mc68020@nonvon.UUCP (07/10/87)
I am having major problems with the latest release of news 2.11.
To begin with, the version I am using has been patched to patch 8.
When the Makefile attemps to link checknews, the loader complains
that _line and _header are not found. In checking, it appears that
someone inadvertantly declared 'line' as an extern int. In checking
the two other source files used in making checknews, I found no other
references to it, so that fiz was easy.
The problem with _header is more difficult. It is defined in param.h
as struct. It is used as a struct in checknews.c, but is UNDECLARED.
How is it that such problems have survived past patch level 8?
(again, I flame about documentation: of course, he says, **IF** the damned
code was DOCUMENTED, so that it was possible to make some SENSE of it without
days of analysis, I might have been able to figure out who should declare
header.)
Solution?rick@seismo.CSS.GOV (Rick Adams) (07/11/87)
None of those problems exist in patch level 8. That's why no one else is complaining about them. Before ranting about the poor quality of the FREE software you have received, why don't you check and see where you screwed up. Because thats what has happened, YOU SCREWED UP. I am guessing that you screwed up in applying patch TWO, because that's where the makefile was changed to add rextern.o to the dependancy list for checknews. (Try doing a "make clean;rm -f Makefile.dst defs.h;make -f Makefile.dst". I'll bet it starts working.) I will resist the temptation to say that if you don't like the quality of something you have been given free, write something better. However, the quality of the newscode is actually better than a lot of the software we have paid for. (This does not mean that the news software is high quality stuff.) Welcome to the real world. That's the kind of program you usually get paid to work on. They all don't look like something out of a textbook. It would be nice if they did, but reality isn't like that. A hell of a lot of people have managed to fix little things like that without "days of analysis". It shouldn't have taken more than 10 minutes to find and fix. First you use grep to find the declarations, then you change the make dependency. (Of course this presumes you understand C programs and make.) I believe that the "Solution" you are looking for is to choose a good college after you have graduated from high school and learn some skills so you future employer can assign you to work on other peoples programs as well as your own. ---rick
karl@grebyn.COM (Karl A. Nyberg) (07/11/87)
In article <608@nonvon.UUCP>, mc68020@nonvon.UUCP (root) writes: > > I am having major problems with the latest release of news 2.11. ... > (again, I flame about documentation: of course, he says, **IF** the damned > code was DOCUMENTED, so that it was possible to make some SENSE of it without > days of analysis, I might have been able to figure out who should declare > header.) > > Solution? I think that sounds like an offer to dig in and write documentation that you feel would be sufficient for aiding in providing such solutions. That would be a worthy contribution in return for your use of the software. Last company I worked for was that way - if you spotted a problem and recommended a solution, you were responsible for implementing it. -- Karl -- DDN: nyberg@ada20.isi.edu INET: karl@grebyn.com - AKA - karl%grebyn.com@seismo.css.gov uucp: {decuac, seismo}!grebyn!karl
don@hnsurg3.UUCP (Don Lawrence) (07/12/87)
The quality of the FREE News software we all use is FAR BETTER than almost
ALL of the commercial software we've purchased here at MASSIVE COST over the
past 6-7 years. Heartfelt thanks again to Mark, Rick, et.al. Don't give
up because of a few Bigmouths; thousands of us look up to you for leader-
ship and are extremely grateful. We usually remain shut up. So I will.
--
Don Lawrence {decvax,cbatt,cbosgd}!hal!ncoast!hnsurg3!don
(This IS my employer's opinion !) seismo!uunet!hnsurg3!donphil@amdcad.AMD.COM (Phil Ngai) (07/12/87)
In article <4668@grebyn.COM> karl@grebyn.COM (Karl A. Nyberg) writes:
<Last
<company I worked for was that way - if you spotted a problem and recommended
<a solution, you were responsible for implementing it.
Oh, what a clever way to discourage people from making helpful comments.
--
Phil Ngai, {ucbvax,decwrl,allegra}!amdcad!phil or amdcad!phil@decwrl.dec.commangler@cit-vax.Caltech.Edu (System Mangler) (07/13/87)
In article <4668@grebyn.COM> karl@grebyn.COM (Karl A. Nyberg) writes: >Last >company I worked for was that way - if you spotted a problem and recommended >a solution, you were responsible for implementing it. In article <17480@amdcad.AMD.COM>, phil@amdcad.AMD.COM (Phil Ngai) writes: > Oh, what a clever way to discourage people from making helpful comments. Yes, "signing up" is an irksome expectation, and I wish Kidder's "Soul of a New Machine" had not popularized it. Another great one is insisting that something work the first time. This merely insures that problems will not be reported to management (at least until they've effectively killed the project anyway). Followups to comp.software-eng Don Speck speck@vlsi.caltech.edu {seismo,rutgers}!cit-vax!speck
michael@psy.vu.nl (Michael Felt) (07/15/87)
In article <17480@amdcad.AMD.COM> phil@amdcad.UUCP (Phil Ngai) writes: >><... a solution, you were responsible for implementing it. > >Oh, what a clever way to discourage people from making helpful comments. >-- Oh, what a clever way to make your own work schedule :->