🚀 go-pugleaf

>>>>> On 31 Oct 92 02:59:05 GMT, xxronr@convx1.lerc.nasa.gov (Ron Rivett) said:
In article <1992Oct31.025905.6047@eagle.lerc.nasa.gov> xxronr@convx1.lerc.nasa.gov (Ron Rivett) writes:

Ron> Nntp-Posting-Host: convx1.lerc.nasa.gov


Ron> Nearly all of the Postscript printing that I have been involved with,
Ron> (either as a user or in setting up Postscript printers on Unix systems
Ron> with Network Print Servers, such as an Emulex P3000 or the Milan Fastport), 
Ron> has been in a "one-way-mode" (i.e. sending a Postscript file to a black & 
Ron> white or color Postscript printer).  

Ron> I understand that many Postscript printers can send status information back 
Ron> to the host, but I would like to know how often this "two-way" flow of 
Ron> information is really needed.  Is the ability to receive information back 
Ron> from a Postscript printer really needed in "day-to-day" usage with 
Ron> applications that generate Postscript, or is mainly used to debug 
Ron> Postscript programs or obtain status or other types of information from the 
Ron> printer ? 

Ron> I am asking this because several vendors of Postscript printers are
Ron> insisting that we buy their TCP/IP network interface hardware, instead
Ron> of using a Network Print Server, because this "two-way" flow of
Ron> information in Postscript printing is critical. After 6 months of
Ron> setting up a few pilot systems with Sun and SGI Unix systems, Network 
Ron> Print Servers, and Postscript printers such as a Tektronix Phaser IIIpxi 
Ron> or Apple LaserWriter, I have yet to see a need for this  "two-way" flow of
Ron> information at a user level. Can someone clarify this or point me towards
Ron> any publications that can clarify this issue ? 
Ron>  
Ron> If this "two-way" flow of information is really needed, how is this 
Ron> integrated into a spooling system, such as a BSD lpr/lpd system, and a 
Ron> "production" environment where users are printing documents very often
Ron> (i.e. Not a limited usage environment where a user could monopolize a printer
Ron> for a given period of the day) ?
Ron>  

Ron> Thank you



Ron> ================================================================================
Ron> Ronald R. Rivett                            voice: (216) 433-8281
Ron> Sverdrup Technology                 
Ron> NASA Lewis Research Center          NASA LeRC FAX: (216) 433-8000 
Ron> 21000 Brook Park Road                      E-mail: xxronr@convx1.lerc.nasa.gov
Ron> Cleveland, Ohio 44135 
Ron> Mail Stop 142-2               
Ron> ================================================================================

Ron>  

	I've found printer communication to be useful. On the RS/6000,
if it's configured right (big if), the print system will `write' a
message back to a user indicating that their job has failed, including
the printer output. However, for UNIX, I haven't seen any program that
needed the two-way communication. We considered mounting a printer on
token-ring, but even then, we were going to drive it with a server,
and just use the ring for speed. (38400 gets slow!)