On Jun 9, 2010, at 12:15 PM, Henrik Johansen wrote:
On Jun 9, 2010, at 11:38 39AM, Stéphane Ducasse wrote:
adrian there are some fixed simple UTF and other pending fixes that we could integrate (was planning for after lunch) Stef
Also, people who know there is something that really needs to be fixed for a 1.1 release should speak up now.
Cheers, Adrian
- ConnectionQueue. (2476) - WeakKeyDictionaries. Easiest would probably be to port the one in Squeak now (with related finalization classes), as Levente did some work on that compared to what was in Squeak 4.0 / Pharo 1.0 which made it faster without breaking it when not registered for finalization. Alternatively wait for VM's to be built with Igor's finalization enhancements, require those for 1.1 WeakRegistries, and rewrite WeakKeyDictionaries to no longer assume they may be used/are always registered for finalization. (1628, 2309) - Encoding for code importing/exporting (setConverterForCode methods 1608 , 1872)
are some fixed?
Stef
No, he asked for important things that _needs_ to be fixed :) 1st is a regression from 1.0. 2nd I explained a few approaches to "fixing". 3rd is not a regression from 1.0, but still a blatant problem for anyone expecting to be able to write their code in anything but ascii, and be able to reliably export/import it somewhere else. Whether it is a blocking issue or not, I suppose depends on whom you ask.
I will tag all the ones you mentioned as 1.1.... so we can assess. The idea is that not releasing 1.1 because of a bug that was already in 1.0 makes no sense. E.g. it means that everyone already using 1.0 did not find it a grave problem for *months*, so I am sure people can wait for 1.2 to have it fixed. Marcus -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.