On 24.08.2012 13:14, Michael T. Pope wrote:
> I am running a lot of automated tests to check that my AI `improvements'
> actually improve.  Speeding this up is worthwhile, so I made a brutal patch to
> make headless mode work again, which yielded a 20% speed up[1].
>
> The patch (attached) just informs the GUI object when headless mode is
> required, and prefixes every GUI use of canvas et al with "if (headless)
> return" or thereabouts.  Before committing though, I have to wonder if there
> is a better way of doing this.  Any thoughts folks, particularly Paolo:-).
>    

I agree with Paolo.

> Cheers,
> Mike Pope
>
> [1] Further suggesting that we are spending a lot of time in I/O and XML
> converters.

This is to be expected. As pointed out before, I think we need to change 
the network code to use an object queue rather than a text protocol with 
full serialization and deserialization in the case of local connections. 
That should increase speed drastically.


Regards

Michael

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Freecol-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/freecol-developers

Reply via email to