179538 - [ps] Default solaris print command should call "lp"

Status: RESOLVED FIXED

(Core :: Printing: Output, defect, P4)

Description

[Greg Onufer]

User-Agent:       Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.3a) Gecko/20021110
Build Identifier: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.3a) Gecko/20021110

lpr is part of the optional UCB/Solaris 1.x compatability packages that are not
guaranteed to be installed on all SunOS 5.x systems.  Mozilla printing does not
work by default on systems without those optional compatability packages
installed.  Note that the packages are for compatability with an OS that hasn't
shipped for many years.


Reproducible: Always

Steps to Reproduce:
1. try to print anything from mozilla
2.
3.

Actual Results:  
Nothing.


Expected Results:  
Something being printed.

Comment 1

[Greg Onufer]

Attachment: workaround for systems that use lp instead of lpr

This is NOT a fix, it's just the changes I made for my SunOS 5 builds.	It will
break systems that use lpr and do not have lp.

Comment 2

[Roland Mainz]

Mhhh, the quiz is how to fix that ?
The default is set via unix.js - we would need a way to do somelike like this in
JavaScript:
-- snip --
#ifdef __sun
#define DEFINE_DEFAULT_SPOOLER "/usr/bin/lp"
#else
#define DEFINE_DEFAULT_SPOOLER "/usr/bin/lpr"
#endif
-- snip --

Any ideas ?

Comment 3

[Roland Mainz]

Well, a working workaround would be to use Xprint instead ("Xprt" on Solaris 2.x
uses /usr/bin/lp since eternity :) ...

Comment 4

[Greg Onufer]

The real pseudo code should look more like this:

#ifdef __sysv
#define DEFINE_DEFAULT_SPOOLER "lp"
#else
#define DEFINE_DEFAULT_SPOOLER "lpr"
#endif

since SunOS 4.x systems, if anyone were to be foolish enough to port mozilla to
them, use lpr and should also satisfy the __sun condition.  Note that the
spooler arguments also change.

Comment 5

[Roland Mainz]

Greg Onufer wrote:
> The real pseudo code should look more like this:
[snip]

OK, but that still does not solve my main headache:
How to get this implemented in JavaScript ?

Comment 6

[Kenneth Herron]

This should be pretty easy to fix now. Since this bug was written, all of the
default prefs have been moved into libpref/src/init/all.js, which goes through a
preprocessing phase that specifically supports per-OS defaults.

Comment 7

[Kenneth Herron]

Attachment: Different print_command prefs for solaris

This sets alternate defaults for the prefs "print.print_command" and
"print.postscript.print_command" for solaris. I've tested this with a fresh
profile; the new print command shows up in the print dialog, and I can still
print.

There's no good reason to have two separate prefs for this. We should probably
standardize on "print.postscript.print_command". But that's an issue for
another day.

Does a patch like this require sr?

Comment 8

[Alec Flett]

Comment on attachment 177016 [details] [diff] [review]
Different print_command prefs for solaris

huh? I don't remember #ifdef's working in all.js, unless we're now
pre-processing it. This isn't C.

Comment 9

[Alec Flett]

Comment on attachment 177016 [details] [diff] [review]
Different print_command prefs for solaris

sorry, my bad on the preprocessing..

Comment 10

[Kenneth Herron]

Comment on attachment 177016 [details] [diff] [review]
Different print_command prefs for solaris

Requesting approval to check in. This just changes a couple of default prefs
for solaris.

Comment 11

[Asa Dotzler [:asa]]

Comment on attachment 177016 [details] [diff] [review]
Different print_command prefs for solaris

a=asa

Comment 12

[Kenneth Herron]

Checked in. Resolving FIXED.