pam src rpm replaced?

Mike A. Harris mharris at redhat.com
Sun Aug 24 16:03:31 UTC 2003


On Sun, 24 Aug 2003, Leonard den Ottolander wrote:

>> These type of fixes are very trivial and are often things someone 
>> will never find in real world usage, generally only via specific 
>> testing to locate such minor bugs.
>
> Well, obviously they are real world usage examples, because they get 
>found, reported and fixed.

Some are, yes.  And some are found by trying installations 
designed specifically to try and find lurking trivial bugs.  Both 
are fine as I said, feel free to report the bugs.  But they're 
not all things that people will really hit in real world 
situations where they're not purposefully trying to find them.  
;o)


>> In other words, if someone makes them known, they're worth 
>> fixing, but IMHO they're not worth us spending a lot of our own 
>> time to track down as they don't really cause problems in general 
>> for the majority of users.
>
> That I can understand. Like most bugs they get exposed by using the 
>system, not by actively looking for them.

Indeed.  There are literally thousands of things we could 
actively try to search for and fix.  It wouldn't be efficient to 
do so however, and it would use up valueable and limited 
engineering resources for little gain, when those resources could 
be allocated elsewhere for bigger-bang-for-the-buck stuff.

That said, we welcome those who want to track down nickel and 
dime bugs and report them too.  Even better with fixes attached.  
;o)

For the record, I leave trivial stuff build up for a while, and 
then go on a "trivial bug fix day" day every now and then when 
things are NOT very busy.  That way I can fix 5, 10, 20, etc. 
trivial bugs all at once quickly and be done with them, and not 
be doing it when I have to rush to meet a deadline or cram 
important stuff into builds to meet a tree freeze or something.

I assume others do that too, which also helps explain why trivial 
things don't necessarily get fixed ASAP.  It's more efficient to 
fix a whackload at once when it's convenenient and when pressing 
matters aren't on the table.


-- 
Mike A. Harris     ftp://people.redhat.com/mharris
OS Systems Engineer - XFree86 maintainer - Red Hat





More information about the fedora-test-list mailing list