | 11:05.25 |
chrisl | Okay, just do: "configure --without-gnu-make && make" | 11:05.59 |
rIMpossible | I will remove and unpack again before .... | 11:06.22 |
| *** Parse error in /home/sk/downloads/ghostscript-9.22: Need an operator in 'directories' (./contrib/contrib.mak:1201) | 11:07.29 |
| *** Parse error: Need an operator in 'directories' (./contrib/contrib.mak:1205) | 11:07.32 |
| *** Parse error: Need an operator in 'directories' (./contrib/contrib.mak:1209) | 11:07.35 |
chrisl | Well, that makes no sense | 11:09.43 |
| Lines 1205 and 1209 in ./contrib/contrib.mak are commented out | 11:10.14 |
kens | make not respecting the comments ? | 11:10.51 |
chrisl | Except it *does* honour the same thing on line 1201 | 11:11.22 |
kens | Not a clue then | 11:11.30 |
rIMpossible | 1205 and 1209 are not commented out | 11:14.14 |
chrisl | If you look at the lines above, they end with a "\" which is the continue on next line indicator | 11:14.48 |
rIMpossible | yeah, I see, that's what is driving me crazy, too | 11:16.07 |
kens | Yeah commented out in my copy. Maybe its not properly honouring '\' for comments | 11:16.12 |
chrisl | but, as I say, it works for the one at 1201 | 11:16.28 |
kens | Yes I was just checking that | 11:16.35 |
| Which is just plain weird | 11:16.41 |
rIMpossible | maybe a hidden control code? | 11:17.05 |
| how may I activate display of hidden control codes in vi? | 11:17.18 |
chrisl | Which doesn't affect *any* other make? That's a stretch | 11:17.35 |
kens | Ther's no random binary in there when I look at it in a binary editor | 11:18.09 |
chrisl | rIMpossible: You want ps2write and co, yes? | 11:18.35 |
rIMpossible | chrisl: as recommended by kens | 11:18.55 |
chrisl | OKay, try "./configure --without-gnu-make --disable-contrib && make | 11:19.36 |
rIMpossible | working ... | 11:20.28 |
| workX :) | 11:20.45 |
| What's in contribution to find? | 11:21.04 |
chrisl | I don't understand the question | 11:22.02 |
rIMpossible | chrisl: What do I not have now, with --disable-contrib ? | 11:22.33 |
chrisl | Oh, mostly devices to drive specific printers | 11:23.00 |
rIMpossible | As long as it is not Brother specific, I can live with it :) | 11:23.47 |
| ./obj/gp_getnv.o: In function `gp_getenv': | 11:24.16 |
| gp_getnv.c:(.text+0x42): warning: warning: strcpy() is almost always misused, please use strlcpy() | 11:24.16 |
| ./obj/gp_unifs.o: In function `gp_open_scratch_file_generic': | 11:24.24 |
| gp_unifs.c:(.text+0x467): warning: warning: strcat() is almost always misused, please use strlcat() | 11:24.25 |
| ./obj/ets.o: In function `ets_create': | 11:24.32 |
| ets.c:(.text+0xdfd): warning: warning: rand() may return deterministic values, is that what you want? | 11:24.32 |
| ./obj/jerror.o: In function `format_message': | 11:24.39 |
| jerror.c:(.text+0x184): warning: warning: sprintf() is often misused, please use snprintf() | 11:24.40 |
| . | 11:24.42 |
chrisl | Please don't post a load of warnings | 11:24.46 |
rIMpossible | chrisl: ok, sorry | 11:25.04 |
| These have been the last outputs, is it finished or unfinished | 11:25.35 |
| ? | 11:25.36 |
chrisl | It'll return you to the command prompt when finished | 11:26.07 |
rIMpossible | after those last warnings it got back to command line | 11:26.39 |
chrisl | check the contents of the "bin" directory | 11:27.05 |
rIMpossible | only gs in there | 11:27.15 |
chrisl | Well, it finished | 11:27.26 |
rIMpossible | ok, so now I remove with pkg tools the old sys from my database and do a make install, yes? | 11:28.09 |
chrisl | To do a make install, you will need to be superuser | 11:28.49 |
rIMpossible | http://sprunge.us/ZSRj | 11:30.35 |
