Mailing list spam - the many dnsapi patches that were mailed out

[Jeremy White]

So Zeb's evil plan to break my mailing list script worked!

The moderately fragile bit that parses the patches relies on a regex 
matching '^From '.  She included a comment in her code referencing 
another patch; a 'From <sha> here'.

That caused it to crash after sending 3 out of 4 emails, and before it 
marked that MR as done.  So every time it woke up, it thought it needed 
to send that MR, would send 3, and then crash.

I have, in theory, mitigated that with this change:
   https://gitlab.winehq.org/winehq/tools/-/merge_requests/15

Sorry for the over flowing mailbox.  Hopefully this patch set will get a 
*very* thorough review :-).

Cheers,

Jeremy

[Jin-oh Kang]

On Fri, Jun 24, 2022, 12:32 PM Jeremy White  wrote:
> So Zeb's evil plan to break my mailing list script worked!

😱
> The moderately fragile bit that parses the patches relies on a regex
> matching '^From '.  She included a comment in her code referencing
> another patch; a 'From <sha> here'.

Doesn't the mbox format auto-escape it as `>From `? Maybe it's in another
format or already parsed...
> That caused it to crash after sending 3 out of 4 emails, and before it
> marked that MR as done.  So every time it woke up, it thought it needed
> to send that MR, would send 3, and then crash.
>
> I have, in theory, mitigated that with this change:
>    https://gitlab.winehq.org/winehq/tools/-/merge_requests/15
>
> Sorry for the over flowing mailbox.  Hopefully this patch set will get a
> *very* thorough review :-).

Maybe I'm too late to the party, but a better way to fix this might be to
remember the delivery state of each patch. Just a suggestion though; feel
free to dismiss it 😅

Also, since you particularly mentioned a *very* through review, a nit: in
Python we usually use `is None` instead of `== None`. Both works, though.
> Cheers,
>
> Jeremy

[Jin-oh Kang]

In general, a separate wine-patches list might help lessen the impact of
future incidents.

[Charles Davis]

On Fri, Jun 24, 2022 at 1:46 AM Jin-oh Kang  wrote:
> In general, a separate wine-patches list might help lessen the impact of future incidents.

We used to have one. It was folded into wine-devel a while back.

[Jin-oh Kang]

On Fri, Jun 24, 2022, 3:55 PM Charles Davis  wrote:
> On Fri, Jun 24, 2022 at 1:46 AM Jin-oh Kang 
> wrote:
> > In general, a separate wine-patches list might help lessen the impact of
>
> future incidents.
>
> We used to have one. It was folded into wine-devel a while back.

Yeah, I was specifically referring to that. I think revival of that list
has been brought up in the GitLab experiment thread before by Alex Henrie.