CERN Accelerating science

Migrating out of Central Printing

  M.Goossens, C. Letertre, M.Marquina and D.Underhill CERN/CN

With the rundown of the CERNVM Service, mainframe-based printing (represented by the Xerox 4050 in the Computer Centre and the IBM 3812 printers spread around the site) must also come to an end.

The Xerox 4050, known under the name of "XEROXDD", has been the main centrally supported device dedicated to bulk printing since many years. Its workload however has decreased quite dramatically during recent years and more modern printers, notably the HP Laserjet, are now able to replace it without important loss of functionality while offering much more accesibility. One such printer, known as "513-cc", is situated centrally in the user area of the Computer Centre.

In view of the relatively high cost of maintaining XEROXDD we intend to stop the rental/maintenance contract of this underused, mainframe connected printer at the end of the year, and so it will not be available as of January 1996. Current CERNVM and VMS users are invited to follow the procedure below in order to migrate to network connected printers, located in the Computer Centre or preferably closer to their place of work.

The IBM 3812 printers have provided excellent service over the years but their lifeblood is the IBM mainframe which dissapears with CERNVM. While they will remain available until the closure of that service, alternative printer capacity will progressively be made available from January onwards. The migration procedure described is valid as well for such printers and we highly recommend that you follow it in order to avoid last minute surprises.

Changing default printer on CERNVM

These are the steps to take in order to replace your default printer in your different CERNVM applications. Let us assume that your default printer is XEROXDD or any of the IBM 3812 printers on-site. You can check this by typing:

DEFAULTS LIST XPRINT

Next get the name of the printer closer to you which can deliver the format you require. Standard printing can still be done in the Computing Centre if you choose 513-cc.

Let us look at an example of how to using this printer for a few applications.

XPRINT

It suffices to set up a default printer:

DEFAULTS SET XPRINT PRINTER 513-cc

MAIL

If you see PF4=Xprint when you run MAIL, the same command as for XPRINT above will already have set the correct printer.

If you see PF4=Print then you must edit the file MAILUSER XEDIT; for this locate the line:

'SETMAIL PRINT.3812 YES'

and change it to:

'SETMAIL PRINT.3812 NO'

From then on MAIL will use the XPRINT and the printer you have chosen for that program.

BOOKIE

Those using SGML/BookMaster via the BOOKIE exec, should type the command

DEFAULTS SET BOOKIE

and specify the printer of your choice in the panel, then exit via the function key PF5

SGML and XSCRIPT

Since neither of these applications supports PostScript printers, users of these utilities will have to make a printed copy of their files before the closure of CERNVM on a 3812 printer and/or generate a ``lineprinter'' copy by the following procedure:

-

set your STORAGE below 16M,
e.g., DEFINE STORAGE 10M, then IPL;

-

choose ``IBM3800'' in the SGML or XSCRIPT panels;

-

after execution of the Script program, a file of type LISTING will be generated on your minidisk. This file can then be printed with the command XPRINT xxxx LISTING.

If you have active documents in one of these formats, the only long-term solution is to convert them to another format using one of the methods suggested at the end of this article.

Changing default printer on VMS

On VMS systems, the default printer, and the road taken to print your file depend on the command used. In particular one can distinguish the following cases for handling a print request:

1. It is passed on directly to the XEROX printer (the default case with XPRINT ).
2. It is passed to CERNVM, which transmits the output to printers directly attached to it (XEROX, any 3812), or forwards the command to the Springer print server when you chose a printer attached to Appletalk. This roas is taken with XPRINT and any of the qualifiers /PRINTER, /NODIRECT, /CLASS, /DEST, /HEADER, /HEX, /HTEXT, /INDEX, /NUMBERED, /PUNCH, /TITLE, /TRC, /UCS, /UPCASE, or /OPT.
3. It is passed to the Springer print servre, which will print the output on the choosen printer (XPRINT/LW or WPRINT). In this case the default printer is taken from the logical variable LPR_PRINTER.

With the removel of XEROXDD at the end of 1995 both the XPRINT default and the option /PRINTER=XEROX will clearly cease to function. When CERNVM stops the only possible route will be that via Springer and so it is recommended that you already now set up your LPR_PRINTER parameters using the following method (where the central printer 513-cc is used as an example).

-

Check your current default printer with the default command
SHOW LOGICAL LPR_PRINTER

-

Define the logical LPR_PRINTER to point to your preferred printer by inserting the following line in your LOGIN.COM
$ DEFINE/NOLOG LPR_PRINTER "513-CC"

-

Use either WPRINT or XPRINT/LW as print command. WPRINT accepts most of the same parameters as XPRINT and offers some new features.

-

To use the PRINT command inside MAIL:

a)

Inside MAIL, set your queue to be either WPRINT$QUEUE (to use WPRINT) or XPRINT$QUEUE (to use XPRINT);

b)

Create in your top directory, either a file WPRINT$DEFAULTS.DAT (if you use WPRINT) that could contain, for instance:

/printer=513-cc/portrait/border

or XPRINT$DEFAULTS.DAT (if you use XPRINT) which could contain, for instance:

/lw/printer=513-cc

For more information, on VXCERN you can type

Migrating your Documents

Simple SGML and SGML/BookMaster documents (i.e., those without tables and mathematics) can be easily translated into the WWW HTML2 standard. However, as HTML2 has no tables and maths, and the table model of HTML3 (implemented partly by Mosaic and Netscape) is quite different from that of Bookie-SGML (which is based on Bookmaster's) it is not straightforward to translate SGML tables into HTML3, while EQN encoded formulae (the math processor in SGML) are quite different from the TeX based box model used in HTML3.

In other words, it is not too difficult to translate simple SGML documents (no tables, no math) into HTML (as a first step, just view your file with your preferred WWW browser, e.g., Mosaic, and correct obvious errors by hand).

For other documents there are two cases:

1. The document is stable, i.e., no more modifications are anticipated. Therefore a good-quality copy on one of the supported printers should be made, and preciously kept as an original for making future hardcopies.
2. Further modifications are anticipated. The source of the document will have to translated into one of the three text processing languages supported at CERN:
   

   FrameMaker

   +

   runs on almost all computer platforms;

   +

   a licence must be obtained;

   +

   quite complex to learn;

   +

   relatively good on simple math;

   +

   ``WYSIWYG'' (you ``see'' what you get);

   +

   styles are supported.

   LaTeX

   +

   runs on all computer platforms;

   +

   completely free;

   +

   quite complex to learn;

   +

   excellent for mathematics;

   +

   batch oriented (you do not see what you get, but previewers are available);

   +

   many styles are available.

   Word

   +

   only runs on Mac and PC;

   +

   CERN has a site-licence;

   +

   relatively easy to learn;

   +

   not very good at math;

   +

   WYSIWYG, with styles available;

   +

   excellent for shorter, unstructered documents.

   All three tools have translators to HTML.

The choice of the right text processing tool is basically one of personal preferences. Options taken by your team, financial considerations, availability on certain machines, level of complexity of the documents, etc., all play a role. In general, any one of these tools is invaluable for certain kinds of documents and not very suited for others.

Finally, the support for CernPaper documents, which were directly based on the Waterloo Script language, had been stopped several years ago. Consequently, one can only try and save as much text of the files as possible, and re-use it with one of the systems discussed above. Alternatively, one could try and store a printable version of the document by creating a LISTING file with the XSCRIPT command (as described at the end of Section ) and keeping the paper and electronic copies.