[OSM-talk] When will bots be removed from history?
Shaun McDonald
shaun at shaunmcdonald.me.uk
Sun Apr 4 19:54:31 BST 2010
On 4 Apr 2010, at 17:57, Valent Turkovic wrote:
> On Sun, 28 Feb 2010 17:11:12 +0100, Polderrunner wrote:
>
>> Why bother whether the changeset was created by a bot or not. Simply
>> offer the user the choice not to display big changesets (say those
>> covering more than 5 degrees in either latitude or longitude). That
>> should get rid of most bot edits and other edits unlikely to touch your
>> area of interest.
>>
>> It could be implemented with a checkbox on the history page. Such a
>> solution shouldn't need much extra CPU power.
>
> As you can see there are lots of implementations that don't use much CPU
> power, but there is still nothing implemented :(
Who is saying that it doesn't take much CPU?
As one of the developers who knows the code, this is a hard problem. Firstly from a UI perspective. Secondly from a backend implementation perspective. As currently whether a changeset is big is determined on the fly before being output, and isn't something that is stored or calculated at the database level, it becomes really difficult to calculate the list of small/big traces, hence why it isn't implemented. Third, you'll also find that there are big changesets that contain changes in that area, so you may actually be interested in them.
>
> Bots still make usage of history and monitoring for changes and also
> vandalism monitoring impossible :(
I think you'll find that tools like ITO World's OSM Mapper would be a better fit for the problem that you are trying to solve.
>
> Please, please, pretty please fix this.
When there is a good solution, it will be implanted. Until then, please use a tool that has been designed to the job.
Shaun
More information about the talk
mailing list