481211 - The order of words should not be taken into account in quicksearches like "Subject, Sender/from, or Recipient" (quickfilter)

Status: RESOLVED DUPLICATE of bug 522985

(Thunderbird :: Folder and Message Lists, defect)

Description

[Frédéric Buclin]

The "Subject or Sender" search field doesn't let me type "foo bar" and return all emails having "foo" and "bar" in them. It's rather looking for "foo bar" as is despite it's very unlikely that these words are besides each other (i.e. it won't find "foo bla bar"). Counter-intuitive, IMO.

Tested with Tb3b2:

Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1b3pre) Gecko/20090223 Thunderbird/3.0b2

Comment 1

[Thomas D. (:thomas8)]

Frédéric, imo what you mention here is one of the biggest problems (or unused potential) we will continue to have in TB3. What you suggest would really boost the efficiency of quickfiltering. Unfortunately, although this has been noticed long ago in Seamonkey bug 123788 (from 2002), so far this hasn't received a fraction of the interest it deserved. It's a major change to quicksearch logic, but I believe it's not even very hard to implement this - maybe you can investigate?

I filed the following bug for this, with detailed proof of concept (please vote for it):

Bug 522985 - [TB] Power Filter: Quick Search should filter for all-words-as-substrings, not phrase (aka Filter Revolution, faceted filter)

This idea is even more interesting given the current search pattern and UI complexitiy of "Search all messages", where you can enter multiple words from different spots of the mail, but you can only enter full words by default and there's a lot of other drawbacks (compared to comfy quick search filters) which I have sketched in Bug 383895, comment #24.

Comment 2

[Thomas D. (:thomas8)]

Frédéric, another thing you may wish to do in support of this is investigate if this suggestion has been made elsewhere (forums, newsserver, blogs...), or in the past (I could imagine there's some "wontfix" corpses or the like in bmo archives).

Comment 3

[Frédéric Buclin]

Why is my bug marked as a dupe of yours, instead of the opposite?? Doesn't make sense to me.

Comment 4

[Thomas D. (:thomas8)]

(In reply to comment #3)
> Why is my bug marked as a dupe of yours, instead of the opposite?? Doesn't
> make sense to me.

Oh, that's easy to explain:
Bug 522985...
- was there before this bug
- contains a lot more detailed information on the problem (including mockup screenshot)
- is more action-oriented than this bug (what exactly do we need to change if we want to implement this). Action-oriented bugs have a higher chance of getting implemented, which I believe is an interest that we share.

I hope that makes sense. Your bug isn't lost, on the contrary, it's now adding strength to our common cause.

Comment 5

[Frédéric Buclin]

(In reply to comment #4)
> Oh, that's easy to explain:
> Bug 522985...
> - was there before this bug

ok, some maths: 481211 < 522985. ;)

Comment 6

[Thomas D. (:thomas8)]

(In reply to comment #5)
> (In reply to comment #4)
> > Oh, that's easy to explain:
> > Bug 522985...
> > - was there before this bug
> 
> ok, some maths: 481211 < 522985. ;)

Yea, I'll need to re-check my maths. It's no secret I am coming from languages, which also correlates with the extent of some of my comments. Obviously yours was there first but I hadn't seen it when I posted mine, sorry for that. I'd still say we'd better keep 522985 because of the other two arguments mentioned in comment #4.