RE: [hpoj-devel] ptal-mlcd and hotplug scripts

[PASCHAL,DAVID (HP-Roseville,ex1) <dav...@hp...>]

Hi, Allen.

Allen Barnett wrote:
> I tried unplugging and replugging my USB cable a few times 
> and it seems
> to work OK. (Subject to my USB hardware performing properly; 
> I'm afraid
> my 3 year old DELL Dimension is wearing out.) The only problem I have
> with this setup is that my hotplug script starts both ptal-mlcd and
> ptal-printd, but only ptal-mlcd exits automatically. So, each replug
> starts another ptal-printd. Would it be possible for ptal-printd to
> receive notification from ptal-mlcd that ptal-mlcd is shutting itself
> down?
That's a bit of a problem.  From what I've heard, systems running LPRng hang
unless ptal-printd is running at startup.  One solution may be to:
- run ptal-printd from an init script (before LPRng is initialized),
- load ptal-mlcd from a hotplug script (if not from an init script), and
- use a descriptive PTAL device name like mlc:usb:hpojg85 instead of the
uncreative mlc:usb:X that I typically use that corresponds to the
/dev/usb/lpX in use.

> Would it be practical to add a SYSLOG output option to your utilities?
I've been thinking along those lines.  I may direct ptal-mlcd's log messages
to syslog regardless, and to standard output if it's available (or to a
remote console instead of stdout if one is open).

> On another subject altogether: should 'ptal-hp scan' work 
> with the G85?
> I get the error message:
> 
> Scan upload state is 2 (i=11)!
No.  "ptal-hp scan" only works on PML (Peripheral Management Language)
scanners (LaserJets and older scrollfed OfficeJets), whereas SANE currently
only works on SCL (Scanner Control Language) scanners (flatbed and newer
scrollfed OfficeJets).  "ptal-hp scan" was basically a quick and dirty hack
so that people with PML scanners wouldn't have to keep waiting for me to
write a proper SANE backend that supports PML scanners.  :-)

David