Larry Hall
Fri Jan 9 16:59:00 GMT 2004

At 12:33 AM 1/9/2004, Christopher Faylor you wrote:
>On Thu, Jan 08, 2004 at 05:33:23PM -0800, Brian Dessent wrote:
>>"H. James de St. Germain" wrote:
>>>at the time...  Its curious that it works in either place and even more
>>>curious that placing it after cygwin starts causes programs like rxvt
>>>to core dump.
>>Why should that be curious?  As Chris pointed out it has to be set
>>before cygwin1.dll initializes.
>It certainly shouldn't SEGV if it is set in an rc file but it is curious
>why someone would comment on the curious nature of something that was
>already acknowledged by Larry as a potential bug.

Yeah, I know it's going out on a limb to suggest that a segv is a bug.
I'm a bit radical when it comes to things like that.  Any segv I tend 
to call "bugs", whether I find them curious or not.  I'm also of the 
notion that all "bugs" should segv.  That would make the job of finding 
and fixing "bugs" allot easier, right?  So when you write some code that 
will cause a "bug", please do so in such a way that it will always segv. 
This will streamline the bug reporting and resolution process immensely.

(Newly self-appointed Cygwin bug standard's officer) 

Unsubscribe info:
Problem reports:

More information about the Cygwin mailing list