chrisl | Pass, it shouldn't be building anything at that point | 11:32.09 |
| Maybe you need to unpack, build and install all as root | 11:33.00 |
rIMpossible | chrisl: ok, will try this, too | 11:33.15 |
| chrisl: same as root | 11:37.44 |
| unpacked, configured and make'ed as root | 11:38.01 |
chrisl | Well, I dunno - seems to me BSD's make is just broken | 11:38.46 |
rIMpossible | Build dependency: gmake 4.2.1 - installed now - how to tell configure? | 11:45.12 |
| http://ports.su/print/ghostscript/gnu | 11:46.02 |
chrisl | You don't, just do "./configure && gmake" | 11:46.30 |
| FWIW, I just tried it using gmake, and it built and installed just fine | 11:47.24 |
rIMpossible | ok, didn't know that dependency, so you were right, I used bsd make and now I first installed GNU make | 11:48.13 |
| mea culpa | 11:48.23 |
chrisl | It's not your fault, really. There is *so* much variation between make implementations. I *try* to keep as many working as possible, but that behaviour is, um, *very* odd] | 11:49.42 |
rIMpossible | Seems so | 11:50.43 |
| Are the gs fonts included, too. Or do I have to download separately. In OpenBSD we have the fonts as separate package | 11:51.25 |
chrisl | The fonts are in the source package, and are built in the executable | 11:51.57 |
rIMpossible | Good, one step less | 11:53.00 |
| now as root, gmake install, yes? | 11:53.18 |
chrisl | Yes | 11:53.36 |
rIMpossible | first try - same file | 11:56.34 |
| with ps2write | 11:56.44 |
| :( empty page - should be 4 pages | 11:57.32 |
chrisl | Presumable this is a PDF? | 11:58.15 |
rIMpossible | chrisl: yes | 11:58.47 |
| with psrgb and pswrite it's good | 11:59.01 |
| with 9.07 was even 1 empty page | 11:59.20 |
chrisl | Can you share the file? | 11:59.45 |
rIMpossible | It's an eticket for my bus drive to brussels next week ... so would say no, sorry. Will try another file | 12:00.43 |
chrisl | What have you used to view the PDF previously? | 12:01.45 |
rIMpossible | mupdf | 12:01.56 |
chrisl | Okay, should be fine then. | 12:02.13 |
rIMpossible | chrisl: I have a file for you, Vaillant atmotec exclusiv | 12:10.03 |
| I can share that | 12:10.09 |
chrisl | Can you upload it somewhere, or do you need an email address? | 12:11.34 |
kens | Sorry, chatting in another window. If you cna post the file I'll take a look | 12:13.16 |
rIMpossible | Ok, ghostscript converts 4 pages to ps, can be viewed with gs [file.ps] all 4 pages are there in the file. lpr -Plp4-PS [file.ps] gives one empty page and on the display of my Brother MFC-9140CDN (Ignore data - press cancel). With pswrite it works, but not with ps2write, the printer has the latest firmware from brother. My printcap is hand written and worked for years now. Maybe I use - for | 12:17.09 |
| ps2write - the wrong command line switches? | 12:17.16 |
| my command: /usr/local/bin/gs -q -dNOPAUSE -dSAFER -sDEVICE=ps2write -sOutputFile=/home/sk/downloads/Vaillant-atmotec.ps Vaillant-atmoTEC-exclusiv-VCW-DE-254-4-7-HL-R2.pdf | 12:17.28 |
kens | Tha'ts about as asimple as it gets | 12:17.45 |
| It 'sounds like' a Brother fault. We've had trouble with teh Brother printers before | 12:18.29 |
| You could try printing each page as a separate file | 12:18.38 |
| It has been known to have problems with large files previously. | 12:18.51 |
| If your PDF file contained transaprency then the PostScript will consist of a large image, which may cause it trouble | 12:19.15 |
rIMpossible | kens: Have a place for sharing the file? | 12:20.03 |
| files. Will upload .ps and .pdf | 12:20.19 |
kens | DropBox ? | 12:20.29 |
rIMpossible | I don't have a login there | 12:20.41 |
kens | We don't run a public ftp server or anything I'm afraid | 12:20.43 |
