usenet@cps3xx.UUCP (Usenet file owner) (05/14/89)
In article <2954@rti.UUCP> bcw@rti.UUCP (Bruce Wright) writes: >In article <2649@ssc-vax.UUCP>, coy@ssc-vax.UUCP (Stephen B Coy) writes: > 4) If they are comparing OS/2 using the Presentation Manager then I > expect it would be slower ... without significant hardware assists > it's hard for a graphics based system to compete with a text based > system. Kind of interesting thing. At work I use a Zenith I*Mclone, 8Mhz 8086. Text output on a two plane screen of the Amiga is very much faster than the clones. Heck, Amiga text output is even faster than the 12Mhz '286 clones in the office. Even with the hardware extras on the Amiga, graphics rendered text is a lot more complex than with a real text display. Makes me wonder, are the Amiga system programmers that much better than MicroSloth's? I think so, and for proof look at how slow in general AmigaBasic is (supposedly written in assembly), and in particular text output! Is there any chance that AmigaBasic could be replaced by some other language in future releases? Perhaps TrueBasic, or a C interpreter/compiler? Hell, A version of LOGO written especially for the Amiga would be a good choice too. One more random note: How about a new feature in the input.device, to support hot-key type programs? A program could request to be called when Left-Amiga-whatever is pressed. Intuition would register a hot-key handler when it puts up yes/no requesters for A-V, A-B; then remove it when the requester goes away. The request msg would have a priority field, and a flag for exclusive or shared access. Intuition would post its a-v, a-b request as shared, with a high priority, covering up any lower priority handlers. Somehow this method seems better than haveing a bunch of little input-handlers, each checking for one particular key-sequence. This is an example of a short .signature jap@frith.cl.msu.edu
wtm@neoucom.UUCP (Bill Mayhew) (05/15/89)
Most IBM-compatible display boards are a study in slowness. There is considerable penalty being stuck in an 8-bit slot on the XT interface bus. The maximum DMA transfer rate on the XT bus is less than 5 megabytes/sec. The tightly coupled architecture of the Amiga graphics chips is a substantial benefit. Curiously, my IBM model 80-171 (16 MHz) machine has 25 wait states on a video write according to the ATPERF progam. Given the piggyness of the display under windows 386 I'm inclined to believe that. You'd think IBM could have done better than that since that model 80 VGA controller is integrated on the motherboard. We recently switched some of our AT&T 6386s over to dual-port VRAM VGA boards with 16 bit I/O. Those cards are fast, in fact qualitatively faster than the Amiga even under windows 386. I haven't had the chance to ATPERF them yet. Bill wtm@impulse.UUCP
ejkst@cisunx.UUCP (Eric J. Kennedy) (05/16/89)
In article <2968@cps3xx.UUCP> porkka@frith.UUCP (Joe Porkka) writes: >Kind of interesting thing. At work I use a Zenith I*Mclone, >8Mhz 8086. Text output on a two plane screen of the Amiga >is very much faster than the clones. Heck, Amiga text output >is even faster than the 12Mhz '286 clones in the office. >Even with the hardware extras on the Amiga, graphics rendered text >is a lot more complex than with a real text display. It just depends on how much effort you put into it. If you were to simply type a file to the screen on the two machines, I don't know that there would be much speed difference. If you want to get fancy on the Amiga, and use some sort of warptext routine, then yes, you can get it blisteringly fast. But to go even faster on a character-based display, all you have to do is set up your screen in ram, and do a block move of 4 KB to to the correct memory location ($B0000 or $B8000 if I recall correctly). Moving 4 KB from one location to another is considerably faster (and easier) than anything that I've seen done on the Amiga, by the very nature of things. This method is not always used, for various reasons, not the least of which is to be compatible with multitaskers. (like, Lotus 123 prints the date and "Ready" overtop Deskqview's windows) -- Eric Kennedy ejkst@cisunx.UUCP