[postgis-devel] Death to Pointless Operators

[Paragon Corporation]

Paul and Nicklas,

Yah - sorry about that.  I meant ~=


I guess that whole ~= mess left a sore taste in my mouth for anything with ~
in it.

Can we just get rid of ~= then and overload = if we can?


-----Original Message-----
From: postgis-devel-bounces at postgis.refractions.net
[mailto:postgis-devel-bounces at postgis.refractions.net] On Behalf Of Paul
Ramsey
Sent: Sunday, December 19, 2010 2:07 PM
To: PostGIS Development Discussion
Subject: Re: [postgis-devel] Death to Pointless Operators

When you say ~, do you mean =~ ?

Because ~ never meant equality of any sort, it meant contained.

P.

On Sun, Dec 19, 2010 at 8:59 AM, Paragon Corporation <lr at pcorp.us> wrote:
> =  would mean bounding box equals as it does now except it would take 
> advantage of a gist index where it can.  I don't think we can change that.
> Its consistent with all operators are bounding box checks.
>
> So regardless of if its doing btree / gist it should mean the same 
> thing is my point.
>
> The distinctions look like just implementation artifacts I as a user 
> don't really care to know about.
>
> My concern with keeping ~ going is that its broken in the sense that 
> if you are a long time user of PostGIS and actualy used it  you still 
> have in your mind that it means geometric equality.
>
> When you port your apps over from say 1.3 (even 1.4 and 1.5 if you had 
> done a soft upgrade) -- the meaning of it has suddenly changed.
>
> Your apps aren't breaking -- there just returning the wrong answers 
> which is much harder to debug than to get back a " I'm sorry  ~ doesn't
exist."
>
> R
>
> -----Original Message-----
> From: postgis-devel-bounces at postgis.refractions.net
> [mailto:postgis-devel-bounces at postgis.refractions.net] On Behalf Of 
> Paul Ramsey
> Sent: Sunday, December 19, 2010 11:38 AM
> To: PostGIS Development Discussion
> Subject: Re: [postgis-devel] Death to Pointless Operators
>
> Would '=' be equivalent to ST_Equals, ST_OrderingEquals or 
> bounding-box-equals?
> P
>
> On Sun, Dec 19, 2010 at 7:32 AM, Paragon Corporation <lr at pcorp.us> wrote:
>> :
>>
>>>> If you get rid of anything you should really get rid of that ~ -- 
>>>> its just confusing because it means something completely different 
>>>> as Nicklas mentioned a while ago (when you are talking about the 
>>>> built-in polygon/point and I think even pgsphere --
>>>
>>> ~ is the opposite of @. But it can be any symbol we like. How about !@ ?
>>>
>>
>>> Actually that would be a bad symbol pairing (the equals/notequals 
>>> parallel
>> is not correct, A = B does not imply B != A). A better one would be 
>> if we used >> and << since they are commutative relations.
>>> But that would break backwards compatibility which above you are 
>>> telling
>> me is a big deal for the operators that people don't actually use!
>>
>>> So, for sheer conservative curmudgoenliness, I would suggest we 
>>> leave ~
>> and @ as they are.
>>
>>
>> Like I said if we can I would prefer to overload the = operator like 
>> ltree does so I don't have to go around explaining why we have ~ and 
>> = which mean the same thing but one uses an index and one doesn't.
>> Or you think there is a benefit there of keeping them separate?
>>
>> http://www.postgresql.org/docs/current/interactive/ltree.html
>>
>>
>> _______________________________________________
>> postgis-devel mailing list
>> postgis-devel at postgis.refractions.net
>> http://postgis.refractions.net/mailman/listinfo/postgis-devel
>>
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-devel
>
>
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-devel
>
_______________________________________________
postgis-devel mailing list
postgis-devel at postgis.refractions.net
http://postgis.refractions.net/mailman/listinfo/postgis-devel