------- Additional Comments From igodard at pacbell dot net  2004-11-24 22:12 
-------
Thank you for the excellent explanation, but your explanation was redundant; I 
do (and did) understand why the nature of the problem is opaque. I'm not filing 
this bug because I don't understand; I'm filing a QOI report because the 
initial 
message wouldn't leave the average user with an understanding.

The actual compiler message in such a case would be much better if it contained 
a brief explanation of the situation like the one you've given here, including 
what the possibilities are, why you don't know, and what to do to narrow down 
the problem on a generic operating system. I am *not* saying you have to fix 
the 
OS; I do feel that you *should* tell the user everything that you know, 
including that you don't know.

This is a situation where message terseness is not a virtue. The existing 
message is confusing and unhelpful, even to will understand the problem and 
what 
to do. I suggested an alternative phrasing; but anything would be better so 
long 
as it explains:

1) was terminated by the OS
2) could have been out of a resource, killed by user request, did a nasty (i.e. 
segv), or something else OS dependent.
3) a generic suggestion for how to find out what was the actual cause
4) how to file a report is nothing else works.

gcc has a generic problem with its messages: they assume the user is 
intimiately 
familiar with the host, the language, and how the compiler works. This makes 
lousy messages for your actual user community.  This is another case of poor 
message QOI.

May I suggest that you circulate this report and its discussion among your 
collegues and see what they think about the general message philosophy and this 
message in particular. I'm giving up here and won't re-open this again, but the 
QOI issue won't go away just because I do.

Ivan

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |UNCONFIRMED
         Resolution|WONTFIX                     |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18639

Reply via email to