Hi,
It's very quiet on this list about printing, so I thought it would be a nice idea to put forward a proposal. I don't know if there are internal redhat discussions about this subject, but I think I'll find out soon :). It's a rahter lengthy explanation with a diagram so I put it up here: www.beryllium.net/~remco/printingproposal.html
Kind regards,
Remco Poelstra
On Sun, 06 Mar 2005 22:36:20 +0100, Remco Poelstra remco@beryllium.net wrote:
Hi,
It's very quiet on this list about printing,
look back in the archives.. there is a thread started on feb 1st Subject: system-config-printer.
-jef
On Sun, 06 Mar 2005 23:23:52 +0100, Remco Poelstra remco@beryllium.net wrote:
A thread a month ago, with no real results (as far as I can judge) is not really something I call 'active'.
i'm just making sure you have read it... as a reference. You wanted to have an understanding of what red hat internal discussion was.... that thread is probably the best discussion to-date on this list concerning printing configuration and was conducted primarily among red hat developers who are spending some effort on the issue. More likely than not more discussion relating to that thread has continued between tim and john elsewhere, since tim gave every indication that he was seeking input on something he was 'actively' working on the issue of replacing s-c-printer.
John's posts in the thread detailing how desktop printing confuration is currently handled seems particularly relevant to your proposal... so i wanted to make sure you have read it.. so that you would be better prepared to talk specifics with the specific red hat developers.
-jef"give me an R. give me an E. give me a S. give me an E. give me an A. give me an R. give me a C. give me an H. What's that spell!"spaleta
On Sun, 2005-03-06 at 18:09, Jeff Spaleta wrote:
On Sun, 06 Mar 2005 23:23:52 +0100, Remco Poelstra remco@beryllium.net wrote:
A thread a month ago, with no real results (as far as I can judge) is not really something I call 'active'.
i'm just making sure you have read it... as a reference. You wanted to have an understanding of what red hat internal discussion was.... that thread is probably the best discussion to-date on this list concerning printing configuration and was conducted primarily among red hat developers who are spending some effort on the issue. More likely than not more discussion relating to that thread has continued between tim and john elsewhere, since tim gave every indication that he was seeking input on something he was 'actively' working on the issue of replacing s-c-printer.
John's posts in the thread detailing how desktop printing confuration is currently handled seems particularly relevant to your proposal... so i wanted to make sure you have read it.. so that you would be better prepared to talk specifics with the specific red hat developers.
We would like to have input on how the printing system should work from outside developers but Jeff is right here. You need to understand how it currently works and what decisions have been made concerning the printing system. Having some design that does not take into account what has already been discussed is of no use to anyone. There have been numerous pitfalls I had run into in designing the current desktop auto configuration system in FC3 and have insight into what needs to be fixed which was somewhat reflected in the previous thread. One month is not a huge amount of time as I don't think everything will be fixed in the core 4 time frame. The core desktop team currently have been busy porting to the new dbus and hal api's. That is almost done and on our agenda is to get back to the printing issues. At that point I will look more closely at your proposal.
BTW you need to illustrate the system/session split in your proposal. On an initial glance the flow looks wrong to me (things like you will never get cups accepting dbus commands directly). It needs to be more detailed. Thanks.
-- J5
On Mar 8 mars 2005 22:36, Remco Poelstra a écrit :
John (J5) Palmieri wrote:
porting to the new dbus and hal api's. That is almost done and on our agenda is to get back to the printing issues. At that point I will look more closely at your proposal.
Good to hear that it's on the agenda, because after some 40 years, unix deserves a good printingsystem.
It's a pity new components like cups solve the driver problem while /dev/nulling parts like routing that were the strength of lprng and friends.
(is cups still unable to do something as stupid as queue alaiasing ?)
Jeff Spaleta wrote:
John's posts in the thread detailing how desktop printing confuration is currently handled seems particularly relevant to your proposal... so i wanted to make sure you have read it.. so that you would be better prepared to talk specifics with the specific red hat
Ah, I see, thank you. Since you quoted another line of me, I tought you meant something else. I've read the thread and I thought my ideas were somewhat compatible with the current ideas/implementation. Since the reply of John was somewhat of the same content, I'll consider my design to be uncompatible and think it all over again :).
Remco
John (J5) Palmieri wrote:
porting to the new dbus and hal api's. That is almost done and on our agenda is to get back to the printing issues. At that point I will look more closely at your proposal.
Good to hear that it's on the agenda, because after some 40 years, unix deserves a good printingsystem.
BTW you need to illustrate the system/session split in your proposal. On an initial glance the flow looks wrong to me (things like you will never get cups accepting dbus commands directly). It needs to be more detailed. Thanks.
Hmm, from your reply I make up that DBUS works in a different way than I think. I'll read into it and come back with a better/different proposal.
Anyway nice to hear that RH is interested in the view of "outsiders".
Remco Poelstra
On Mon, Mar 07, 2005 at 12:11:18AM -0500, John (J5) Palmieri wrote:
There have been numerous pitfalls I had run into in designing the current desktop auto configuration system in FC3 and have insight into what needs to be fixed which was somewhat reflected in the previous thread. One month is not a huge amount of time as I don't think everything will be fixed in the core 4 time frame.
Well, it won't be fixed in the Fedora Core 5 time frame either unfortunately.
We seriously need to be designing a new printer configuration tool or mechanism now to stand any chance of getting this fixed for Fedora Core 6.
Tim. */
-
Jef Spaleta -
John (J5) Palmieri -
Nicolas Mailhot -
Remco Poelstra -
Tim Waugh