[Vm-dev] VM Maker: VMMaker-oscog.46.mcz

Mariano Martinez Peck marianopeck at gmail.com
Tue Jan 18 08:43:35 UTC 2011


On Tue, Jan 18, 2011 at 9:09 AM, stephane ducasse <
stephane.ducasse at gmail.com> wrote:

>
> >>
> >
> > Just a note of appreciation - keeping the oscog branch numbering
> > (currently at 46) separate from the trunk branch numbering (currently
> > at 216) has been very helpful and much appreciated from my point of view.
>
> David
>
> I'm curious. can you explain you last sentence?
> Because understanding different process benefits is always interesting for
> me.
>


I am not David, but I can explain since I was catched by this also.
A part of cog concist in a fork of the VMMaker package. Eliot started from
one VMMaker version. Let's take as an example that he started cog from
VMMaker.dtl.100. Instead of calling it VMMaker.eem.101  he named it
VMMaker.oscog.1

Notice two things:
- He used oscog as name and not eem. I think oscog comes from open source
cog.
- he started from 1 (or something like that).

Now, in the same repository and the same package of VMMaker you see both
"branches" of VMMaker, the standard one which is in the normal form
VMMaker.dtl.xxx  (or any other commiter), and those which are
VMMaker.oscog.yyy

So, if you want to search the latest cog version, you need to search the
latest VMMaker.oscog.yyy.  What it is not fun is that since eliot restarted
with the numbers, cog version are always at the end. For example, the lates
cog version .46 is muuuch before (in the monticello browser) that the
lastest of standard vm (246)

In the squeaksource page, when you see "latest" you see really the latest,
which in this case is the 246 of the normal squeakVM, It seems "latest" uses
the order you see in the monticello browser (version number) instead of the
timestamp of the commit.

Cheers

Mariano
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.squeakfoundation.org/pipermail/vm-dev/attachments/20110118/2d43f7ce/attachment.htm


More information about the Vm-dev mailing list