Klaus D. Witzel wrote:
I think that the below must be considered for inclusion into 4.0.
I think first we need a working implementation again, then we need to do some measures about speed and space and then we can start thinking about adoption. It's a long way to adopt such fundamental changes.
Cheers, - Andreas
/Klaus
On Tue, 03 Oct 2006 21:22:38 +0200, tim Rowledge wrote:
On 3-Oct-06, at 3:02 AM, Tony Garnock-Jones wrote:
Andreas Raab wrote:
Well, in that case all I can say is: "Code speaks" ;-) If you still have that code from '98 and can update it to work with method properties I'd like to see it and I'd like us to have the discussion about which approach people like more.
I'd very much like to see the code from 98 as it stands, without any changes, to get some feel for the scale of the change as proposed.
See
http://www.rowledge.org/tim/squeak/SqFiles/packages/NewCompiledMethod-Archiv...
for the original released files. The new CM class is very simple and it is the cloner changes that were the really tricky part - the earlier cloner was very limited and need quite a bit of sneaky work to make it possible to do the needed transmutations.
Note that this bunch of files don't have anything much by way of clever tricks for handling source; I can't find anything on my system from the experiements I did at Interval. Also note that the chances of being able to filein this code to any recent image is essentially zero. Don't waste your time!
tim
tim Rowledge; tim@rowledge.org; http://www.rowledge.org/tim Useful Latin Phrases:- Gramen artificiosum odi = I hate Astroturf.