rIMpossible | frre service? | 12:20.46 |
| s/frre/free | 12:20.57 |
kens | Anything you can find | 12:20.58 |
| wetransfer ? | 12:21.27 |
| But to be honest, if Ghostscritp can process the PostScript, I'd have to guess its a limitation of the Brother printer. Their PostScript implementation is not exactly great in my opinion. | 12:22.18 |
rIMpossible | https://ufile.io/y61hd .ps | 12:22.41 |
kens | 22MB I'd have to guess your original PDF has transparency | 12:23.30 |
| That's a huge file for 4 pages | 12:23.43 |
rIMpossible | https://ufile.io/gndxy pdf | 12:23.46 |
| kens: this has not 4 pages, the eticket had 4 pages ;) | 12:24.36 |
kens | Well that seems to be a 24 page file, not 4 | 12:24.37 |
rIMpossible | but I dont share a valid eticket ;) | 12:24.57 |
| 1.2M --> 22.9M | 12:25.16 |
kens | The PostScript file look sfine to me | 12:25.49 |
rIMpossible | can you print it, that it looks like the pdf | 12:26.04 |
| ? | 12:26.05 |
kens | Looks ot be getting slow. I suspect something is not freeing up VM. | 12:28.02 |
| I would suggest printing each page separately, but if you can't even print page 1 that's not likely to work. | 12:28.32 |
| Possibly your printer simply doesn't have enough memory for the task | 12:28.47 |
| Hmm, no it just seems to be distiller. | 12:29.53 |
| Its only using 64MB of memory | 12:30.08 |
chrisl | kens: any images that look 1 bit in there? | 12:32.50 |
kens | Could well be there are someline drawings | 12:33.04 |
chrisl | Our git history shows Brother printers not handling CCIT G4 compression well | 12:33.34 |
kens | Yes.... Possibly there's something we can do to stop them being compressed that way. But if its not even doing page 1 that seems unlikely to be the problem | 12:34.09 |
chrisl | "-dNoT3CCITT -dEncodeMonoImages=false" | 12:34.15 |
kens | Only 8-bit on page 1 I think | 12:34.17 |
chrisl | CIDFonts? | 12:35.12 |
kens | Not sure I haven't looked much at the file | 12:35.38 |
| It does get very slow with other PostScript itnerpreters towards the end of the job. I suspect there may be full page images in there | 12:36.06 |
| My 'other' PostScript rip can handle it just fine though | 12:36.19 |
chrisl | I'm reluctant to print 24 pages.... | 12:37.00 |
kens | Distiller is still running. I suspect it is either trying to merge images (my other rip mentioned that) or identify duplictae images | 12:37.27 |
| Processing large images to do either of those tasks would be slow | 12:37.36 |
| But we know Distiller does that | 12:37.43 |
| chrisl you could chop page 1 out and print it. | 12:38.02 |
| I could go warm up my laser printer I guess | 12:38.10 |
| Hmm, there can't be any *really* large images in there, it doesn't even start page 1 till about 800=% through the file | 12:39.05 |
| Ah, I suspect I see the problem | 12:39.17 |
| We define a *lot* of fonts | 12:39.37 |
| all type 3 fonts, so I'd have to guess the original contains CIDFonts | 12:39.53 |
| I expect that the Brother simply doesn't have the capacity to store all the fonts we create, because we create them all up front | 12:40.16 |
| So printing each page to a separate file woudl reduce the burden and may well work | 12:40.33 |
chrisl | but the glyphs will be imagemasks | 12:40.38 |
kens | pswrite, of course, doesn't use fonts | 12:40.44 |
rIMpossible | wait, I will look for techdata of the printer | 12:40.52 |
kens | chrisl yes, but eahc font will use VM | 12:40.56 |
chrisl | kens: Yes, but it was bitmap glyphs in a Type 3 that caused us to add the "-dNoT3CCITT" for Brother printers | 12:41.31 |
kens | Good pooint yes, it could be that that's the problem also | 12:41.47 |
rIMpossible | 192 MB memory and 333 MHz processor | 12:42.19 |
