Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!yoyo.aarnet.edu.au!myall.awadi.com.au!myall!blymn From: blymn@mallee.awadi.com.au (Brett Lymn) Newsgroups: comp.os.386bsd.misc Subject: Re: NetBSD+FreeBSD merge urged! Date: 24 Sep 1993 00:21:04 GMT Organization: AWA Defence Industries Lines: 68 Message-ID: <BLYMN.93Sep23182109@mallee.awadi.com.au> References: <1993Sep18.204206.29629@diana.ocunix.on.ca> <CDn8xn.33q@sztaki.hu> <1993Sep20.213150.13549@diana.ocunix.on.ca> NNTP-Posting-Host: mallee.awadi.com.au In-reply-to: db@diana.ocunix.on.ca's message of Mon, 20 Sep 1993 21:31:50 GMT >>>>> On Mon, 20 Sep 1993 21:31:50 GMT, db@diana.ocunix.on.ca (Dyane Bruce) said: Dyane> In article <CDn8xn.33q@sztaki.hu> pink@fsz.bme.hu writes: >In article 29629@diana.ocunix.on.ca, db@diana.ocunix.on.ca (Dyane Bruce) writes: >>In article <Ygam_WG00WBME1btYl@andrew.cmu.edu> "Alex R.N. Wetmore" <aw2t+@andrew.cmu.edu> writes: [..snip...] Dyane> DOS will always be a dog to run under unix... Not true. On a commercial unix with a commercial dos-under-unix I could run four dos tasks at once and still be comparable with a 386/25 running DOS (my machine was a 486/25) Dyane> DOS has no concept Dyane> of being to do things in the background. No concept of asynchronous I/O Dyane> completion i.e. it spin loops waiting for I/O completion, Dyane> waiting for keyboard ready is brutal. True but busy waits can be detected and the process priority lowered. Dyane> Still, it can be very usable Dyane> for running the "dusty deck" old Miserable DOS program, or Dyane> the occasional word processing. Again, on a commercial unix, I *preferred* using Microsoft Word on a dos-under-unix to using it on a 286 machine because the d-u-u process was *much* faster. Dyane> You aren't going to compile Dyane> anything anyway under DOS... You'd be using the native unix. Wrong... just think of this: using the unix development environment, i.e. your favourite editor (vi/emacs/whatever), awk, sed, whatever. Doing your software build using a unix make file that fires off a dos process to do the compile and link. I have done this, basically to avoid having to deal with some compiler author's idea of an integrated development environment (Wordstar command keys for doing cut/paste...gimme a break!) AND their brain-dead/naive idea of what a make should do (like regenerate dependencies *every* time it makes ... life is too short). By running d-u-u and being able to use the unix tools I am familiar with I won in terms of being able to get on with the job instead of struggling with a concept on an editor that hails from the early '80's. Dyane> Re-writing the virtual BIOS can help a lot, as anytime a BIOs Dyane> trap is taken you can suspend the virtual DOS... Hmmm I think you would have problems with this concept, some DOS things still expect timely events... Dyane> It is going to take a decent DOS/Windows emulator to wean people Dyane> away from an program loader system of the 60's to a real Dyane> operating system, of at least the 70's :-) Ooops, I sugest a dose one of the folk-lore newsgroups for you... unix precedes DOS by a good margin, in fact if you reverse your 60 and 70 you would be closer to the mark. Dyane> Of course, the dreaded legal challenge is always possible.... There Dyane> is a legal case on in the states that might enable MS to Dyane> make WABI illegal.. (As they can argue that the Windows API can't Dyane> be copied.) Sun is not small fish so Microsoft may have their hands full trying to win that case. It will be interesting to watch. -- Brett Lymn