Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.1 6/24/83; site utah-gr.UUCP Path: utzoo!watmath!clyde!bonnie!akgua!sdcsvax!dcdwest!ittvax!decvax!linus!philabs!pwa-b!utah-gr!donn From: donn@utah-gr.UUCP (Donn Seeley) Newsgroups: net.flame Subject: Re: dbx Message-ID: <1344@utah-gr.UUCP> Date: Mon, 11-Feb-85 20:41:44 EST Article-I.D.: utah-gr.1344 Posted: Mon Feb 11 20:41:44 1985 Date-Received: Mon, 18-Feb-85 04:15:04 EST References: <160@umich.UUCP> Organization: University of Utah CS Dept Lines: 42 > *** REPLACE THIS LINE WITH YOUR MESSAGE *** Why should I if you refuse to do it yourself? But I will anyway. > <<>> > > What !(#&^$% foisted this piece of trash on us? A fellow named Mark Linton, originally at Berkeley, now at DEC Western Research Labs and Stanford, wrote 'dbx'. It appears to have developed out of the P-code debugger 'pdx'. > ... Use it on a program of any size and it bombs miserably. Funny, I use it on the f77 compiler and it works fine. The f77 binary is about two-thirds the size of the kernel (or gemacs, another notoriously large program). Do you compile with the '-g' flag? I've noticed that dbx dies in a couple situations (e.g. when you try to evaluate an un-normalized floating point value or try to analyze a core dump from a program that trashed its stack) but I haven't found it unusable by any means. > And is sdb still around? Fat chance. On my system sdb is in /usr/old, source in /usr/src/old. I'm pretty sure it was distributed that way. I won't cry at the loss of sdb. There are a few features I miss like context search through source, but recent (post-4.2) versions of dbx have this. The bugs in the way dbx handles f77 binaries are mostly f77's fault (sigh). (The ones I know about have been fixed.) > Why don't people thoroughly debug a piece of software BEFORE > distribution instead of after? Where would all the fun be then? This IS net.flame, after all, Donn Seeley University of Utah CS Dept donn@utah-cs.arpa PS -- If you want buggy software, read net.sources... (I've seen things in net.sources that would make 'lint' run away, screaming.)