On 18 Apr 2009, at 20:24, Richard Gaskin wrote:
The CGI is indeed fast, but if the timing is being measured inside
the script it's not accounting for the biggest difference between
the CGI and on-Rev: on-Rev has no load time to bring the engine
into memory and initialize it since it's already loaded and
running, while the CGI engine has to be loaded fresh each time it's
called.
Even with that extra overhead the Rev CGI measures well against
equivalent CGIs, but I'd be surprised if it could beat on-Rev.
--
if you put this in a button you can see another test:
on mouseUp
put "http://marksmith.on-rev.com/mashash/hashmac.irev?
data=somedata&key=somekey&action=md5hmac" into tIrevUrl
put "http://marksmith.on-rev.com/cgi-bin/hashmac.cgi?
data=somedata&key=somekey&action=md5hmac" into tCgiUrl
put the millisecs into ts
put url tIrevUrl into tResA
put the millisecs - ts into tTimA
put the millisecs into ts
put url tCgiUrl into tResB
put the millisecs - ts into tTimB
put "irev:" & tTimA && tResA & cr & "cgi:" & tTimB && tResB
end mouseUp
I'm seeing the cgi taking 190-200 ms and the irev taking 170-180 ms.
The irev is 'including' a textified version of my hash/hmac library,
and the cgi is loading a stack which inserts the library (and a few
others) into back, so perhaps the test is slightly skewed in irev's
favour.
I'll leave it up for a few hours if anyone wants to try it out (I'd
also be interested in other people's timing from different places -
I'm in London).
Best,
Mark
_______________________________________________
use-revolution mailing list
[email protected]
Please visit this url to subscribe, unsubscribe and manage your subscription
preferences:
http://lists.runrev.com/mailman/listinfo/use-revolution
