Problems in Squeak 3.9

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

Problems in Squeak 3.9

Jerome Peace
Problems in Squeak 3.9


Hi Roland,

Welcome to squeak.

***
>squeak-dev.5.pris at spamgourmet.com
squeak-dev.5.pris at spamgourmet.com
>Wed Jan 30 05:27:49 UTC 2008
>
>Squeak 3.9
>latest update #7068
>I'm running this on Ubuntu Linux.
>
>1) First a problem with the Squeak GUI.
>This happens if you pressed caps-lock without
noticing. To replicate
>the problem, open a browser and make a change in the
workspace. Now
>press Caps-lock. Then try to close the browser window
and you can't!
>When this happened to me I wasn't aware of the
caps-lock and ended up
>shutting down Squeak the hard way. Try it out to see
what I mean.
>
This annoyance is probably worthy of a bug report.*


>2) A bug with SketchMorph
>I was following the tutorial, which uses SketchMorph:
>http://squeak.preeminent.org/tut2007/html/078.htm
>
>Unfortunately I wasn't aware that this class is not
working
>properly(more on that at 3). I finally switched to
ImageMorph and got
>the problem fixed.
>When I read the documentation of ImageMorph I found
the following:
>
>Relationship to SketchMorph: ImageMorph should be
favored over
>SketchMorph, a parallel, legacy class -- see the
Swiki FAQ for details
>( http://minnow.cc.gatech.edu/squeak/1372 ).
>
>It would be sensible to put a corresponding notice in
the SketchMorph,
>otherwise you won't have a clue about this, and btw
update the
>tutorial to use ImageMorph also.  Btw this took me
over 10 hours to
>debug, because I first thought this was some problem
with Squeak
>events.

Yes the transparent pixel feature is frustrating.
>
>3) Bug with SketchMorph.
>
>The problem is in this message:
>
>containsPoint: aPoint
> ^ (self bounds containsPoint: aPoint) and:
>  [(self rotatedForm isTransparentAt: aPoint -
bounds origin) not]
>
>Even if the point is in SketchMorph the following
expression returns
>true: (self rotatedForm isTransparentAt: aPoint -
bounds origin)
>and so the MouseClick isn't detected.

Another good bug report, I think.
Hmm. If the point you are clicking over is not
transparent and therefore the mouse should see it.

Is the rotated form different than the original form?
or the same?
With rotation transformers in the middle of things
there is lots of room for confusion.
>
>This message is sent from the MorphicEventDispatcher
to check the mouse click:
>aMorph containsPoint: anEvent cursorPoint event:
anEvent
>
>4) Bug with Morph class. The message
wantsHaloFromClick always answers
>true. It should be possible to turn this off, I don't
want halos from
>submorphs.

I think this is a preference issue.
Preferences disable:  #wantsHaloFromClick (or whatever
the name of the preference really is)
should do the trick.
>
>5) Question: In the GUI editor how do I change
instance messages into
>class messages? The only way I know is copying by
hand then deleting
>the old one.

Fastest way I know is the cut and paste. Though if you
are thinking to move them you probably need to think
again.

most class mesages of this nature get implemented:

MyClass class>>#makeMyDay

^self new makeMyDay

where the second makeMyDay is implemented on the
instance side.

classes are generally for creating instances.
>
>6) What would be the best way to accomplish the
following: I have a
>morph and in it I want to have a canvas where I can
draw, fill
>rectangles and also write text in various positions.
How do I best
>embed a Canvas into a Morph?

In 3.9 I implemented a snapshot feature (connected to
the halos as Shift click on the blue viewer halo.)
So I would assemble a morph with the pieces I want
embedding them into the morph and moving them around
with the brown halo and then take a snapshot to get an
imagemorph like thumbnail.

This is not exactly what you asked to do but it is the
easy way to get close to the same results.
>
>Roland
***

* FAQ: Is this a known issue ? Where is the place to
report bugs (or check if
some have already been raised) ?

The best place for this info would be to start a
Mantis report. (You can get a mantis acct freely and
easily).

A good place to start is:

http://bugs.squeak.org/my_view_page.php

Mantis provides a patient persistent way to focus on
an issue.
I use it to accumulate data on a problem until a
solution can be found.
It provides a place to alert the community to a
problem;
-accumulate facts and clues from the analysis;
-publish preposed solutions and get feedback;
-get solutions harvested and included into the main
stream.

This FAQ is available at the Squeak wiki site.
http://wiki.squeak.org/squeak/Mantis%20FAQ%20and%20Tips
aka:
http://wiki.squeak.org/squeak/5912
Mantis FAQ and Tips



      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page.
http://www.yahoo.com/r/hs

Reply | Threaded
Open this post in threaded view
|

Re: Problems in Squeak 3.9

squeak-dev.5.pris
On Jan 31, 2008 5:20 AM, Jerome Peace - [hidden email]

> Hi Roland,
>
> Welcome to squeak.

Thanks!


> >3) Bug with SketchMorph.
> >
> >The problem is in this message:
> >
> >containsPoint: aPoint
> >       ^ (self bounds containsPoint: aPoint) and:
> >         [(self rotatedForm isTransparentAt: aPoint -
> bounds origin) not]
> >
> >Even if the point is in SketchMorph the following
> expression returns
> >true: (self rotatedForm isTransparentAt: aPoint -
> bounds origin)
> >and so the MouseClick isn't detected.
>
> Another good bug report, I think.
> Hmm. If the point you are clicking over is not
> transparent and therefore the mouse should see it.
>
> Is the rotated form different than the original form?
> or the same?
> With rotation transformers in the middle of things
> there is lots of room for confusion.

Well, I didn't rotate the form at all and AFAIK the point where I
clicked isn't transparent.

> >4) Bug with Morph class. The message
> wantsHaloFromClick always answers
> >true. It should be possible to turn this off, I don't
> want halos from
> >submorphs.
>
> I think this is a preference issue.
> Preferences disable:  #wantsHaloFromClick (or whatever
> the name of the preference really is)
> should do the trick.

Well I looked at the sourcecode of the Morph class it looks like this:

wantsHaloFromClick
        ^true.