Unfortunately, I can't post to apple's bugtracker about CUPS, so I'd try to post bugreport here.
I've discovered, that CUPS-1.6.2, 1.6.3 and 1.7rc1 has some issue in their IPP backend.
When I using my printer (Epson Stylus TX210) via usb backend — it works flawlessly.
But when I setup it as IPP (and other symlinked backends like http/https/ipps) even on the same host (!) cups silently finishes jobs (like successfully completed), but don't print anything. All, that it says in error log is:
>> E [22/Aug/2013:19:52:18 +0700] [Job 643] Invalid printer command "!
After some debugging I discovered that:
>> I [22/Aug/2013:21:20:09 +0700] [Job 648] Started filter /usr/libexec/cups/filter/commandtoepson (PID 3330)
>> I [22/Aug/2013:21:20:09 +0700] [Job 648] Started backend /usr/libexec/cups/backend/ipps (PID 3331)
>> D [22/Aug/2013:21:20:14 +0700] Discarding unused job-progress event...
>> D [22/Aug/2013:21:20:14 +0700] [Job 649] PID 3333 (/usr/libexec/cups/filter/gziptoany) exited with no errors.
>> E [22/Aug/2013:21:20:14 +0700] [Job 649] Invalid printer command "!
That was make me think, that error can be somewhere in compression and encryption, but unfortunately even with force-disabled encryption and compression i get:
>> D [22/Aug/2013:21:38:10 +0700] [Job 655] PID 6324 (/usr/libexec/cups/filter/commandtoepson) exited with no errors.
>> E [22/Aug/2013:21:38:10 +0700] [Job 655] Invalid printer command "!
One time, after long debug session, I even get something more informative then quote-exclamationmark error:
>> D [22/Aug/2013:21:48:44 +0700] [Job 663] Set job-printer-state-message to "Invalid printer command "<1B>@<1B>(R<08>"!", current level=INFO
I mostly debugged it on cups-1.6.3 (since, 1.6.2 has broken russian web-interface and don't wont to work at all for me), but as I mentioned above, cups-1.7rc1 affected too. When I downgraded to, for example, 1.5.2 and removed cups-filters (and keeped gutenprint, conatining commandtoepson filter, for example) — it start to work fine again (but, it seems, every client should downgrade to 1.5).
Is it possible to patch it somehow to restore functionality? :)
ping? Any news?
// By the way, upstream recommends to... just not use PPDs/filters for IPP printing. Which is strange way to solve the problem :(