| can read german? so I post the link to the doc | 12:42.44 |
kens | As chris suggested you could try disabling the CCITT G3 fax compression, because we know Brother's implementation of that is poor | 12:42.49 |
| I cna read German yes, but its not hugely useful probably | 12:43.07 |
rIMpossible | will try to get it in en | 12:43.30 |
kens | Not a problem, I don't htink the specs will help | 12:43.39 |
| I'd try printing just one page | 12:43.45 |
chrisl | FWIW, my HP printed the first four pages of the valiant boiler document in ~30 seconds | 12:44.01 |
kens | ad use the switches Chrisl suggested | 12:44.01 |
| Well, that sort of sounds more like its the CCITT fax problem, again :-( | 12:44.23 |
chrisl | There's a reason I keep buying HP (however much I dislike the company and their business practices) | 12:45.12 |
rIMpossible | http://support.brother.com/g/b/spec.aspx?c=eu_ot&lang=en&prod=mfc9140cdn_eu_as_cn | 12:45.20 |
kens | So: | 12:45.26 |
| gs -sDEVICE=ps2write -sOutputFile=out.ps -dFirstPage=1 -dLastPage=1 -dNoT3CCITT -dEncodMonoImages=false input.ps | 12:45.26 |
| Sorry EncodeMonoImages not EncodMonoImages | 12:45.49 |
rIMpossible | EmulationsPCL6, BR-Script3 (PostScript ® 3â¢) | 12:46.42 |
| Resolution600 dpi x 600 dpi, 2400 dpi (600 dpi x 2400 dpi) quality | 12:46.44 |
kens | rIMpossible : it claims to support PDF 1.7, so why not just send it the PDF file / | 12:46.55 |
| You could also use the pxlcolor device to produce PCL for it instead | 12:47.32 |
| Note that BR-Script3 is Brother's own (pathetic) implementation of PostScript. | 12:48.10 |
| We should send Scott to them and get them to buy Ghostscritp instead..... | 12:48.24 |
rIMpossible | kens: That is 9330/9340 | 12:48.30 |
| I have 9140 | 12:48.42 |
kens | Oh OK | 12:48.46 |
| Well you have anumber of alternatives (I cna't quite remember what the original problem was) | 12:49.10 |
| You could carry on using pswrite | 12:49.16 |
| You could try using ps2write with the -dNoT3CCITT | 12:49.38 |
| You could use the pxlcolor device to create PCL output instead of PostScript | 12:49.55 |
rIMpossible | kens: The original problem was, I had 9.07 and you suggested to use 9.22 because of the better ps2write device | 12:50.00 |
kens | Yeah sadly that's not true when using CIDFonts | 12:50.16 |
rIMpossible | And then we ran into compilation/installation problems ;) | 12:50.18 |
kens | Because ps2write doesn't support CIDFonts, because they were added after the original PostScript 2 specification | 12:50.37 |
rIMpossible | kens: Ok, so I have to setup the pcl queue in printcap | 12:50.49 |
kens | So until the stalled CIDFont support in ps3write gets finished, there sis less advantage in using ps2write | 12:51.14 |
rIMpossible | ok, I will try the command you suggested above | 12:51.25 |
kens | Trying one page at a time, and then 4 pages should tell us whether its the known Brother problem with CCITT G3 fax, or simply overwhelming the VM with too many fonts | 12:52.17 |
| I have a sneaky suspicion chrisl is correct and it will turn out to be the compression | 12:52.44 |
rIMpossible | 4 pages converted - first page empty - no following | 12:56.38 |
| 1 page converted - first page empty - no following | 12:56.56 |
kens | Well that isn't great | 12:57.23 |
| It does, unfortunately, seem like its a Brother printer. We've got 4 other PostScript interpreters (including a physical printer) which are happy with the original file, so I don't htink its us. | 12:58.00 |
rIMpossible | but viewing out.ps with gs on screen shows the first page | 12:58.01 |
kens | Yeah its pretty clearly something about our PostScript that the Brother interpreter isn't happy about | 12:58.29 |
