[BUG] net_cls: Panic occured when net_cls subsystem use

jamal hadi at cyberus.ca
Sun May 31 06:17:10 PDT 2009


On Sun, 2009-05-31 at 09:55 +0200, Jarek Poplawski wrote:

> Sure, after fixing it properly we should get rid of unneeded checks.

> > > Anyway, it's worked for other classifiers like this for some time...
> > 
> > Would you agree that it is/was a bandaid?
> > Or maybe you have some other fear that this may break something else and
> > prefer the workaround instead?
> 
> If somebody decided to do it this way instead of the "proper" fix then
> it looks to me more like a bandaid "by design". 

I think your and my definition of "proper" are at odds here ;->
My logic says there's a causality effect and you always fix the cause
in such a situation. 
Here's the anology of our conversation (some captured above) as i see it
centred around a bug of leaky pipe which just messed up the carpet
overnight in some room:
 
handyman Jamal: fix the pipe so it doesnt leak.
handyman Jarek: put a little bucket below the dripping spot
handyman Jamal: fixing this pipe is cheap - it is on warranty too
handyman Jarek: I fixed two other rooms by putting buckets
handyman Jamal: Proper: fix the pipe - you dont need the buckets forever
handyman Jarek: Proper fix is to put the bucket below any drip

Of course i am exagerating to make a point on our logical approaches:
leaking pipes dont quiet match software bugs i.e.
a dripping pipe will have the short term sense of emergency of needing
buckets but in this case the cost of the damage and time is the same
if you put a bucket or fixed the pipe.

> And, yes, I have some
> fear we could break some strange configs, which could depend on this
> wrong but working design.

To continue our discussion
handyman Jamal: What is your fear about fixing the pipe?
handyman Jarek: Someone may have plants which depend on the drips;
so if you fix it his plants wont get water anymore.

I hope the above doesnt offend you - it is meant in good spirit.

cheers,
jamal



More information about the Containers mailing list