Re: GnomePrinterSelector private?

[David Cooke <d j cooke lineone net>]

Hi,

I've just picked up on this thread.  I sent in a message a year or so ago 
asking why the gnome-print team seemed to be limiting developers' options, 
but I did not get a reply.

At 13:10 08/01/03 -0500, you wrote:

> I have been wanting to declare the API of gnome-print stable [*1], and
> are feeling _very_ close to it. I stopped installing headers that I
> wasn't sure why app developers would need to use them, this is something
> that I'd like to discuss further.

This seems an inherently conservative approach, in which you are not 
leaving options open that you have not foreseen.

I released an app (GWintree) using gnome-print v0.35, with a separate mini 
shared library providing a slight extension to gnome-print to do what I 
wanted.  As it happened, I'd worked originally with 0.34, and might have 
done things differently with 0.35; but I wanted to port a program from 
Win32 and reproduce the same behaviour.

The distinctive feature of the program is, again, that the output is 
formatted for a precise page size: changing the printer after the 
formatting is done will prevent output if the printable area is known and 
is smaller.  The program therefore requires the printer to be set, and 
rather than prompt the user beforehand, sets a default printer.  I found 
that under gnome-print, I could not at the time set the printer, which 
seemed to me bizarre.

I haven't looked to see how gnome-print has evolved lately.  I would 
suggest, though, that the usability of any framework is likely to benefit 
from a certain generosity, in providing the facilities that a developer 
might need, rather than only those that you can be certain they will.

Regards,
David