[PATCH v3 10/15] crypto: index encrypted parts when indexopts try_decrypt is set.

[David Bremner]

Daniel Kahn Gillmor <dkg at fifthhorseman.net> writes:

>
> "index.auto.decryption" has at least two different meanings at first
> blush.  For example, it might mean "this message was automatically
> decrypted", which isn't necessarily the case.

I think in either case you need to document what the properties
mean. I don't think "index-decryption" by itself is unambiguous either.

> second confusion:
>
> Even if we used a less ambiguous term like "flush-on-reindex", it would
> seem weird to need to say "notmuch search
> property:index.flush-on-reindex.decryption=failure",

Yes, you seem to envision interactive searching for these
properities. I'll have to take your word for it. I tend to think about
properties as more like APIs than as end user facing.

> We could maybe avoid both of these confusions by dropping "auto"
> altogether and defining the following rule:
>
>  * properties prefixed with "index." are set during indexing and will be
>    cleared (and possibly re-set) upon reindexing
>
> wdyt?
>
> If you're ok with that, i could try to update the series to work that
> way (and include documentation along those lines).

I don't object to this.