threshold for away mailings

[Michael McAndrew]

Hello people,

Any reason why the bounce threshold for away is 3 before a contact is put on hold?  That seems a bit quick to me.  I would say that 10 is a more sensible level given that, as far as I know, this isn't really a bounce.

or how about setting it at 'infinity', given that this isn't actually a bounce?

Michael

[Piotr Szotkowski]

Any reason why the bounce threshold for away is 3 before a contact is put on hold?

I think it might’ve been conservative thinking (if we have a 10k mailing and 2k people are away three times, we’d rather not burden ourselves with mailing them), but I agree it doesn’t make much sense to me.

or how about setting it at 'infinity', given that this isn't actually a bounce?

Patches welcome (I think ‘infinity’ handling will need a code fix). The default table contents are in xml/templates/civicrm_data.tpl.

[Michael McAndrew]

Hey, this forum thread turned into an issue with patch http://issues.civicrm.org/jira/browse/CRM-6405

30 seems like a decent approximation to infinity in this case.

Although (and I don't know the answer to this because I'm ignorant of the bounce processing calculation) is there a time limit set when calculating whether a contact reaches the bounce processing threshold?  i.e. is it '30 bounces in the past [time period]' or '30 bounces (ever)'.  If it is the latter, then it is conceivable that someone could reach this limit within a year or so, depending on how much they were away.

Interested to hear thoughts on this...

Michael

[Piotr Szotkowski]

Hey, this forum thread turned into an issue with patch http://issues.civicrm.org/jira/browse/CRM-6405

This is exactly the thing we want to encourage; thanks for leading the way, Michael!

30 seems like a decent approximation to infinity in this case.

civicrm_mailing_bounce_type.hold_threshold is an UNSIGNED INT(10), so for ‘true’ ∞ you might want to go with 4294967295.

Although (and I don't know the answer to this because I'm ignorant of the bounce processing calculation) is there a time limit set when calculating whether a contact reaches the bounce processing threshold?

Yeah, it’s actually set on a per-recipient basis and held in civicrm_email.reset_date; I think the date is, um, reset on every successful delivery.

[Michael McAndrew]

So then I think that 30 is probably a decent threshold.  Because if you receive 30 consecutive I am out of the office notices, then we should put them on hold IMO.

Not sure whether submitting a patch that adds a single character is actually worth submitting as a patch, but hey it's a start

[Piotr Szotkowski]

So then I think that 30 is probably a decent threshold.  Because if you receive 30 consecutive I am out of the office notices, then we should put them on hold IMO.

Agreed.

Not sure whether submitting a patch that adds a single character is actually worth submitting as a patch, but hey it's a start

No, you have it backwards – the smallest patch that actually fixes an issue is the most welcome. Great work.

[Michael McAndrew]

Any way you can work out if I get the prize for smallest patch ever?  I am guessing I am probably joint winner with a few others.  Perhaps a special prize for me given that my character was a '0'?

[Piotr Szotkowski]

Any way you can work out if I get the prize for smallest patch ever?  I am guessing I am probably joint winner with a few others.  Perhaps a special prize for me given that my character was a '0'?

Oh, definitely; I think a couple¹ of drinks should be a good start towards a proper prize. (As we can’t do deliveries, you surely agree that it’s only reasonable for you to come for your prize to Warsaw.)

¹ hm, maybe the number should we somehow related to the patch itself…