[j-nsp] MX480 troubles.

[Justin M. Streiner]

On Wed, 13 Apr 2011, Chris Evans wrote:

> It seems like alot of folks run bleeding edge code with some if these major
> bugs popping up.I also get the impression that a lot of shops don't test
> code before they deploy.

That depends on your definition of 'test' :)  Seriously, though.  I've 
seen shops do a "does-it-boot" test, some very impressive regression 
testing, or somewhere in the middle.

People have to strike a balance between the number of person-hours spent 
regression-testing a new code release and the impact of tripping a bug 
that snuck through your testing.  Many places I've seen don't have the 
ability to perfectly replicate their production environment in their 
testing lab, so the testing becomes about having a reasonable level of 
confidence.

I'll agree that bleeding-edge code should be avoided in production, 
however there are cases for some people where there isn't a choice. 
Support for a new piece of hardware or a new feature that suddenly 
becomes critically important is one example.  Another could be where a 
bug fix is needed, and that bug is only being fixed (initially) in newer 
code releases.

Sadly, we've all probably been there... :(

jms