Path: utzoo!attcan!uunet!deimos!ux1.cso.uiuc.edu!brutus.cs.uiuc.edu!psuvax1!news From: ehrlich@cs.psu.edu (Daniel Ehrlich) Newsgroups: comp.sys.ibm.pc.rt Subject: Re: Risc System/6000 Message-ID: Date: 21 Feb 90 16:44:11 GMT References: <90051.101319ANKGC@CUNYVM.BITNET> <24325@princeton.Princeton.EDU> <20635@netnews.upenn.edu> Sender: news@cs.psu.edu (Usenet) Distribution: na Organization: Computer Science Department, Penn State University Lines: 49 In-Reply-To: bradley@grip.cis.upenn.edu's message of 20 Feb 90 23:08:52 GMT In article <20635@netnews.upenn.edu> bradley@grip.cis.upenn.edu (John Bradley) writes: John> Yes, but has anybody actually PLAYED with these things? I attended the John> Philadelphia unveiling and got to try out a 320, a 530, and an X terminal. Yes. We have one here in the Computer Science Department. John> All three of them exhibited X performance that ranged from mediocre to John> sluglike, depending on the operation. What IBM told us when we complained about the performance of the X software was that there is a large amount of debugging code that is still in place but will be removed by the time the product really ships. Wink, wink, nudge, nudge. ;-) John> In particular NONE of the three beasties could move bits to/from the screen John> quickly enough. (as in, displaying/refreshing an image, and scrolling a John> region) Actually we can make the X server crash on demand my moving the mouse quickly while keying the buttons in rapid succession. John> The scrolling was particularly bad, mainly because it's the thing you're John> most likely to be using. I was unable to scroll an 80x24 xterm (say a John> resolution of approx 600x350 (or so)) at much better than 2-3 lines per second. John> This is right up there with a color Sun running the original cfb code from John> X11R1... John> While I can think of a host of reasons why it might be slow on the 320 and John> the 530 (lots of annoying background jobs, or something), the Xterminal was John> just as slow. I killed all other clients that were going to the Xterm, and John> it STILL couldn't scroll the screen at an acceptable rate, making me think the John> thing is utterly useless. Not to mention exspensive. The ~$US 2K price IBM is touting is without a monitor! John> Anyhow, I can't imagine HOW IBM could make the X server SO slow. The X11R4 John> server for my RT 125 Megapel is MUCH faster... It is all that debugging code. ;-) John> Or, perhaps the machines here were set up by complete bozos... John> John Bradley - University of Pennsylvania - GRASP Lab -- Dan Ehrlich Voice: +1 814 863 1142 FAX: +1 814 865 3176