Re: Devin's effort

On 4/22/2015 11:25 AM, Tanstaafl <[email protected]> wrote:
> On 4/21/2015 10:20 PM, Bernie Maier <[email protected]> wrote:
>> So Thunderbird *does* attempt to index something from the entire
>> mail hierarchy. And I don't want it to do that.

> Actually, I'll go ask and confirm precisely how this (TB Indexing) works
> in both cases. I *think* it does some indexing of the headers and
> whatever is available, only because I know I can search a huge mailstore
> that is not sync'd offline in just a few seconds. I'll report back if I
> can get a definitive answer...

According to someone who should know (but I've asked if there is any
official documentation that fully explains indexing/Gloda), Gloda does
still index as much as it can even if full copies of the emails aren;t
there.

In my opinion this is a good thing. Indexes simply speed things up.

> That was my main point. You CAN disable GLODA completely, and still have
> very fast searching of all of the headers/meta-data (Subject,
> Sender/Recipients, etc). This is how I have TB configured now.
> 
> But, I'm now unsure of whether or not there is still some minimal
> indexing going on. If there is, I can tell you that you don't even know
> it is happening... or... hmmm... maybe it could be responsible for some
> of the memory/processor usage differences (as well as the perception by
> some here as it being 'slow and clumsy')...

Ok, so, apparently when Gloda is disabled, there is no indexing going
on, but as I said, searching on the headers meta-data, the results are
still really fast even on large mailboxes...

I just enabled Gloda (for the first time in... forever), and will see if
my searches speed up noticeably.

Also, there is a nice Addon called 'Glodaquilla' that allows you to get
very granular on what is and is not indexed:

https://addons.mozilla.org/en-us/thunderbird/addon/glodaquilla-search-indexing-en/