<br><br><div class="gmail_quote">On Tue, Jan 4, 2011 at 10:20 AM, Igor Stasenko <span dir="ltr"><<a href="mailto:siguctua@gmail.com">siguctua@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div><div></div><div class="h5">On 4 January 2011 08:45, Levente Uzonyi <<a href="mailto:leves@elte.hu">leves@elte.hu</a>> wrote:<br>
> On Mon, 3 Jan 2011, Eliot Miranda wrote:<br>
><br>
>> On Sun, Jan 2, 2011 at 11:38 PM, Levente Uzonyi <<a href="mailto:leves@elte.hu">leves@elte.hu</a>> wrote:<br>
>><br>
>>> Quoting Chris Muller <<a href="mailto:asqueaker@gmail.com">asqueaker@gmail.com</a>>:<br>
>>><br>
>>> Levente wrote:<br>
>>>><br>
>>>> I think we should wait for the new VMs, before releasing the new<br>
>>>> images.<br>
>>>>><br>
>>>><br>
>>>> Hi Levente,<br>
>>>><br>
>>>> I did agree, but now that Eliot is on a tear with a new code<br>
>>>> generator, the first official release might still be some weeks away.<br>
>>>> Someone please correct me if that is a misstatement.<br>
>>>><br>
>>>> It seems unlikely 4.2 will be incompatible with whatever ends up being<br>
>>>> the first official Cog release. We should consider pushing it out the<br>
>>>> door so we can re-open the trunk to new featurey.<br>
>>>><br>
>>><br>
>>> That's right, Cog's recent improvements introduced some instability. So I<br>
>>> think we should release Squeak 4.2 with SqueakVM only and postpone the<br>
>>> multi-VM release. We can also make the release cycle of Squeak 4.3<br>
>>> shorter<br>
>>> to catch up with the two release per year policy and deliver the VM<br>
>>> improvements earlier. I'm also sure that a lot of improvements will be<br>
>>> done<br>
>>> till April/May on both image and VM side.<br>
>>><br>
>><br>
>> It would be my preference to release with the Cog VM using the naive<br>
>> SimpleStackBasedCogit code generator. After all that's worth at least 3x<br>
>> over the standard VM whereas the unstable StackToregisterMappingCogit is<br>
>> only ~ 8% faster than SimpleStackBasedCogit. It's quite easy for me to<br>
>> generate either; the two coexist in the same VMMaker.<br>
><br>
> Okay, then I guess we should make sure that the two VMs (SqueakVM and<br>
> CogVM with SimpleStackBasedCogit) behave the same way in most cases. Here's<br>
> an incomplete list of VM differences:<br>
> - CogVM does compile the FloatMathPlugin properly (no crash), but SqueakVM<br>
> doesn't or it does without optimizing the code. Wwe can apply some<br>
> optimization for most methods, see the makefiles of Cog.<br>
> - CogVM doesn't support the new finalization scheme<br>
<br>
</div></div>this is easy to fix, if Eliot would merge my changes to VMMaker,<br>
like VMMaker-oscog-Igor.Stasenko.43 and build new VM's using merged version.<br></blockquote><div><br></div><div>Soon :)</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div><div></div><div class="h5"><br>
> - there's no SoundPlugin in CogVM on windows<br>
> - the SocketPlugin of CogVM seems to be missing the new primitives<br>
> - AllocationTest >> #testOneGigAllocation is still failing with CogVM<br>
> - I'm not sure if SqueakVM has all of the mirror primitives<br>
><br>
><br>
> Levente<br>
><br>
> P.S.: does the ~8% speedup come from the recompilation benchmark? Is the<br>
> code for the Computer Language Benchmarks Game available somewhere? It<br>
> would be good to test the new stdio support with it.<br>
><br>
<br>
<br>
<br>
</div></div><font color="#888888">--<br>
Best regards,<br>
Igor Stasenko AKA sig.<br>
<br>
</font></blockquote></div><br>