[squeak-dev] Bootstrapping closures

Eliot Miranda eliot.miranda at gmail.com
Mon Jul 7 18:25:36 UTC 2008


(subject changed; was Anyone know the following about Slang?)

On Mon, Jul 7, 2008 at 3:18 AM, Keith Hodges <keith_hodges at yahoo.co.uk>
wrote:

> Eliot Miranda wrote:
>
>> David,
>>
>>    thanks a lot!  I have some integration to do :)   Not least I want to
>> integrate the pragma support in 3.9 back into 3.8 because using pragmas
>> instead of the null statements (self var: #foo type: #barf) etc is so much
>> nicer.  I also have to produce bootstraps for the closure compiler in a few
>> images, e.g. Croquet 1.0, 3.9, 3.10 & Spoon.  Once I get the Croquet
>> bootstrap done I would welcome volunteers for the others.
>>
>> cheers!
>>
> I am curious as to how you intent to distribute these backported
> bootstraps? Is this something
> http://installer.pbwiki.com/LevelPlayingField can help you with?


I'm not yet sure how I'll distribute.  Ideally I think a set of Monticello
packages plus a Monticello package containing a bootstrap script.  This is
what I have internally for our Qwaq Forums image bootrtap (derived from
Croquet 1.0, in turn derived from 3.8).

But when I tried to modify this bootstrap to boot a minimal 3.8 for testing
the new VM I failed to get it working and gave up after about a day and a
half.  The thing I know will work is a set of change sets but that's not
ideal.

Thanks for the pointer to LPF.  That could indeed help.  But the real issue
with this bootstrap is the related changes straddling Compiler (a new
compiler for Closures), Kernel (a new BlockClosure and related changes in
ContextPart, InstructionStream, CompiledMethod et al), System (new
specialObjectsArray etc) and Tools (new interface between debugger and
contexts).  It is really easy to break the debugger when loading the
bootstrap incorrectly and that makes working out what's wrong tricky.  So
progress is slow.

Moving to change sets is easier because one doesn't have to filter out as
many extraneous changes in a package which are the result of other changes
unrelated to closures.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20080707/e76f39f0/attachment.htm


More information about the Squeak-dev mailing list