> From [EMAIL PROTECTED] Thu Nov  2 15:08:55 2000
> To: [EMAIL PROTECTED]
> cc: [EMAIL PROTECTED]
> cc: [EMAIL PROTECTED]
> Subject: Re: lprng user survey 
> Date: Thu, 02 Nov 2000 18:08:17 -0500
> From: Scott Schwartz <[EMAIL PROTECTED]>
>
> You wrote:
> | > * lpc (kill,abort,etc) doesn't make the subserver daemon stop, when
> | >   a large file is being written to the printer by the filter (ifhp).
> | >   So we wait a long time for the filter to finish.
> | 
> | I am curious what version you are using.  The current implementation
> | sends a killpg()  (kill process group) that should kill off the
> | subserver AND the server as well as all processes in the process group.
> | What version of Solaris are you using?
>
> Solaris 7, sparc.  (LPRng-3.6.26; ifhp-3.4.1)

It is hung in a write operation... it has been killed off, but it
is in zombie mode,  and Solaris will keep trying to put the data
in the TCP/IP buffer out.  It will not 'release' the process
until this is done.

This is NOT a bug,  it is a FEATURE of Solaris.

>
> | > * We miss the "clean" command, from bsd's lpc.
> | 
> | What does this do? 
>
> $ /usr/sbin/lpc
> lpc> help clean
> clean           remove cruft files from a queue
>

This is in checkpc:

checkp -r -A 1d


> | By the way,  which of the 28 versions of BSD lpd's
> | are you talking about :-)
>
> Oh basically the one that shipped with SunOS 4.1.3, but I can
> live with whatever RedHat shipped with 6.2. :-)

GAHHHKKK.  You could at least be using SunOS 4.1.4...

>
> [ buffer overruns ]
> | No... not at all... I am VERY puzzled...
>
> Most likely this was due to pilot error: we subsequently changed all
> the cables and made sure the printer was configured the way we wanted,
> and things seem to be working smoothly now.
>
> | > * When a non-lprng (Solaris) system does an lpq on us, they get pages
> | >   of status back.  That seems unhelpful.  The documentation talks about
> | >   Solaris reversing long vs short listings, but that doesn't seem correct.
> | >   We want to see the user's name, the file name, etc.  But not pages
> | >   and pages of debugging output from the spooler and filter; that should
> | >   be reserved for LPRNG verbose output, unless a filter will parse that
> | >   stuff and extract sensible messages to present (like, "out of paper").
> | >   Somehow HP's own print filter does a nice job of this; I wish LPRNG
> | >   would do the same.
> | 
> | The problem here is trying to make something that is usable by
> | everybody.
>
> I understand.  At the same time, I hope you don't mind if I gripe a
> little. :-)
>
> My feeling is that, when using a traditional lpq client, the output from
> lprng is intolerable in both the "short" or "long" format.  Relying on
> lpq to parse and reformat the status it gets from the network is, IMHO,
> a very bad default, because smooth interoperating with traditional lpr
> clients must be a consideration.

Yep.

>
> | I have tried several solutions to this problem and I get equally large
> | complaints from all sides.
>
> It would be helpful it you put an analysis of the various strategies
> you've tried in the howto, to get new users up to speed, because I don't
> understand the problem.  My feeling (principle of least surprise) is that
> if you do exactly what traditional lpd does, that should be satisfactory,
> while lprng enhancements (like returning all the debugging stuff from all
> the log and status files) can be provided with purely optional extensions.

Yep.  Tell me which of the 28 versions you want to simulate,  get a support
contract,  and I will ship you a version that does this.

Now of course,  if I change this,  all of the LPRng users will scream
as this is NOT what they want and it will not be backwards compatibile
with known LPRng operation.

You can compile and install LPRng on any known UNIX system.  I this is
not satisfactory,  then I have some desperation options that will
work as well.

>
> | The only solution I have found that seems to work
> | is to use the LPD option: 'return_short_status=HOSTSPEC'
> | where HOSTSPEC is a wildcard or whatever that matches
> | the various Solaris Systems.
>
> I have that turned on now, but only to silence lpq, so that lpq -a isn't
> completely intolerable.  It is most certainly not a fix, because lprng's
> "short" status omits the brief and useful information that we wanted,
> and that our other spoolers provide.

You have the source.  Enjoy.  That is why it is called 'Open Source'.

>
> | > * Moreover, the "long" status is not just status for the last job.
> | >   It's the concatenation of the last several jobs worth of status.
> | >   Very unhelpful, and also confusing for users.
> | 
> | Unless, of course,  their job is one of the previous ones...
>
> No, not even then, because few if any users is interested in seeing,
> much less decyphering, pages and pages of junk.  If someone wants verbose
> status returned to them for a particular job, maybe you could arrange
> a way to mail that information back to them.  If someone wants to see
> extra-verbose debugging information for a queue, you could arrange an
> option (-ll !) for that.  But the normal queue status report should be
> generally useful per-se.

lpq jobid

>
> | Ummm... your comment about this and the one about  the filter
> | not being killed off seem to be in conflict.
>
> Different occasions.  It's hard to distill days of screwing around into
> a coherent narrative. :-)
>
>

Patrick Powell                 Astart Technologies,
[EMAIL PROTECTED]            9475 Chesapeake Drive, Suite D,
Network and System             San Diego, CA 92123
  Consulting                   858-874-6543 FAX 858-279-8424 
LPRng - Print Spooler (http://www.astart.com)

-----------------------------------------------------------------------------
YOU MUST BE A LIST MEMBER IN ORDER TO POST TO THE LPRNG MAILING LIST
The address you post from MUST be your subscription address

If you need help, send email to [EMAIL PROTECTED] (or lprng-requests
or lprng-digest-requests) with the word 'help' in the body.  For the impatient,
to subscribe to a list with name LIST,  send mail to [EMAIL PROTECTED]
with:                           | example:
subscribe LIST <mailaddr>       |  subscribe lprng-digest [EMAIL PROTECTED]
unsubscribe LIST <mailaddr>     |  unsubscribe lprng [EMAIL PROTECTED]

If you have major problems,  send email to [EMAIL PROTECTED] with the word
LPRNGLIST in the SUBJECT line.
-----------------------------------------------------------------------------

Reply via email to