Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site lanl.ARPA
Path: utzoo!watmath!clyde!burl!ulysses!bellcore!decvax!decwrl!pyramid!pesnta!phri!cmcl2!lanl!crs
From: crs@lanl.ARPA
Newsgroups: net.internat
Subject: Re: "Bad magic number"
Message-ID: <104@lanl.ARPA>
Date: Wed, 22-Jan-86 09:09:18 EST
Article-I.D.: lanl.104
Posted: Wed Jan 22 09:09:18 1986
Date-Received: Fri, 24-Jan-86 21:37:46 EST
References: <515@cheviot.uucp> <51400005@uokvax.UUCP> <486@ihdev.UUCP>
Organization: Los Alamos National Laboratory
Lines: 35

> >I'd say one should take advantage of the opportunity and do a French version
> >of what the error message should have been all along, i.e. something like
> >"not a .o file for this processor." 
> 
> Aaah, but that does not cover the scope of errors covered by `bad magic
> number' which can also tell you that you are trying to load a library
> module from a file that is not a library and various other things.

Perhaps 'Bad magic number' is trying to do too much then.  Isn't it
possible to distinguish among these errors?

> `Bad
> magic number' says it all for me,  but what I would like to see instead
> is an on-line program to explain errors (perhaps in more than one
> language).

Sounds a reasonable approach but then the basic error message can be even
more cryptic than 'Bad magic number'; how about 'Error number 1492' as they
do with the micros?  How about an environment variable or the like to say
whether a user wants the error messages automatically run through this "on
line program" for translation to human language?  And, perhaps, another to
specify the language?

> >If one of the results of international-
> >ization is the elimination of cretinous messages like "bad magic number,"
> >it deserves support from even the most isolationist of users.

Absolutely!
-- 
The opinions expressed are not necessarily those of my employer,
the government or your favorite deity.

Charlie Sorsby
...!{cmcl2,ihnp4,...}!lanl!crs
crs@lanl.arpa