| Finding out what it is could take a long time and a great deal of experimentation | 12:58.50 |
| Unless you really want to go down that road, I'd suggest trying one of the other solutions. | 12:59.15 |
rIMpossible | know how to unpack that on unix? sk | 0 | ~/downloads | $ file BrMain485.exe | 13:02.41 |
| BrMain485.exe: MS-DOS executable PE for MS Windows (GUI) Intel 80386 32-bit | 13:02.47 |
kens | A DOS executable ? I don't see how you can 'unpack' that | 13:03.31 |
chrisl | If it's a self-extracting zip then unzip ought to do it | 13:03.59 |
kens | Apparently its a driver download ? | 13:04.15 |
| But its a Windows device driver ? | 13:04.32 |
rIMpossible | kens: its a newer firmware, thought I had latest, but end of july a new one came out | 13:04.48 |
kens | Umm, well that might help | 13:04.56 |
| You never know, they may even have fixed a bug | 13:05.16 |
chrisl | <cough>!! | 13:05.27 |
kens | Well, when I run that here, it runs a Brother firmware update too | 13:06.56 |
| tool | 13:07.01 |
| Which searches for a printe rto update, and unsurprising.y doesn't find one | 13:07.12 |
| But its running a Windows executable | 13:07.25 |
| So you'd have to run it on a Windows box | 13:07.33 |
rIMpossible | kens: no there is java inside | 13:07.38 |
kens | Ah, but RAR will open it | 13:07.54 |
rIMpossible | but I rarely unpack those pe archives, and don't remember | 13:08.05 |
| how I did in the past | 13:08.13 |
kens | Lots of XML for different languiages | 13:08.18 |
| Quite a few DLLs | 13:08.30 |
rIMpossible | unrar x file did it | 13:08.55 |
kens | I have to say I can't see any obvious firmware in there | 13:09.25 |
rIMpossible | No, that's only the updater, they collect it from a .jp server | 13:11.07 |
kens | Right, well that would explain why it appears to be only a windows executable plus XML to configure it | 13:11.34 |
| FWIW Distiller did, eventaully, create a PDF file from teh 24-page PostScript. I think it spent an awfully long time trying to decide if the thousands of imagemasks for the text were the same or not. | 13:12.22 |
| Hmm, oddly the fonts in the original PDF seem to be straight type 1 fonts, not CIDFonts. | 13:13.23 |
| We are definitely producing a type 3 fot it though. | 13:14.37 |
| ah, they are type 1C (CFF) fonts | 13:15.44 |
| chrisl do you recall when CFF support was added to PostScript ? | 13:16.11 |
chrisl | Thought it was Level 3, but could be wrong | 13:16.31 |
kens | Well anything later than level 2 would explain why we revert to type 3 btimaps | 13:16.51 |
chrisl | Level 2 only lists Type 0, Type 1 and Type 3 | 13:19.24 |
kens | Well that explains *why* we are producing type 3 | 13:19.37 |
| We really could use a decent ps3write. | 13:19.48 |
chrisl | The Type 2 spec was published in 2000 | 13:21.16 |
kens | Pretty definitive | 13:21.25 |
rIMpossible | Hmm, I have something interesting to configure the ps queue deeper | 13:21.55 |
chrisl | And the LL3 reference lists it as a LL3 feature | 13:21.57 |
kens | OK so we would need a ps3write then | 13:22.07 |
chrisl | Well, we've been trying..... for years now :-( | 13:22.23 |
kens | is still deep in PDF 2.0 | 13:22.42 |
rIMpossible | https://ufile.io/0wc0h | 13:25.12 |
| https://ufile.io/mkc9t | 13:25.59 |
kens | No filter seem the correct setting to me | 13:26.17 |
rIMpossible | https://ufile.io/eusr8 | 13:26.37 |
| hope all 3 are different | 13:26.45 |
kens | In all honety these seems like htings you shouldn't have to change | 13:27.26 |
| The 3rd image seems to match the 1st | 13:27.44 |
| Has the same name too | 13:27.59 |
| Hmmm | 13:28.51 |
| So if you create a file with the following: | 13:29.06 |
