[asterisk-dev] [design] Matching algorithm

[Russell Bryant]

Leif Madsen wrote:
> Tony Mountifield wrote:
>>> In an ideal world, I choose #2. In reality, I choose #1, because I don't 
>>> want to break more dialplans that we have to.
>> Fine, if that's the case, then I agree #2 would be too big a change.
>> And thanks to Lief for the _mai[n]-X. trick. I hadn't thought of that one.
> 
> Also agreed with the #2 option -- is there anyway we could end up 
> deprecating the lowercase usage for a version of make #1 a reality in 1.8?
> 
> A warning at dialplan reload each time that stated
> 
> [WARNING] Use of lowercase pattern matching is deprecated and will be 
> removed in the next version of Asterisk. Please use uppercase for 
> pattern matching.
> 
> Also, would it be a good idea to have it as an option in the [general] 
> section of the dialplan?
> 
> lowercasepatternmatch=<yes|no>
> 
> And we can thank Jared for the [n] trick -- he was the one who gave me 
> the idea when I ran into the same issue you described.

I'd rather not.  I don't see any real benefit of deprecating it.  If we 
already support it, let's just leave it that way.

I think we as a community have leaned a little bit too far on the side 
of "make it as perfect as we can", versus the opposite side, which is 
"maintain compatibility at all costs".  I'd like to start moving back 
the other way.

-- 
Russell Bryant
Senior Software Engineer
Open Source Team Lead
Digium, Inc.