| | <<<Back 1 day (to 2017/02/10) | 20170211 |
| emendelson | Hello. I came across a problem with gpcl6win32.exe and some old special-purpose PCL soft fonts. Briefly: WordPerfect 5.1 for DOS had special Arabic and Hebrew versions that used scalable PCL fonts to print to PCL-compatible printers. The PCL output prints correctly on current LaserJets (I use a P3015), but GhostPCL produces PDF output that has mostly stray marks. I can't imagine why, because GhostPCL produces perfect (bitmapp | 13:37.35 |
| | I'll post a link to some sample output on the next line in case anyone is interested: | 13:37.57 |
| | https://www.dropbox.com/s/f143yy106fcd9o4/ARABIC.PCL?dl=1 | 13:38.08 |
| | I'm trying to put together a system that will produce PDFs from PCL output from Arabic and Hebrew WPDOS. This works perfectly with European-language WPDOS, but not with these versions. | 13:39.23 |
| | Is there any setting that I can use to make GhostPCL work correctly with this output? Am I merely being incompetent? (That's happened before.) Thanks for any help. | 13:39.55 |
| emendelson_ | (That incomplete line up there was supposed to say that GhostPCL produces perfect (bitmapped) PDFs from other scalable PCL fonts.) | 14:13.40 |
| HenrySti` | emendelson: create a bug on bugs.ghostscript.com, attach all the data and a command line. | 14:56.11 |
| kens | emendelson I downloaded the PCL file you linked to and I get exactly the same output from gpcl6win32.exe for rendering (to the display device) as I do from pdfwrite. | 15:29.21 |
| | The pdfwrite device ca't do better than the rsult from rendering, and in some cases (RasterOPs) will not do as well. However, in this case, if thre's a bug I don't think its pdfwrite. Note that PCL is a notoriously badly specified language and you may have stumbled over a grey area in the documentation, or something which simply isn't documented at all. | 15:31.04 |
| emendelson | If you're interested in this, I can send a link to a complete DOSBox-based system containing the printer drivers, etc. | 15:34.35 |
| | But if this is the first time it's come up, it may not be worth the effort. | 15:34.45 |
| kens | Err, no, not interested in that. If you think you';ve found a bug, then please open a bug report at https://bugs.ghostscript.com. Attach the file and a command lien there and someone will look at it. | 15:35.27 |
| | It would be helpful to keep the example file small..... | 15:35.46 |
| | Since I don't speak PCL it won'#t be me that looks into it. | 15:36.02 |
| emendelson | OK, I'll make a file with two or three characters and will post a bug report. | 15:42.18 |
| kens | Perfect, thanks | 15:42.24 |
| henrys | kens: have been on #ghostscript for a while, looks like it has grown | 15:43.17 |
| kens | There's often peoplelurking, sometimes more, sometimes less. Today seems like a busy day :) | 15:43.46 |
| kens | fetches additional caffeine | 15:45.09 |
| | huh network hiccupped | 15:49.26 |
| emendelson | kens, I posted a bug report, somewhat incompetently, using three comments instead of one single report, but I hope it's all clear enough. | 16:00.12 |
| kens | Yes, I just receicved the emails | 16:00.23 |
| | Doesn't matter how many comments as long as all the info is there :-) | 16:00.39 |
| emendelson | The attached file has ten characters in it, but it's still pretty large. | 16:00.43 |
| kens | Yeha well I guess the character descriptions are large. Its more a question of complexity than size. Its easier to debug a few characters than a few hundred... | 16:01.11 |
| emendelson | OK, it would be nice if it could be fixed, but I'm not sure the real world will be affected... | 16:01.32 |
| kens | I'm afraid this is kind of outside my scope, on for Henrys :-P | 16:02.07 |
| emendelson | Thanks in advance to Henrys! | 16:02.25 |
| kens | However, the page renders blank in GhostPCL for me. I'll take the file and try it on a real laserjet later, need to boot up a different computer for that | 16:02.36 |
| emendelson | The page converts to a blank PDF for me, not like the big file I posted earlier. But it works on a real LaserJet. | 16:03.15 |
| kens | Well, I rendered it, rather than sending it to pdfwrite. When pdfwrite matches the rendered output I regard it as not being a pdfwite problem (that bit is mine ;-) | 16:04.04 |
| emendelson | For what it's worth, here's a utility that raw-prints to the default Windows printer under Windows 8 and 10: | 16:04.22 |
| kens | So if there's a bug I think its in the interpreter, which I know well, not a lot about... | 16:04.23 |
| emendelson | http://www.columbia.edu/~em36/windowsrawprint.html | 16:04.24 |
| kens | Not a problem I'll just send it direct to the pritner. | 16:04.40 |
| | Its just that that printer is attached to a different computer (downstairs) which isn't turned on at the moment | 16:05.08 |
| emendelson | I was pretty sure you had a way to send files to a printer (!!) but I thought lurkers might find the utility useful. It certainly makes life easier for me. | 16:06.51 |
| kens | :-) | 16:15.41 |
| | I see Henry confirms there's a problem | 16:15.48 |
| henrys | emendelson: I confirm it's a bug but I'll have to spend some time digging into, it's not immediately obvious what is wrong. Sorry but I can't give you a schedule for a fix, free users can attention "as time permits" | 16:16.24 |
| | I meant "free users get" | 16:17.08 |
| emendelson | I'm not impatient! I'm only grateful for GhostPCL, which has done wonders for me for years and years. And I share your views on free users... | 16:17.21 |
| kens | henrys : I was only looking because it might have been a pdfwrite problem. I was going to dig into why PCL crashes with pdfwrite when using UFST as well, it shouldn't. Might not work right, but shouldn't crash. So if I was going to dig there for one bug, might as well be for 2 :-) | 16:17.25 |
| henrys | kens: it shouldn't crash no, grab the assignement if you want to work on it | 16:18.53 |
| kens | Different problem, henrys, the UFST one has no bug assignment yet. It turned up while Robin was working on the regression testing. | 16:19.58 |
| | Previously we weren't testing UFST+pdfwrite+PCL | 16:20.13 |
| | When I can build a UFST to test with I'll open a bug report. It'll be some light relief from graphics states. | 16:20.49 |
| henrys | kens: I misunderstood you I thought you meant you wanted to look at emendelson 's problem while looking at the UFST issue | 16:21.41 |
| kens | Well, I thought it miught have been a pdfwrite problem, so I'd look at both together, but it seems its not pdfwrite.... | 16:22.06 |
| | To be honest I wasn't expecting you to be online until Monday.... | 16:22.54 |
| henrys | it is a pcl symbol set or downloaded font data parsing problem ... oh right, what the hell am I doing here? | 16:24.10 |
| kens | :-D | 16:24.18 |
| | Forward 1 day (to 2017/02/12)>>> | |
Log of #ghostscript at irc.freenode.net.