home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.amiga.datacomm
- Path: sparky!uunet!hela.iti.org!usc!rpi!pooler
- From: pooler@vccsouth18.its.rpi.edu (Robert Peter Poole)
- Subject: Re: Weird problem with Term 2.4 and CPUblit
- Message-ID: <h8p1v-#@rpi.edu>
- Nntp-Posting-Host: vccsouth18.its.rpi.edu
- Organization: Rensselaer Polytechnic Institute, Troy, NY
- References: <olsen.5173@sourcery.mxm.sub.org> <jgn1pbq@rpi.edu> <92315.204742DXB132@psuvm.psu.edu>
- Date: Wed, 11 Nov 1992 16:03:02 GMT
- Lines: 58
-
- In article <92315.204742DXB132@psuvm.psu.edu> <DXB132@psuvm.psu.edu> writes:
- >In article <jgn1pbq@rpi.edu>, pooler@vccsouth29.its.rpi.edu (Robert Peter Poole)
- >says:
- >
- >>In article <olsen.5173@sourcery.mxm.sub.org> olsen@sourcery.mxm.sub.org (Olaf
- >>Barthel) writes:
- >>>In article <t7h1+w#@rpi.edu> pooler@aix02.ecs.rpi.edu (Robert Peter Poole)
- >>writes:
- >
- >>>Scrolling will not be faster, though. If you really wish to use 16 colours h
- >>wit
- >>>`term', try an A4000.
- >>>
- >
- >>Not everybody can afford to get rid of his Amiga 3000 and buy a 4000.
- >>Besides,
- >>the blitter in the 4000 is essentially unchanged. Granted, chip bandwidth has
- >>improved because of AGA. But I find it hard to believe that The cure-all is
- >>AGA. As I said, if the CPUBlit hack worked before, it should work now.
- >>Something is amiss.
- >
- >Blitting is inherently very slow in ECS 16-color mode. It is very very slow
- >if you combine that with overscan. CPUblit doesn't solve the problem.
- >You could get fast scrolling with copper assist, but the OS doesn't really
- >support it.
- >
-
- *sigh* Why is it that people seem to totally miss the point of what I am
- saying? Are people on the net really that stupid, or is it just that I'm not
- explaining myself clearly?
-
- YES, I F*CKING UNDERSTAND THAT ECS HAS LIMITATIONS IN BLITTING THINGS ON A
- 16 COLOR HIRES INTERLACE SCREEN. I also understand that CPUblit is not a magic
- cure.
-
- However, do YOU understand that CPUBlit does improve the performance of Term
- 2.4 significantly the first time I run Term? And that after I exit Term and
- start it up again, CPUBlit suddenly stops working? Do you understand this?
- Do you? Good. Then stop getting off the topic and start posting something
- relevant. If I wanted to hear about the limitations of ECS I wouldn't need to
- get it from the net, I could read a spec sheet.
-
- What I would LIKE to hear is why this happens. Is it a bug in CPUblit that
- only appears under Amigados 3.0 on a 3000? Is it a bug in Term 2.4 that only
- shows up under these (admittedly strange) circumstances? One of the two
- programs is doing something screwy. I honestly am not sure which, but I think
- it bears looking into, since this seems to be a side effect of Term 2.4's
- cleanup before exiting. (Otherwise, why would Term 2.4 run fine with CPUblit
- once and not again unless I reboot the machine?)
-
- (Incidentally, I'm not using an overscanned screen. REPEAT, NOT using
- overscan.)
-
- (Ranting over.)
-
- Rob Poole
- pooler@rpi.edu
- pooler@cs.rpi.edu
-