| %!PS | 13:29.06 |
| showpage | 13:29.06 |
| and send that to your printer, what happens ? | 13:29.15 |
| Not that it matters really. If the pswrite output worked, then so shold the ps2write output | 13:29.30 |
rIMpossible | Hope that's the right now: https://ufile.io/4x82f | 13:31.04 |
kens | Yes, but..... | 13:31.53 |
| THe fact remains that however you had it configured and working with pswrite, it should continue to work with ps2write using the same configuration | 13:32.25 |
rIMpossible | In the printer queue settings, I didn't change anything since I bought the printer. Wanted to show you, maybe there are options to change, that it works. But seems not to ... | 13:34.20 |
kens | If it worked before, the it should continue to do so | 13:34.37 |
rIMpossible | ps2write seems not to work with brother, pswrite, psrgb, psgray does | 13:34.53 |
kens | I think you;ve simply stumbled over a bug in the Borther interpreter. Realistically we don't have time to try and figure out what's wrong with it (but we've had trouble with it before). I'd suggest you explore other alternatives as I mentioned. | 13:35.38 |
| You could always send it to Brother as a bug, if you're feeling lucky | 13:36.05 |
rIMpossible | I have 6 queues, maybe I test the binary queue | 13:36.30 |
| or pcl queue | 13:36.38 |
| kens: I know, that you cannot test all printers with your software, we all 3 (including chrisl) spent time on that issue. And I will try further and report back. I believe, brother has an implementation error of the protocol, where ps2write is more strict, the ps* devices are more chilled ;) | 13:38.35 |
kens | Well the pswrite devices are much simpler. A lot more stuff ends up being rednered to images | 13:39.09 |
| Which is simpler, but cruder | 13:39.18 |
chrisl | or worse, rectangles | 13:39.25 |
kens | and generally larger too | 13:39.29 |
| Yeah rectangles is even more hideous | 13:39.40 |
rIMpossible | the output from firefox, printing with lpr works, but they have their own implementation | 13:39.57 |
kens | Well, seems you have some solutions anyway, I really need to get back to work | 13:40.19 |
rIMpossible | *selecting .ps as output and print to file | 13:40.21 |
chrisl | IIRC, firefox just renders an image, and wraps it in some PS | 13:40.52 |
rIMpossible | kens: thank you for spending time with that, if I can give positive feedback, I will give detailed information. I am always willing to help to improve software | 13:41.47 |
kens | Thanks, if you can figure out what's wrong we'll endevaour to fix it. | 13:42.06 |
diverdude | I have 2 pdf files. in each there is 1 page. On each of those pages there is only information on the top half of the page. Is it possible to open each of the files and cut out the top half of the page and then merge the 2 half pages into 1 full page in a new pdf file? | 14:40.23 |
| in terminal | 14:40.33 |
kens | Yes, with some PostScript programming | 14:41.18 |
| It would take me some time to figure out a complete solution, and I'm kind of busy at the moment | 14:41.52 |
| The easiest way to do it is multiple passes. First set up a fixed media the size of the area you want. Set a PageOffset whch moves the content into the media space (negate the height basically) | 14:42.48 |
| Create the two new PDF files from teh original | 14:42.59 |
| Then do one more pass. Set the media to be the size required for the composite, then | 14:43.58 |
| Then look at this Stack Overflow answer | 14:44.05 |
| https://stackoverflow.com/questions/30545382/ghostscript-for-pdf-how-to-fit-4-pages-into-1/30546987#30546987 | 14:44.07 |
diverdude | ok nevermind...thanx anyway :) | 15:00.16 |
| Forward 1 day (to 2017/10/07)>>> | |