incubator-couchdb-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Randall Leeds <randall.le...@gmail.com>
Subject Re: geocouch: limit number of points in an area
Date Sun, 01 Aug 2010 21:25:47 GMT
On Sun, Aug 1, 2010 at 13:54, Volker Mische <volker.mische@gmail.com> wrote:
> On 01.08.2010 22:48, J Chris Anderson wrote:
>>
>> On Aug 1, 2010, at 1:43 PM, Volker Mische wrote:
>>
>>> On 01.08.2010 17:24, Benoit Chesneau wrote:
>>>>
>>>> On Sun, Aug 1, 2010 at 2:17 PM, Volker Mische<volker.mische@gmail.com>
>>>> wrote:
>>>>>
>>>>> On 01.08.2010 10:53, Benoit Chesneau wrote:
>>>>>>
>>>>>> I'm looking for a way to limit number of points retrieved for an
area
>>>>>> depending on its size. Ie When playing with the zoom on the map I
want
>>>>>> to be able to only display and retrieved only main points instead
of
>>>>>> all the points in this area. Is there a simple way to do that actually
>>>>>> ?
>>>>>>
>>>>>> - benoit
>>>>>
>>>>> Hi Benoit,
>>>>>
>>>>> there a two ways. Either introducing a "limit" parameter, but that's
>>>>> probably not what you want. It would limit to a arbitrary selection
>>>>> (not the
>>>>> "main points" as you'd like to have it).
>>>>> You probably want clusters. My suggestion is to do it on the client
>>>>> side
>>>>> with OpenLayers' cluster strategy [1][2].
>>>>>
>>>>> [1] http://openlayers.org/dev/examples/strategy-cluster.html
>>>>> [2] http://openlayers.org/dev/examples/strategy-cluster-threshold.html
>>>>>
>>>>> Cheers,
>>>>>  Volker
>>>>>
>>>> This solution doesn't work when you have a lot of points though. Do
>>>> you plan to add such things to geocouch ?
>>>>
>>>
>>> Clustering isn't that easy. I think some layer on top of GeoCouch
>>> should/will be built to handle such things. It can even be in any
>>> programming language and communicate with GeoCouch via HTTP.
>>>
>>
>> If there is a threshold that can be emitted with the points, maybe a _list
>> function is the way to filter it server side.
>>
>
> The problem is, that the points are potentially returned in random order. So
> you never know if there will be a point that is close to anther one at a
> later stage.

I think Chris was referring to a zoom threshold of sorts, which is
independent of nearby points.
R

Mime
View raw message