home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!Germany.EU.net!unido!cat!incom!orfeo!hotb.sub.org!fasel!gunnar
- From: gunnar@fasel.hotb.sub.org (Gunnar Zoetl)
- Newsgroups: comp.sys.acorn
- Subject: Mult-O-printing (Was: Re: ARC vs PC - no debate. (long!))
- Message-ID: <2-zeXwj010n@fasel.hotb.sub.org>
- Date: 21 Jul 92 20:43:29 GMT
- References: <1YWCpMj010n@fasel.hotb.sub.org> <16846@acorn.co.uk>
- Organization: El Gunthor Pasa
- Lines: 43
-
- osmith@acorn.co.uk (Owen Smith) writes:
-
- > In article <1YWCpMj010n@fasel.hotb.sub.org> gunnar@fasel.hotb.sub.org
- > (Gunnar Zoetl) writes:
- >
- > >This would not solve the problem, as it is not the printer driver locking up
- > >the machine, but the application that is printing, as someone from acorn
- > >stated the last time a similar thread occurred. Try it yourself: setup the
- >
- > Sorry to disappoint you, but most of the time taken in printing *IS* the
- > time taken to crank the bytes down the wire to the printer and wait for the
- > printer to do the business
-
- This heavily depends on what you are printing. For pure text, this is true,
- but for dvi files, draw files and such, that require a lot of computation,
- the printing time depends mostly on the time required to generate the page.
-
- I know the point is that the printer driver spends too much time waiting for
- the printer, and this time could be used in a better way.
-
- > To be able to effectively multi-task while printing, we would need
- > pre-emptive tasks with prioritised scheduling (such that printing would be
- > low priority) and we would also need easily switchable graphics contexts.
-
- I am not quite sure if this is necessarily so. I think, if a spooler was
- implemented within the driver, and the spooler would be interrupt driven,
- whereas the application interface of the printer driver (I mean the part
- that gets the data from the apps) just relied on the application to do the
- multitasking, it would provide enough means to multitask while printing.
- It is not quite as primitive as the thing I just wrote, but I think you
- get the idea.
- On the other hand, getting an application to do the printing in multi-
- tasking mode would require a lot more effort. However, if you have coop-
- erative multitasking, as we do, you are free to implement the multitasking
- while printing in any way you like, heading for maximum printing performance
- or maximum usability of the machine while printing. You do not have this
- option when doing preemptive multitasking.
- Now, please do not start this preemptive vs. cooperative multitasking war
- again, I do know that both have their benefits, and also both have their
- shortcomings. (I work with Unix boxes, too) The problem is, we are stuck
- with the cooperative way, so we have to make the best out of what we have.
-
- Gunnar
- --
- Seid nett zueinander ,-----. ___o o
- ,-'----+--`--. _'\ <,_ (\
- '-(_)----(_)-' (_)/ (_) /> Sonst gibts auf die Backen.
-