[OSM-dev] Potlatch - really BAD bug
shaun at shaunmcdonald.me.uk
Tue Jan 13 12:19:04 GMT 2009
On 13 Jan 2009, at 09:22, Ed Loach wrote:
> Richard wrote:
>> It has been explained many times in the
>> past, not just
>> by me, that database freakiness _will_ happen because we don't
>> transactions, our server is often under high load, and there
>> may be memory
>> leaks or blocking processes in some of the software we use.
> Do I understand correctly that the introduction of changesets in API
> 0.6 should go some way towards addressing this problem? As I
> understand it the whole changeset needs to make it to the server for
> it to be applied, and therefore I've been assuming that when
> updating a way the addition, updating and removal of nodes to a way
> would all be part of a single changeset. Or have I misunderstood?
Changesets are groups of edits. A whole changeset is not atomic (i.e.
does not need to be sent to the server in one go).
Each edit will be atomic (i.e. the whole request will all complete, or
will all fail if there is a problem with one part of the data, or
someone has modified the data since the start of the request or the
user last downloaded.)
The exposing of version numbers will help too.
More information about the dev