Hannes,
First a piece of magic: Yesterday, I used SqueakMap to load Connectors 2.0 
followed by BabySRE. It failed. I then tried Connectors1.9, BabySRE. It 
failed because SqueakMap did not retrieve all of the Baby Monticello file. 
There is some invisible interference. May be through the Monticello cache??

Today, I started again with a fresh 3.7-5989 image, fixed the maxWidth bug 
and loaded Connectors 1.9, BabySRE. It worked! Mysteriouser and 
mysteriouser. I clearly have to look into the installation process. I 
probably do something horribly illegal somewhere.

I do not understand your queation about decompiler. The decompiler is used 
when there is no source code. It converts the bytecodes of the 
CompiledMethod to text, using t1, t2, ... as names for the formal 
parameters and local variables.

I need more info to be of use...

Cheers
--Trygve

--



At 21:37 21.12.2004, Hannes wrote:
>Trygve,
>
>Trygve Reenskaug wrote:
> > Hannes,
> > I tried Squeak3.8a-5987.image and the collaboration worked.
> >
> > I then tried Squeak3.8g-6527 and the loading of connectors failed due to
> > an array out of bounds bug in StrikeFontSet>>maxWidth.
> >
> > Try to compile this change before you load Connectors 1.9.
> >    StrikeFontSet>>maxWidth
> >         ^ (fontArray at: (2 min: fontArray size)) maxWidth.
> > Load Connectors 1.9
> > Load BabySRE-TRee11
> >
>
>Yes this worked! The collaboration diagram tool is very useful.  It emphasizes
>the fact that everything is an object in Smalltalk. It allows mind-map like
>browsing/navigating and documentation of designs. On can think of having
>graphical browsers now.
>
>I did an example today and put the diagram here
>http://minnow.cc.gatech.edu/squeak/1934
>
>It is very easy and straightforward to use. Not more than 10 minutes to learn
>using it.
>It extends the IDE in an important way.
>
>
>The downside so far:
>The installation process is not smooth.
>I have installed  in a 3.8g-6527 image
>- Monticello223
>- RefactoringBrowser3.8.10 (or was it 11?) (RFB)
>- Fix for RFB
>- maxWidth-Fix in StrikeFontSet (see other thread)
>- Connectors 1.9
>- and finally SRE
>
>
>Remarks:
>
>a) SRE works fine. But when I develop the decompiler does not work correctly
>anymore. The parameters and local variables are replaced by t1,t2,t3....
>I do not know where this comes from.
>
>b) RefactoringBrowser
>Today Marcus Denker put out 3.8.12 where you no
>longer need to install the fix (mentioned in another thread today).
>
>c) Redoing an 3.8base image with additional packages
>I then tried to redo the same thing from scratch again this evening
>but this time SRE didn't load anymore.
>
>d) Current work method
>So for the time beeing I will use the image with SRE that works by loading my
>code from the Monticello repository do draw the diagrams wheras I will develop
>it in another image. Not ideal but a big step forward in any case.
>
>e) base image plus MC, RFB, Conn1.9, SRE
>If somebody other is successful assembling a 3.8gamma image with Monticello,
>RFB, Connectors and SRE without the aforementioned limit I will happily 
>provide
>webspace that other poeple can get started right away. This will be for a few
>months as I think later we will have better solutions (New version of
>SqueakMap).
>
>
>Thank you , Trygve, for this path-leading contribution!
>
>Hannes


-- 

Trygve Reenskaug      mailto: trygver <at> ifi.uio.no
Morgedalsvn. 5A       http://heim.ifi.uio.no/~trygver
N-0378 Oslo           Tel: (+47) 22 49 57 27
Norway