Re: Future of gccbug

[Daniel Berlin <dberlin at dberlin dot org>]

On Nov 7, 2003, at 5:45 PM, Gabriel Dos Reis wrote:

> Daniel Berlin <dberlin@dberlin.org> writes:
>
> | > Also, bugzilla changes the message-id, also a pain.
> | >
> | It has to in order to make it do what it does now.
> | It's the way it constructs message ids that let's it thread.  It
> | doesn't store the message ids of messages it's sent out (and this
> | wouldn't help, since we don't know what comments are replies to other
> | comments).
>
> It could just let through the References: contained in the messages it
> gets.
I'll see what i can do, i probably won't get to it till after the merge 
this weekend.

>   Precisely because you don't let those references get through,
> the recipients don't know which messages are
> repliers to other.  It
> becomes a pain to follow.

I'll see about adding it to the existing reference line if there is one.
What are references supposed to be seperated by? Single spaces or 
commas?

> | This is fairly common in bug-tracking systems. In fact, i'm not 
> really
> | aware of one that does threaded messages in a WWW interface.
>
> Please WWW is not all about bug submission, reading and tracking.
> Consider that other developpers use more mail agents than WWW
> interfaces.

I didn't say it was, i'm saying it's difficult, and that i'm not aware 
of anything that *does* do it.

Please stop assuming that because i argue about something, i won't do 
it or think i'm right.  I'm simply stating facts.