[V3dot10] Experimental 3.10 image without Etoys

[Lex Spoon]

"Edgar J. De Cleene" <edgardec2001 at yahoo.com.ar> wrote:
> El 4/26/07 12:04 PM, "Damien Cassou" <damien.cassou at gmail.com> escribió:
> > StringSocket depends on CanvasEncoder which is not in the image.
>
> The only place is :
> addToOutBuf: arrayToWrite
[...]
>     CanvasEncoder at: 1 count: arrayToWrite size + 1.
[...]
> The line could be commented out for no Undefined.
> CanvasEncoder is for Nebraska use

Ugh!  Yes, I think this is just for logging purposes.  An immediate fix
is to check for whether CanvasEncoder is loaded.  Something like:

Smalltalk at: #CanvasEncoder ifPresent: [ :canvasEncoder |
	canvasEncoder at: 1 count:  arrayToWrite size + 1 ]


There are better fixes in the long run, but I don't think anyone is
available to really work on this right now, so I won't go into it in
this email.


"Damien Cassou" <damien.cassou at gmail.com> wrote:
> Do you mean it would be interesting to have a separate squeaksource
> package for them and not include them in the base image?

Yes, I think so.  For now, package universes-based images would always
have this package loaded, but logically it should be separate.  In the
future, maybe the package universes editor will use a different
protocol....

I guess "ObjectSocket" would be a good name for the package, or
"Network-ObjectSocket" if that works better with Monticello.

Lex