[FreeBSD-Announce] FreeBSD bug tracking moves from GNATS to Bugzilla

Eitan Adler lists at eitanadler.com
Tue Jun 3 20:50:54 UTC 2014


On 3 June 2014 13:41, John Marino <freebsd.contact at marino.st> wrote:
> On 6/3/2014 22:06, Stephen Hurd wrote:
>> Matthias Andree wrote:
>>> Surely there may be valid and sometimes useful comments by people who
>>> aren't the original reporter, but for the common case that the bug
>>> evolves into a discussion between a developer and a reporter having a
>>> valid BIDIRECTIONAL communications channel up front helps a lot.
>>
>> I would venture that docs bugs and certain classes of website bugs
>> rarely have communications.  For example:
>>
>> Those bugs would simply not have been submitted if I had to create an
>> account first.

This is a legitimite concern.  This is exactly why we have anonymous
bug reporting on at the moment.  On the hand we're *already* seeing
tens of spam per hour via this exact mechanism.


> If the submitter's interest in getting the bug known in order to be
> fixed exceeds the distaste for registration, then yes, the report would
> still get submitted.

This is true.

>   (For example, if that person *really* wants to
> access FreeBSD forums via IPv6 and nobody knows it's busted, they would
> probably bite the bullet).

There are also other mechanisms.  For example, mailing lists.

> If nobody else that has an account already bothers to report it, it must
> not really be that big of an issue.

This is the wrong rhetoric to be using here.  We explicitly *don't*
believe this is true.

The concern from bugmeister's side is that we want to make sure that
the bug database is *useful*.  We know that some low-effort
no-reply-needed bugs come in.  This is true for both ports and doc.
It is slightly less true in src but not non-existent in source.
On the other hand the question is does this value outweigh the the
value of knowing that the submitter has a valid email address.

>> I strongly support anonymous bug reporting, but I'm not interested in a
>> protracted discussion on it.  The comitters can decide if they want
>> trivial bugs reported or not as they prefer.

At the moment this is absolutely enabled.  On the other hand we will
be closely watching these types of bugs.  I'm not sure we have
meta-data to track the source of an account (anon-report, or not) but
we will soon.

> I don't concede killing anonymous means killing trivial bug reporting,
> but if that was the case: Oh well, I guess we have to focus on
> non-trivial bugs.

Trivial bugs are important too.

-- 
Eitan Adler


More information about the freebsd-ports mailing list