Tim's New Compiled Method format

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

Tim's New Compiled Method format

Tapple Gao
 
While talking to Craig about spoon, I learned that Tim made a
new Compiled Method format using normal objects, and (I think) a
VM that ran it. I tried searching for it and came up with:

New Compiled Method Format and 3.0 Image
    http://wiki.squeak.org/squeak/750
VI4
    http://wiki.squeak.org/squeak/2119
Version 4
    http://wiki.squeak.org/squeak/3716

I have been discussing what would go into Squeak 4 since we now
need to create such a system, due to the urgency of the license
situation, and I am now leaning toward Spoon rather than
KernelImage as a basis for Squeak4, after learning the magnitude
of the relicense effort on Monday.

So, where could I find this code? Does it work? Is there just a
CompiledMethod format change, or is there a whole other VM
project I never heard of? I can't really tell, but now would be
the time to use it if it works.

--
Matthew Fulmer -- http://mtfulmer.wordpress.com/
Reply | Threaded
Open this post in threaded view
|

Re: Tim's New Compiled Method format

Michael van der Gulik-2
 


On Thu, Mar 27, 2008 at 7:06 AM, Matthew Fulmer <[hidden email]> wrote:

While talking to Craig about spoon, I learned that Tim made a
new Compiled Method format using normal objects, and (I think) a
VM that ran it. I tried searching for it and came up with:

New Compiled Method Format and 3.0 Image
   http://wiki.squeak.org/squeak/750
VI4

   http://wiki.squeak.org/squeak/2119
Version 4
   http://wiki.squeak.org/squeak/3716

I have been discussing what would go into Squeak 4 since we now
need to create such a system, due to the urgency of the license
situation, and I am now leaning toward Spoon rather than
KernelImage as a basis for Squeak4, after learning the magnitude
of the relicense effort on Monday.


Tell us about the magnitude of this relicense effort... is it unachievable? Would we spend less time on the relicensing if we "just did it" using vanilla 3.10, or would it be less time in total to include this refactoring first and then do the relicensing?

Personally, I will eventually (1-2 years maybe?) need a refactored and tidied up CompiledMethod format for my own project, so I have a vested interest in this.

Gulik.


--
http://people.squeakfoundation.org/person/mikevdg
http://gulik.pbwiki.com/
Reply | Threaded
Open this post in threaded view
|

Re: Tim's New Compiled Method format

Tapple Gao
 
On Thu, Mar 27, 2008 at 09:27:45AM +1300, Michael van der Gulik wrote:

>    On Thu, Mar 27, 2008 at 7:06 AM, Matthew Fulmer <[hidden email]> wrote:
>
>      While talking to Craig about spoon, I learned that Tim made a
>      new Compiled Method format using normal objects, and (I think) a
>      VM that ran it. I tried searching for it and came up with:
>
>      New Compiled Method Format and 3.0 Image
>         http://wiki.squeak.org/squeak/750
>      VI4
>         http://wiki.squeak.org/squeak/2119
>      Version 4
>         http://wiki.squeak.org/squeak/3716
>
>      I have been discussing what would go into Squeak 4 since we now
>      need to create such a system, due to the urgency of the license
>      situation, and I am now leaning toward Spoon rather than
>      KernelImage as a basis for Squeak4, after learning the magnitude
>      of the relicense effort on Monday.
>
>    Tell us about the magnitude of this relicense effort... is it
>    unachievable? Would we spend less time on the relicensing if we "just did
>    it" using vanilla 3.10, or would it be less time in total to include this
>    refactoring first and then do the relicensing?

Currently, August 2008 is looking like a reasonable release
time. For details, see
http://installer.pbwiki.org/LicenseAuditing

We have 2 months or so in which no relicensing work can happen
(in Plan B), and that time could be spent getting a VM bug-fix
into the image.

>
>    Personally, I will eventually (1-2 years maybe?) need a refactored and
>    tidied up CompiledMethod format for my own project, so I have a vested
>    interest in this.
>
>    Gulik.
>
>    --
>    http://people.squeakfoundation.org/person/mikevdg
>    http://gulik.pbwiki.com/


--
Matthew Fulmer -- http://mtfulmer.wordpress.com/
Reply | Threaded
Open this post in threaded view
|

RE: Tim's New Compiled Method format

Ron Teitelbaum
 
Can we get a clear statement about what a rewrite process would look like?

Can the new author read the old method?  Can someone else produce a
requirements document that specifies what the method needs to do?  Can
someone write tests from the old method that need to pass with the new
method?

I still do not understand what we are allowed to do in order to reverse
engineer an unclean method.

