From owner-freebsd-hackers Fri May 24 13:44:43 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id NAA25788 for hackers-outgoing; Fri, 24 May 1996 13:44:43 -0700 (PDT) Received: from kitten.mcs.com (Kitten.mcs.com [192.160.127.90]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id NAA25776; Fri, 24 May 1996 13:44:38 -0700 (PDT) Received: from venus.mcs.com (root@Venus.mcs.com [192.160.127.92]) by kitten.mcs.com (8.7.5/8.6.9) with SMTP id PAA13530; Fri, 24 May 1996 15:44:19 -0500 (CDT) Received: by venus.mcs.com (/\==/\ Smail3.1.28.1 #28.5) id ; Fri, 24 May 96 15:44 CDT Message-Id: Subject: Re: The view from here (was Re: ISDN Compression Load on CPU) To: terry@lambert.org (Terry Lambert) Date: Fri, 24 May 1996 15:44:18 -0500 (CDT) From: "Karl Denninger, MCSNet" Cc: davidg@Root.COM, dennis@etinc.com, jkh@time.cdrom.com, hackers@freebsd.org, freebsd-isp@freebsd.org In-Reply-To: <199605241946.MAA01597@phaeton.artisoft.com> from "Terry Lambert" at May 24, 96 12:46:38 pm X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >> Jordan's talking about other quality issues, like a poorly designed bus and >> rail design that allows for cards to become unseated fairly easily and the >> general lack of quality with 99.9% of the motherboards on the market. In this > > case, I'm refering to poor Q/A at the factory (bad cache ram, etc), poor > > BIOSes, etc. ...and then there's interoperability problems between various >> cards with various motherboards. This is one advantage that Sun machines have >> that PCs will never have - all the hardware is pretty much made by one vendor > > and this cuts down the interoperability/reliability problems dramatically. > > For those of us that come from minicomputer backgrounds (DEC PDP and VAX), > > it's all too obvious to us where the shortcomings are in PC hardware. > > On the other hand, I've had to reseat SBUS-connected TIGA boards on > AT&T/NCR X terminals on a 3-4 month rotation because thermal > expansion walks the right out. > > I personally prefer the MCA connector technology, but of course, we > "all" hate MCA. > > The Apple NuBUS cards and slot edge connections are also pretty > cool as far as good connector technology, but again, there's enough > play in the engineering that you can get impossible-to-install cards; > definitely a QC issue. The thing that really pisses me off about the PCs is the "edge-card" type of connector. Those things just don't hold their quality and connectivity over the long haul. VME-style connectors are a LOT more reliable. As in million-insertion reliable. Look at the backplane of a CISCO sometime (the 7xxx series) and you'll know what I mean. And yes, I know connectors aren't free. But I'd GLADLY pay $5/board more for connectors that HAD to mate 100% and couldn't bend the pins or jam during insertion. That would also mean that case and board tolerances would have to improve, or you'd end up with an impossible-to-insert card -- which will quickly drive the maker out of business. > All in all, the PC cards don't suck too badly if you use full length > cards and loking end-rails internally. With the screw on the other > end, the card isn't going anywhere. But almost none of them are full length now, and PCI makes the problem worse since they have shortened the "inserted" part of the board. EISA isn't too bad, as there is a LOT of contact area -- but PCI has the same problems ISA has, only worse, due to the small area of contact and poor tolerance control on the part of the component folks. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | T1 from $600 monthly; speeds to DS-3 available Voice: [+1 312 803-MCS1] | 21 Chicagoland POPs, ISDN, 28.8, much more Fax: [+1 312 248-9865] | Email to "info@mcs.net" WWW: http://www.mcs.net/ ISDN - Get it here TODAY! | Home of Chicago's only FULL Clarinet feed!