launchpad-dev team mailing list archive

[Martin Pool <mbp@xxxxxxxxxxxxx>]

On 25 January 2011 12:22, James Westby <jw+debian@xxxxxxxxxxxxxxx> wrote:
> On Tue, 25 Jan 2011 12:10:43 +1100, Martin Pool <mbp@xxxxxxxxxxxxx> wrote:
>> I had a bit of a go at this over the weekend.  It is gratifyingly fast
>> compared to what I expect to see with launchpadlib clients, just
>> through doing fewer requests and not unnecessarily blocking on them.
>> I like that a lot.
>>
>> The code is in <http://launchpad.net/wrested> and lp:wrested.
>>
>> For instance:
>>
>>  ./wrestler.py https://api.launchpad.net/devel/bugs/1
>
> To be fair I believe lplib would perform approximately as well if you
> just did lp.load('https://api.launchpad.net/devel/bugs/1').

Right, I don't have any magic to make the packets faster, so if LP
does just one request it will be just as fast.  I think the
differences are:

 * for high-latency services, using some kind of async thing is very
good, and lplib (afaik) doesn't support either twisted or multiple
threads(?)
 * the natural way to use lplib has you do several transactions on the
way to asking for that url
 * if you don't have a known-fresh cache of the wadl, it will spend a
long time fetching that

So the first is a reason to do a client in Twisted; the other two are
reasons to reconsider what API we present to the application.

-- 
Martin