[printing-discuss] Re: [Inkjet-list] Joining of this list with printing-driver@freestandards.org

[Raph Levien]

On Fri, Dec 07, 2001 at 12:16:17AM +0100, Till Kamppeter wrote:
> Oi,
> 
> this mailing list was initiated for discussion about the IJS 
> architecture, which is a protocol for plugging drivers into GhostScript 
> (without needing to rebuild GhostScript) and fulfilling the needs of 
> modern inkjet printers.
> 
> Since the last OSDN Open Source Printing Summit we have founded the 
> Printing Working Group at www.freestandards.org 
> (http://www.freestandards.org/printing/) and are discussing the 
> architecture of printing in three subgroups via mailing lists and weekly 
> phone conferences. Here we came to the conclusion that IJS is not only a 
> good architecture for a renderer (as GhostScript) communicating with an 
> inkjet driver but for the renderer communicating with printer drivers in 
> general (also for lasers, dot matrix, ...). as being an architecture for 
> plugging in all kinds of printer drivers without recompiling any already 
> installed software and fulfilling the needs of all types of modern 
> printers it gets more and more a point of discussion in our printer 
> driver subgroup. So we decided to join this mailing list with the 
> mailing list of our printer driver subgroup. The address is
> 
>    printing-driver at freestandards.org
> 
> and subscriptions can be done one
> 
>    http://base.freestandards.org/mailman/listinfo/printing-driver
> 
> where you can also consult the archives.
> 
> All participants of this list and also people interested in IJS and 
> printer drivers are encouraged to subscribe for the printer-drivers 
> mailing list and to participate in the weekly phone conferences.

Hmm. I'm not sure what to make of this. Looking at the list, it
seems that there's another very ijs-like protocol being cooked up,
with a lot of participation from the IBM'ers.

I took on the IJS project with the explicit goal of making a very
simple, lightweight protocol for connecting renderers with raster
printer drivers. I think that we are extremely close to achieving
this goal.

However, it may be that the "extreme simplicity" approach I'm
taking with IJS is not the best thing for users. I don't have much
ego invested in IJS - if people want something different, I do not
want to stand in the way.

I see two paths for myself:

1. Finish IJS. I have a 14-page spec now, and there's some good code
implementing it out there. I think it's possible to ship something
very soon that will work well for a large nuber of users.

I'd be happy to turn over the reins to a group who would like to
extend IJS after this is done. Ideally, the extensions would be done
in a backward-compatible way, so that existing clients and servers
would work just fine, if they didn't need the new features.

2. Abandon IJS. Let the printing working group come up with a
protocol, and accept the client-side implementation as a patch
to Ghostscript when it's done. That way, we avoid the unpleasant
situation of having two competing protocols, which will just create
more complexity for users and implementors alike.

It's unlikely that I will actively participate in this PWG<2>,
because my time is quite limited these days, and I don't consider
working as part of such a committee to be much fun. But I'd be
happy to contribute the draft document and sample code as raw
material for the new project.

I'll post my latest draft tomorrow, and think about this some
more.

I realize I brought this situation on myself to a large extent, by
plowing ahead with IJS even when Omni was actively doing their
implementation. Ah well, live and learn.

Raph