Thanks,
Ron Teitelbaum

> -----Original Message-----
> From: Matthew Fulmer
>
>
> On Thu, Mar 27, 2008 at 09:27:45AM +1300, Michael van der Gulik wrote:
> >    On Thu, Mar 27, 2008 at 7:06 AM, Matthew Fulmer <[hidden email]>
> wrote:
> >
> >      While talking to Craig about spoon, I learned that Tim made a
> >      new Compiled Method format using normal objects, and (I think) a
> >      VM that ran it. I tried searching for it and came up with:
> >
> >      New Compiled Method Format and 3.0 Image
> >         http://wiki.squeak.org/squeak/750
> >      VI4
> >         http://wiki.squeak.org/squeak/2119
> >      Version 4
> >         http://wiki.squeak.org/squeak/3716
> >
> >      I have been discussing what would go into Squeak 4 since we now
> >      need to create such a system, due to the urgency of the license
> >      situation, and I am now leaning toward Spoon rather than
> >      KernelImage as a basis for Squeak4, after learning the magnitude
> >      of the relicense effort on Monday.
> >
> >    Tell us about the magnitude of this relicense effort... is it
> >    unachievable? Would we spend less time on the relicensing if we "just
> did
> >    it" using vanilla 3.10, or would it be less time in total to include
> this
> >    refactoring first and then do the relicensing?
>
> Currently, August 2008 is looking like a reasonable release
> time. For details, see
> http://installer.pbwiki.org/LicenseAuditing
>
> We have 2 months or so in which no relicensing work can happen
> (in Plan B), and that time could be spent getting a VM bug-fix
> into the image.
>
> >
> >    Personally, I will eventually (1-2 years maybe?) need a refactored
> and
> >    tidied up CompiledMethod format for my own project, so I have a
> vested
> >    interest in this.
> >
> >    Gulik.
> >
> >    --
> >    http://people.squeakfoundation.org/person/mikevdg
> >    http://gulik.pbwiki.com/
>
>
> --
> Matthew Fulmer -- http://mtfulmer.wordpress.com/

Reply | Threaded
Open this post in threaded view
|

Re: License stuff

Tapple Gao
 
On Thu, Mar 27, 2008 at 09:00:48AM -0400, Ron Teitelbaum wrote:

>  
> Can we get a clear statement about what a rewrite process would look like?
>
> Can the new author read the old method?  Can someone else produce a
> requirements document that specifies what the method needs to do?  Can
> someone write tests from the old method that need to pass with the new
> method?
>
> I still do not understand what we are allowed to do in order to reverse
> engineer an unclean method.

A draft of "the rules" is at
http://installer.pbwiki.org/SqueakRelicenseParty

This has not yet been reviewed. Also, a rewrite effort will not
happen until July if we indeed go with spoon.

If you could, please wait until I write down the results of
meeting four (which was 3 days ago, on Monday). I'm compiling a
web site with all this stuff, but I've been pretty busy this
week. I'll announce on squeak-dev when it's done.

--
Matthew Fulmer -- http://mtfulmer.wordpress.com/
Reply | Threaded
Open this post in threaded view
|

RE: Re: License stuff

Ron Teitelbaum
 
> -----Original Message-----
> From: Matthew Fulmer
> A draft of "the rules" is at
> http://installer.pbwiki.org/SqueakRelicenseParty 

Thank you that was exactly what I was looking for.  Very nice.

Ron

Reply | Threaded
Open this post in threaded view
|

Re: Re: License stuff

NorbertHartl
In reply to this post by Tapple Gao
 
On Thu, 2008-03-27 at 06:28 -0700, Matthew Fulmer wrote:

> On Thu, Mar 27, 2008 at 09:00:48AM -0400, Ron Teitelbaum wrote:
> >  
> > Can we get a clear statement about what a rewrite process would look like?
> >
> > Can the new author read the old method?  Can someone else produce a
> > requirements document that specifies what the method needs to do?  Can
> > someone write tests from the old method that need to pass with the new
> > method?
> >
> > I still do not understand what we are allowed to do in order to reverse
> > engineer an unclean method.
>
> A draft of "the rules" is at
> http://installer.pbwiki.org/SqueakRelicenseParty
>
> This has not yet been reviewed. Also, a rewrite effort will not
> happen until July if we indeed go with spoon.
>
> If you could, please wait until I write down the results of
> meeting four (which was 3 days ago, on Monday). I'm compiling a
> web site with all this stuff, but I've been pretty busy this
> week. I'll announce on squeak-dev when it's done.
>
The mailing list that is mentioned on the page does not exist.


Norbert