Switching to Arch: Outstanding issue w/ printing

Question

I've spent the last six months or so switching to Arch and intel (from OS X/Debian and ppc). I did think I was getting there but I've got stuck on some issues which seem small in some ways but are really problematic given the kind of work I need to do. Part of the problem is that I don't know where to start trying to diagnose or solve them.
I currently have three such issues. One has to do with getting my internal mic to work in usable fashion. The second is to do with the use of external monitors (e.g. for LCD projection). The third is the one which is really, really problematic. When it really matters, I can use one of the university's windows boxes for the mic. And the external monitor is manageable because it works and only causes issues on disconnect. But the third I have no work around for and I'm getting kind of desperate.
Generally speaking, printing works fine. My machine is set up to use a number of different printers: a Dell networked printer on my home LAN, an HP in my office via USB and another HP via USB in another office I work in. Printing is generally fast and good quality.
There are two problems. First, printing periodically gets "stuck" and the only way to get it unstuck is to kill all jobs in the queue, disconnect the printer, switch the printer off and then restart, reconnect and reprint everything. Even this is not impossible. I can get everything printed - it is just a question of figuring out what hasn't printed and remembering to recheck everything each time the printer stalls.
Second, certain documents print so slowly that it is, in practice, impossible to actually print them. I'm not sure what is special about these documents. Usually this happens with PDF files but it is certainly not all PDF files. It affects any which contain graphics but it also affects a lot which do not. Other PDFs print extremely fast.
For example, I am currently trying to print a PDF version of a paper. The job was sent at 18:42:37 BST. It is now 19:16 BST and it has so far managed to print 6 sides (3 pages). I'm printing it from acroread because it refused to print at all from Okular or the command line. This particular PDF is an older JSTOR file and these often seem to be affected. Recent stuff from JSTOR prints fine but older archives do not.
pdfinfo for the file gives:
Producer: iText 2.0.1 (by lowagie.com)
CreationDate: Tue Dec 4 07:13:41 2007
ModDate: Tue Dec 4 07:13:41 2007
Tagged: yes
Pages: 16
Encrypted: no
Page size: 595 x 842 pts (A4)
File size: 339426 bytes
Optimized: no
PDF version: 1.4
pdffonts:
name type emb sub uni object ID
Helvetica Type 1 no no no 134 0
DRRZGT+Code2000 CID TrueType yes yes yes 135 0
Times-Roman Type 1 no no no 140 0
Times-Italic Type 1 no no no 141 0
Times-Bold Type 1 no no no 6 0
acroread's properties for the document suggests that the non-embedded fonts are being substituted. For example, Arial MT is being used in place of Helvetica and Times New Roman PS MT is being substituted for Times.
Here's the command being run from ps:
hp-LaserJet-1320-series 410 <username> A9ROojFUrV 1 Duplex=DuplexNoTumble finishings=3 noHPBookletBackCover noHPBookletFilter HPBookletPageOrder=Normal HPBookletPageSize=A4 HPBookletScaling=Proportional HPEconoMode=PrinterDefault HPOption_PaperPolicy=PromptUser HPOption_Tray2 noHPOption_Tray3 noHPStraightPaperPath InputSlot=Auto InstalledMemory=16_31MB noManualFeed media=A4 MediaType=Unspecified number-up=1 PageRegion=A4 PageSize=A4 Resolution=1200dpi sides=two-sided-long-edge Smoothing=PrinterDefault job-uuid=urn:uuid:18b2396b-4dca-38bc-561d-a85b5fc4d85d job-originating-host-name=localhost time-at-creation=1336758157 time-at-processing=1336758157 /var/spool/cups/d00410-001
Any sort of insight or suggestion will be gratefully received and I am more than happy to post further information if somebody tells me what is relevant. I've enabled debugging info in CUPS but I'm not sure what might be of interest and there is a lot of output.
I'm looking for, ideally, a solution which will allow me to just print (preferably solving both problems but at least the second). A second best outcome would be some sort of work around which would allow me to convert these files or something and then print them. (But convert to what?) Failing either of those, I am definitely interested in knowing what I might read, inspect or experiment with.
Given that printing other files works just fine, there must be something about how CUPS is handling files with some particular characteristics, mustn't there? I just have no idea which characteristics of the files are relevant or what it is doing. "I didn't manage to read the/your paper because Linux doesn't like older PDFs" just doesn't sound like a good excuse...
Last edited by cfr (2012-05-11 18:34:07)

Answer 1

ewaller wrote:@cfr:  Which interface are you using?
Do you mean am I using cups or something else or do you mean to configure the printers or something else?
I usually use the web interface to configure the printers for cups. I'm using the "recommended" HP drivers for the HP printers (and I think I'm using a proprietary ppd for the Dell printer from Dell but I use that printer much less than the others and not for printing these sorts of files).
I see the same problem with an HP LaserJet 1320 and a P3010 series (I think it is a 3015).
I discovered that if I switch to a gutenprint driver things are better for at least some JSTOR files on the 1320 but I still have issues. For some reason, it doesn't then want to print the last page (no matter how many pages there are) and this driver can only print at 600dpi and has fewer options. The printer is waiting for more data but cups thinks everything has been sent successfully. If I resend the last page to the printer, I then get two copies but one prints much later than the other. I assume this is another way in which something to do with the spooling or the way that cups sends data to the printer is going wrong and maybe connected with why printers get "stuck". Aside from the slowness issue there is definitely something not quite right about the communication between computer and printer. It is as though cups has two speeds: super quick (spools so fast the printer can't keep up); and super slow.
I've tried printing problematic PDFs from at least Okular, acroread and using lp on the command line.
I don't understand how the suggestion concerning converting them to png is supposed to work. Do you convert each page to a separate png file? I guess you could script it but it still seems like a convoluted process. (And I assume it rules out double-sided printing.)