First impressions

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

First impressions

Chris Uppal-3
Well, I've been playing with D5 for a couple of hours now (instead of
getting on with the housework that I *ought* to be doing...).  I think I'm
going to be very pleased with all this new stuff.  Loverly!

I've not got so far as actually trying to load my own code in (there are
some problems like name clashes), but have found a few small, or very small,
problems so far:

Package Browser:
=============
I agree with Ian that automatically selecting all the packages in each
branch is too much.

The 'Save All..' command is never enabled.  I guess that's because the
#saveAllPackages method is not provided (possibly because it thinks it's an
evaluation edition?).

Clicking on the 'Open System Browser' toolbar item does nothing, although
the same operation does work from the menu.  (BTW, I'm assuming that the two
browsers in the toolbar will have different icons by the real release?).

I'm not convinced that the refactorings menu is worthwhile here; if it is,
then why have method refactorings but no class refactorings ?

"Remove" in the class context menu means remove-from-package, but in the
method menus it means delete-entirely.

The PB is (as it was in D4) the only "long-lived" tool from which it isn't
possible to save the image and exit Dolphin.

Dolphin Options:
============
The PAI includes the collection's "sort block" as one of the Dolphin
options -- looks a bit odd.  (It seems to be because
SortedCollection>>sortBlock is a "getter", and so #sortBlock is included in
the published aspects).

Transcript:
========
(Just thought I'd mention this -- I've been meaning to for a while but never
got around to it).  The Transcript Tools>Options menu shows 'Remember this
size' and 'Forget size' items, but they are never enabled.  D4 is the same.

Splash Screen:
===========
(This is *really* minor)  Is the use of the 'xp' initials likely to cause
problems and/or confusion with Windows XP ?

Inspectors:
========
Not a problem, just a suggestion.  Would it be a good idea if (say) holding
down <shift> changed the interpretation of inspectIt to force the use of a
basic inspector always ?

That's all for now...

    -- chris


Reply | Threaded
Open this post in threaded view
|

Re: First impressions

Bill Schwab
Chris,

> Inspectors:
> ========
> Not a problem, just a suggestion.  Would it be a good idea if (say)
holding
> down <shift> changed the interpretation of inspectIt to force the use of a
> basic inspector always ?

Interesting.  Maybe that could bring up the #alternateInspector??

Have a good one,

Bill

--
Wilhelm K. Schwab, Ph.D.
[hidden email]


Reply | Threaded
Open this post in threaded view
|

Re: First impressions

Blair McGlashan
In reply to this post by Chris Uppal-3
Chris

I'm determined to get through all these posting so I can get some of them
fixed, so I'm going to keep my answers brief.

You wrote in message news:[hidden email]...
> Well, I've been playing with D5 for a couple of hours now (instead of
> getting on with the housework that I *ought* to be doing...).  I think I'm
> going to be very pleased with all this new stuff.  Loverly!
>
> I've not got so far as actually trying to load my own code in (there are
> some problems like name clashes), but have found a few small, or very
small,
> problems so far:
>
> Package Browser:
> =============
> I agree with Ian that automatically selecting all the packages in each
> branch is too much.

I'm going to leave Andy to comment on that at a later date, but in the
meantime please try to think of the package tree as a filter.

>
> The 'Save All..' command is never enabled.  I guess that's because the
> #saveAllPackages method is not provided (possibly because it thinks it's
an
> evaluation edition?).

#493

>
> Clicking on the 'Open System Browser' toolbar item does nothing, although
> the same operation does work from the menu.

#494

>...(BTW, I'm assuming that the two
> browsers in the toolbar will have different icons by the real release?).

They do actually have subtly different icons.

> I'm not convinced that the refactorings menu is worthwhile here; if it is,
> then why have method refactorings but no class refactorings ?

It is to permit certain refactorings to be scoped to operate within a
package or packages. The only refactorings for which this makes sense (that
we have identified so far) are method refactorings. For example, you want to
rename a method within a group of packages without affecting other packages.

> "Remove" in the class context menu means remove-from-package, but in the
> method menus it means delete-entirely.

Are we talking about the 'Refactorings' sub-menu? The methods menu itself
says 'Remove from Package'. On the Refactorings sub-menu it is activating
the "Remove Method" refactoring, scoped to operate within the selected
packages (i.e. the method will only be removed from the system, and the
package, if it is not possibly invoked from within those packages).

> The PB is (as it was in D4) the only "long-lived" tool from which it isn't
> possible to save the image and exit Dolphin.

#495.

> Dolphin Options:
> ============
> The PAI includes the collection's "sort block" as one of the Dolphin
> options -- looks a bit odd.  (It seems to be because
> SortedCollection>>sortBlock is a "getter", and so #sortBlock is included
in
> the published aspects).

#496.

> Transcript:
> ========
> (Just thought I'd mention this -- I've been meaning to for a while but
never
> got around to it).  The Transcript Tools>Options menu shows 'Remember this
> size' and 'Forget size' items, but they are never enabled.  D4 is the
same.

#497

>
> Splash Screen:
> ===========
> (This is *really* minor)  Is the use of the 'xp' initials likely to cause
> problems and/or confusion with Windows XP ?

I couldn't possibly comment. :-)

> Inspectors:
> ========
> Not a problem, just a suggestion.  Would it be a good idea if (say)
holding
> down <shift> changed the interpretation of inspectIt to force the use of a
> basic inspector always ?

Good idea. #498

>
> That's all for now...

Thanks!

Regards

Blair


Reply | Threaded
Open this post in threaded view
|

Re: First impressions

Chris Uppal-3
In reply to this post by Bill Schwab
Bill,

> Interesting.  Maybe that could bring up the #alternateInspector??

That would be nice...

> Bill

    -- chris


Reply | Threaded
Open this post in threaded view
|

Re: First impressions

Chris Uppal-3
In reply to this post by Blair McGlashan
Blair,

> >...(BTW, I'm assuming that the two
> > browsers in the toolbar will have different icons by the real release?).
>
> They do actually have subtly different icons.

They do ?!  "subtle" is right...

> > I'm not convinced that the refactorings menu is worthwhile here; if it
is,
> > then why have method refactorings but no class refactorings ?
>
> It is to permit certain refactorings to be scoped to operate within a
> package or packages. The only refactorings for which this makes sense
(that
> we have identified so far) are method refactorings. For example, you want
to
> rename a method within a group of packages without affecting other
packages.

Ah, I hadn't realised that the refactorings were scoped.  Nice!

Still, that suggests that it'd be worthwhile making what the operations
really do clearer.

> > "Remove" in the class context menu means remove-from-package, but in the
> > method menus it means delete-entirely.
>
> Are we talking about the 'Refactorings' sub-menu? The methods menu itself
> says 'Remove from Package'. On the Refactorings sub-menu it is activating
> the "Remove Method" refactoring, [...]

But on the class context menu it uses the same words to mean something less
damaging (if misapplied).  All I'm really suggesting is that the class menu
should be changed to read "Remove from package" too.

> [...] scoped to operate within the selected
> packages (i.e. the method will only be removed from the system, and the
> package, if it is not possibly invoked from within those packages).

That sounds like a very useful operation to have.

But, again, I don't think the fact that it is "smart" is sufficiently
obvious from the context.  At least I don't think *I* would ever have
guessed if I hadn't been told (or got curious enough to check out the code).

> Blair

    -- chris