From owner-freebsd-hackers Tue Jul 18 23:52:44 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.11/8.6.6) id XAA09154 for hackers-outgoing; Tue, 18 Jul 1995 23:52:44 -0700 Received: from saul4.u.washington.edu (saul4.u.washington.edu [140.142.83.2]) by freefall.cdrom.com (8.6.11/8.6.6) with SMTP id XAA09148 for ; Tue, 18 Jul 1995 23:52:42 -0700 Received: by saul4.u.washington.edu (5.65+UW95.05/UW-NDC Revision: 2.33 ) id AA16361; Tue, 18 Jul 95 23:52:30 -0700 X-Sender: spaz@saul4.u.washington.edu Date: Tue, 18 Jul 1995 23:52:29 -0700 (PDT) From: John Utz To: Ed Hudson Cc: hackers@freebsd.org, elh@spnet.com Subject: Re: FreeBSD Killer Apps (was Re: TCL vs...) In-Reply-To: <199507181651.JAA05380@p54c.spnet.com> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: hackers-owner@freebsd.org Precedence: bulk Hello Ed ! Really nice to hear from you on this subject! On Tue, 18 Jul 1995, Ed Hudson wrote: > > howdy. > > i'm not certain this is appropriate for > this news group... Damn right it's appropriate! it is a mailing list, tho... umm unless this is gated onto usenet.. :-) > i build ic cad applications (original work), and > use FreeBSD as my primary development platform... YIPPEE! it is great to here that stuff near and dear to my heart is being done on freebsd... > sadly, all of my distributions (sales) are to IRIX and > SunOS platforms (to date)... i'm trying to talk a major > cad vendor into making their primary product available > under FreeBSD... the answer is (so far) 'when we > can get FlexLM support, then maybe we'll proceed...' > > FlexLM is a license manager/library application > that most IC cad vendors use. as such, it maybe > a critical 'sub-application' Yes, i think u are hitting a *really* important point here! It is these kind of really boring, obnoxious little detail applications that are probably key to gaining commercial acceptance. It has been my experience that commercial ee type applications have always been at bleeding edge of software copy control, at least in the dos realm. The first time i ever heard of a "dongle" ( those little doolybobs that plug into the parrallel port and allow the software to work ) was in conjunction with orcad or tango ( commercial circuit design and pcb layout tools ). I think the high cost and high utility of these packages made the temptation to bootleg them well nigh irresistable ( why do u think i found out what a damn dongle was? :-) ) I can also state with confidence that verilog's mom and dad are very heavy into license management. On the ee network this spring we had a major crisis when the number of students with verilog assignments due the following monday exceeded the number of legal seats... a major bummer was had by all... > i know of at least one vlsi design startup here > in silicon valley that uses NetBSD for running > significant logic simulation regressions on a > large array of pentia's (private verilog2c) > (i know of a couple of other startups considering a > similar methodology - i'm trying to steer them to > FreeBSD). > > i know of several people using FreeBSD boxes as > X-terminals for isdn links to their work places > (again, sadly, i know more people using linux > for this, so far... but hope this will change > as people try to actually run local compiles instead > of just X - they'll want a more SunOS like environment) I think this is true, I can point out that i use -DSUN as a compilation flag first choice when i come across something ( usually ee applications oddly enuf :-) ) that has been ported to the usual suspects and linux, but not net or freebsd. I got a linux-pusher ( i am sorry about the potential flamish verbage, but it is a fact, i find a noticeable number of linux users seem to think that if u are running a pub os other then linux, then u must be ignorant or stupid ) really pissed off when i pointed out to him that when i attempted to take a linux port of the ic design package magic over to freebsd it took 3 weeks of fruitless labor before i got in touch with the developer and found the original source, which i had compiled in two days and running in a couple of days after that ( this was on 115 0r 20, i forget which ) > FreeBSD is just missing a couple of key applications > to be generally usefull as an ic-development environment > commercially (ie, commercial use of public domain, > or near-public-domain (university) (eg, spice3) tools). Umm, what exactly are u saying here? I read u as saying that the tools dont exist? Then u mention that it does? There is spice3f4 ( and there is an amazingly stupid hack that u can only find if u debug it to find the comment that tells u what environment variable to create to prevent it from dumping core ). There is magic, which is a full fledged real estate editor. It is slightly less then perfectly optimized, however. My personal favorite is the chipmunk tools. They completely rock!!! It is a collection of digital and analog drag and drop graphical design and simulation tools. I f*n love it! I predesign my circuits for school and then print the schematic out via ghostcript. > in particular, a good public-domain verilog is needed > the most (i understand that such an effort is underway > at stanford?) There is probably several various efforts in several various locations :-)... The one i am most familiar with is not currently functional with FreeBSD only because *i* can't get SWI-prolog to save any worlds! grr! This is a problem with the unique nature of prolog in general ( a prolog "program" appears to not be a neat, tidy sequence of instructions. it appears to be a literal copy of the address space in memory it was using at the time it decided it was finished doing something. i dont really know. i do know that SWI-prolog compiles and that Prasad's vhdl parser compiles, but that nothing can happen because it cant find anything ( sigh ). This is one of my big projects after i finally get my box recompiled with dlmalloc.... > for many of the univeristy tools, ports exist, but > there's no central distribution available. for many > of these university tools in the USA, such as spice3, > there are hold-over export restrictions that are about > as bad as DES. Umm is this a clarification of your previous statement? For us us folk it is not that big of a problem, eh? I dont think it gets in the way of for profit use..... > -elh > anyway it is really great to hear from u about this. I love running all this stuff under freebsd, and the output from chipmunk was the first thing i was able to show anybody over at the uw college of ee that ever sparked any interest from my fellow classmates ( well , that *and* the ability to run verilog over the phone and get a display up at home! ) let me know what u think! ******************************************************************************* John Utz spaz@stein.u.washington.edu idiocy is the impulse function in the convolution of life