[bind10-dev] [sprint planning] tentative estimate results for sprint ending 2012-11-20

[JINMEI Tatuya / 神明達哉]

At Mon, 05 Nov 2012 10:22:55 +0100,
Jelte Jansen <jelte at isc.org> wrote:

> 2450:
> 
> ranges from 1 to 5; is this more than just adding a few lines about
> the possible error in a document somewhere?

According to my local copy I gave it a 5, but I don't remember why I
thought so:-)  Maybe I thought it included figuring out specifically
in which condition it happens (btw, looking at the ticket now, I've
realized I don't understand what this means: "We should take the error
that we get when this is not done").

> 2394:
> 
> ranges from 1 to 5; I'm one of the fives; this was under the
> assumption that it would also mean going through existing uses of the
> exceptions and changing them if necessary. Or is the intent to only
> add a bit of clarification to the doxygen?

I intended:
- (maybe after discussion) update documentation of Unexpected
- unify InvalidParameter and BadValue. update both doc and code for that

> 2423:
> 
> ranges from 3 to 13 (!). But this may not be too strange, since it's
> about a segmentation fault, and tracking those down can be really
> simple, or it can take days.

I thought this is difficult and would basically take time.  It only
happens on a specific system (so depending on who takes on it, we may need
to count the time for setting up an environment for reproducing it).
And, the stack trace doesn't seem to indicate the direct cause of the
error.  It may or may not turn out to be a simple bug, but based on
what we know I have to err on the side of caution.

Same for 2424.

> 2425:
> 
> Not a segfault, but a similar weird issue where the cause is unknown
> as of yet.

non-reproducible problems are generally difficult (and take time).
And the failure doesn't seem to reveal the direct cause.

---
JINMEI, Tatuya