[Suns-at-Home] PPD file for HP-4050N printer needed for StarOffice 5.2 on Sparc box

[der Mouse]

>>> %!PS-Adobe-3.0

>> <RANT>
>> If a supposedly-PostScript printer pays any attention to comments,
>> it is broken.

One response said

> If you can suggest a better way for a printer that can print both
> ASCII and Postscript to decide which is which, I'd be interested to
> hear it.

and another said

> Since you are ranting, I'll comment.  This is a hypocritical comment
> coming from a unix user.  Most unix systems use this mechanism to
> determine how to interpret a file.  Look at the start of shell
> scripts -- they are supposed to start with:

> #!/path/to/shell

> Strangely, this looks like a comment to me ....

That's because it _is_ a comment.  And it's ignored by the shell, as it
should be.

Both of you ignored text I wrote saying that such behavior had a place
in

>> attempts to guess what sort of content an unknown file has.

This applies to the shell script when treated as an executable (and if
explicitly passed to the shell, the #! comment *doesn't* need to be
there) and, as I pointed out elsewhere in my message, to printers that
try to be multiple things at once.

> Even odder it looks like PostScript comment above.

That is hardly a surprise; I imagine %! is modeled on #!.

> If the printer had only one personality, or was locked to only one
> personality, then all the files would be treated as if they were.
> Thankfully most HPs are built with multiple personalities, and they
> are normally configured to auto-detect the content.

> These are:

> HP-PCL - [...]
> PS - text
> Text - text

> How do you tell these last two apart ? The printer does its best by
> looking at the incoming text.

> So, this is reasonable behaviour. 

Iff there is a way to override the guess!  Indeed, I have on occasion
wanted to print exactly the same file twice, once treated as text and
once interpreted as PostScript code.  This sort of "guess" renders that
sort of (reasonable, IMO) thing impossible unless there's an override
mechanism.

> If you don't agree with this kind of behaviour then you would be
> better using an OS which doesn't do this.

And indeed, the way I have printing set up on my machines, it doesn't
try to guess anything.  (Imagine that.  I actually live the position I
claim to support.)

> And if you don't like your printer's decision, just tell it what you
> want to do.

Exactly.  If you can.  Most of the auto-guessing environments I've run
into don't provide any way to override the automatic guess.

> You can always [...] attach the appropriate command sequence to the
> start to force a particular mode.  (Of course, how do most systems
> achieve this?  They use this "magic" detection ......)

Which is reasonable - or at least not unreasonable - provided that (a)
there is a "magic sequence" to force each possible mode without
otherwise affecting its output and (b) all these magic sequences are
documented.  Most auto-guessing systems I've seen (eg, lpr) fail (a) as
far as I can tell, and I think all of them fail (b).

					der Mouse

			       mouse@rodents.montreal.qc.ca
		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B