Beta 2

Hi Andy, Blair,

Just some remarks from my brief walking awound in B2.

I've got it running on both XPsp2 and on W2K without major issues.

The initial system folder shown after unlocking had a surprising order
of the tools; Panic was next to Additional Tools. After going into the
Additional Tools and back up to the main folder they were sorted by name
and I can't reproduce the original order (without probably completely
uninstalling and reinstalling and unlocking).

The issue that I reported in beta1 wrt the flickering of the screen when
opening a view in the IdeaSpace has completely disappeared - impressive!

When selecting the Executable browser in the system folder, it opens an
Environment Browser (as is described by the help). Wouldn't it be better
to just stick to one name?

One of my apps that I deployed in D6 reports the following error:
'Attempt to update read-only object'. I've got the same app deployed
under D5 which doesn't report that error. The error is raised by
Array>>at:put: so primitive 61 appears to be failing. In the offending
method I create an array with #(0 0 0 0 0 0) and then replace some of
the zero's by a different number using at:put:. I deployed this app in
Beta1 as well but didn't come across this issue (afaicr). Let me know if
you require more info/testing regarding this issue.

For the rest, I like it a lot even though some of the goodies that I
rely upon will need updating before I can fully move over to this version.
More specifically:
- Steve Waring's drop containers don't work because of the changes in
LayoutManager (#layoutContainer: being replaced by
#layoutContainer:context:)
- Solutions Software Editable ListView doesn't load out of the box.
Managed to get the bits that I use working by installing the individual
classes that I use.

Ted

Ted wrote: Some more info, if I create that array with Array>>new:withAll: then the
  problem doesn't occur.

Ted

Ted wrote:

> One of my apps that I deployed in D6 reports the following error:
> 'Attempt to update read-only object'. I've got the same app deployed
> under D5 which doesn't report that error. The error is raised by
> Array>>at:put: so primitive 61 appears to be failing. In the offending
> method I create an array with #(0 0 0 0 0 0) and then replace some of
> the zero's by a different number using at:put:.

Presumably this is the (intended) effect of the new immutability stuff.  See
Object>>isImmutable[:].

    -- chris

Hi,

Does DX6 support object inmutability ?

> Presumably this is the (intended) effect of the new immutability stuff.  See
> Object>>isImmutable[:].

I have simple tested this with anOrderedCollection it seems to be
working ok.

But "Processor constWriteSignal" answer a generic instance of Signal.
I think it would be good to answer InmutabilitySignal or something like
that.

Excellent enhancement !!!

Regards Bruno

"Ted" <[hidden email]> wrote in message
news:[hidden email]...
In X6 beta 2 array literals in methods are read-only (immutable). It has
always been an error to write to a literal constant in a method, it's just
that this error was not detected before.

Note that the immutability support in X6 currently only works for the
indexable fields of an object. It's possible this may change in the final
release (to include named variables as well), but I doubt it.

Regards

Blair

"Blair McGlashan" <[hidden email]> wrote in message
news:[hidden email]...
> ...
> In X6 beta 2 array literals in methods are read-only (immutable). It has
> always been an error to write to a literal constant in a method, it's just
> that this error was not detected before.
>
> Note that the immutability support in X6 currently only works for the
> indexable fields of an object. It's possible this may change in the final
> release (to include named variables as well), but I doubt it.

An updated on this. In the final X6 release immutability support will be
extended to named instance variables as well as indexable instance
variables.

Regards

Blair

Blair McGlashan escribió:
Excellent !!!

Regards Bruno