Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!gatech!ukma!psuvm.bitnet!cunyvm!nyser!mstr!pwa-b!philabs!linus!mbunix!bds
From: bds@mbunix.mitre.org (Barry D Smith)
Newsgroups: comp.ai
Subject: Re: Questions about Lisp boards (chips)
Message-ID: <54603@linus.UUCP>
Date: 25 May 89 16:25:17 GMT
References: <89May24.104032edt.11713@neat.ai.toronto.edu>
Sender: news@linus.UUCP
Reply-To: bds@mbunix (Smith)
Organization: The MITRE Corporation, Bedford, Mass.
Lines: 29

In article <89May24.104032edt.11713@neat.ai.toronto.edu> wang@ai.toronto.edu (Huaiqing Wang) writes:
>Does anyone have any experience with using Symbolic MacIvory Lisp boards?
>
>Specifically, we want to know:
>     - what the board's performance is like (e.g. is it slow?)
>     - what release (e.g. 6.0 or 7.2) are you running using the board?
>     - what are your overall comments on the boards.
>

We've been using a MacIvory since Beta test.  The board's speed is
roughly that of a Symbolics 3620, ie, slow, but Symbolics has recently
announced an upgrade that is about 2x faster.  In any case, i/o is
terrible since the ivory board uses 40 bit words and must pass them
over a 32 bit bus to the Mac.  This requires dissambling the words and
reassembling them at the Mac end.  Trying this with large bitmaps is
horribly slow.  Also, there is no real color support right now, and
the system is VERY sensitive to different color monitors (most won't
work).  The Ivory uses Genera 7.4I, which allows it to make use of the
Mac file system and Nu-bus.  On the positive side, the development
environment is far and away the best thing other than a Symbolics
machine.  Symbolics has implemented the entire Genera environment,
including dynamic windows, the window debugger, and the window
inspector.  Right now, the only thing that can compete is a Sun 4
running SPE 1.1(Beta) and Lucid 3.0.  (I haven't used Allegro
CL/Composer yet, so I can't make comments about it other than I
believe it doesn't have any direct color support, which Lucid 3.0
does.)

I hope this helps.