310155 - mail_delivery_method should have a "temporarily disable" option

Status: RESOLVED WONTFIX

(Bugzilla :: Email Notifications, enhancement)

Description

[Gervase Markham [:gerv]]

It would be cool if the mail_delivery_method Param had an option to temporarily
disable mail - in other words, it didn't send mail but it also didn't update the
last-changed date. This allows you to make big changes to Bugzilla much faster.

Dave did this for the auto-unco stuff, so he should be able to say which bits of
code the if() test would be wrapped around.

It could be that future work on the mail system will eliminate the performance
problem; but it seems to me that not sending mail will always be quicker than
sending mail. And this enhancement would be pretty cheap and simple.

Gerv

Comment 1

[Gervase Markham [:gerv]]

Dave said:

"gerv: for the record, what I ended up doing was setting $enableSendMail = 0 at
the top of Bugzilla/BugMail.pm, then commenting out the SendSQL call that
updates lastdiffed."

Gerv

Comment 2

[Max Kanat-Alexander]

I generally disagree with implementing this. It will confuse admins, and lead
them to making mistakes. The best thing to do is to just queue mails and then
send them after the page processes, which should eliminate all latency from
sending emails.

Comment 3

[:glob ✱]

i'm with max on this one, it would be too confusing.

the real fix for this is bug 249121

Comment 4

[Gervase Markham [:gerv]]

Well... is bug 249121 one of those "wouldn't it be nice if" bugs that'll never
get done because it's a heck of a lot of work, or is there a simple way to do it?

Gerv

Comment 5

[:glob ✱]

> Well... is bug 249121 one of those "wouldn't it be nice if" bugs that'll never
> get done because it's a heck of a lot of work, or is there a simple way to
> do it?

just between you and me i've started working on it.
most of it's simple.

Comment 6

[Gervase Markham [:gerv]]

Adding a dependency; when the other bug is fixed, I'll mark this one WONTFIX.

Gerv

Comment 7

[Gervase Markham [:gerv]]

glob: is comment #5 a case of "famous last words"? :-)

Gerv

Comment 8

[Gervase Markham [:gerv]]

Looks like this did get implemented in bug 284184.

Gerv