Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!think.com!mintaka!bloom-beacon!eru!hagbard!sunic!mcsun!ukc!strath-cs!prlhp1!gupta
From: gupta@prlhp1.prl.philips.co.uk (Ashok Gupta)
Newsgroups: comp.sys.apollo
Subject: Re: sr 10.2 xon/xoff problem.
Message-ID: <1314@prlhp1.prl.philips.co.uk>
Date: 12 Apr 91 09:22:34 GMT
References: <1991Mar25.152157.25840@ccs.carleton.ca>
Reply-To: gupta@prlhp1.UUCP (Ashok Gupta)
Followup-To: comp.sys.apollo
Organization: Philips Research Laboratories, Redhill, UK
Lines: 44

In article <1991Mar25.152157.25840@ccs.carleton.ca> ardavan@ccs.carleton.ca (Ardavan Tajbakhsh) writes:
> 
>     We have a Laserwriter II NT connected to a network of Apollos running    
>  sr 10.2 (patch level PSK5) and we are using aegis print services. Everything
>  is fine except when a large postscript file is sent to the printer.  The 
>  printer acknowledges that it is receiving something, but doesn't generate
>  any output.
>
> I suspected that because of the size of the file (150355 bytes), the printer
> tried to XOFF the line but failed.I put a terminal in place of the Laserwriter
> verified that I was talking to it, sent the file and saw the output scrolling
> on the screen.  Typing <CTRL> <S> on the terminal had no effect.
> 
>  Incidentally, here are the characteristics of the line connected to the 
>  Laserwriter:
>
>New line delay: 0 
>Speed: 9600
>                                       Ardavan <Ardavan@ccs.carleton.ca>

I've had similar problems when printing a largish paper (40 pages with 6 
drawings, composed with DOC and PICED) on our DataProducts laser printer.

The printer would take forever processing a page and then spit out a blank 
page ! Or a page which printed on one day would not print the next !  It'll 
help you home in on the problem if you find you get different results by 
changing the fonts or printing parts of the document at a time.  The printer
may have a preference for some fonts and may need to do a lot of computation
if other fonts are used.  Its memory is also limited and may not be up to the
task. 

It may help to increase the `delay' parameter. The characteristics you 
showed in your message did not indicate the time the server waits for, 
between sending stuff to the printer.  On our system there is a parameter
`nlc_delay' which when set to 100 (the max) *really* helps.  This delay gives
the printer enough time to process the current page before its buffers are 
overwritten by new data from the server.  I guess the problem stems from the 
communication error protocol between the printer and server being necessarily 
of minimal functionality.

Another thing to look out for is the version of Postscript.  Make sure its
the latest.
 
 - Ashok