[Q] MCP - Messages sent but not implemented

Daniel Vainsencher danielv at netvision.net.il
Sat Mar 8 02:21:37 UTC 2003


If not, don't wait too much - try to become as confident as you can in
what you've figured out, and then try to find a specific reviewer that
can look at your work and give you guidance about the rest.

Fishing at the list is very often effective, but detailed couching might
be better sought personally, and done off-list.

Daniel

diegogomezdeck at consultar.com wrote:
> German, did you receive any answer?
> 
> Diego
> 
> > Hi everybody,
> >
> > Following the 1st step in the Morph Cleaning Project (small SLint
> > suggested refactorings), I was trying the "Messages sent but not
> > implemented" SLint rule only on Morph.
> >
> > I came across some interesting results, for example things that
> > shouldn't work as they are, or perhaps work but only on some Morph
> > subclasses or given some context.
> >
> > I'm interested in comments from experienced people on why this things
> > are implemented this way - perhaps there are obscure reasons out of my
> > reach.
> >
> > Here is the list:
> >
> > #addStandardHaloMenuItemsTo:hand:
> > sends: #addWorldHaloMenuItemsTo:hand:
> > which is defined in: PasteUpMorph
> > comment: checks for #isWorldMorph before sending
> > solution: move part to PasteUpMorph and call super
> >
> > #beThisWorldsModel
> > sends: #model:slotName:
> > which is defined in: MorphicModel
> > comment: checks for #isMorphicModel before sending
> > solution: move beThisWorldsModel to MorphicModel; add some mechanism to
> > #buildMetaMenu: so subclasses can add menu options. the menu options
> > will probably be mixed.
> >
> > #buildDebugMenu:
> > sends: #model
> > which is defined in: MorphicModel
> > comment: checks for #isMorphicModel before sending
> > solution: add some mechanism to #buildDebugMenu: so subclasses can add
> > menu options. the menu options will probably be mixed.
> >
> > #chooseNewGraphicCoexisting:
> > sends: #form
> > which is defined in: some Morph subclasses
> > comment: doesn't work for some Morph subclasses
> > solution: welcome
> >
> > #choosePartName
> > sends: #nameMeIn: #renameMe
> > which are defined in: Component, ComponentLikeModel
> > comment: checks for #self world model isKindOf: Component before
> > sending solution: welcome
> >
> > #convertAugust1998:using:
> > sends: #bePossessive
> > which is defined in: TileMorph
> > comment: checks for key == #possessive before sending
> > solution: welcome
> >
> > #enforceTileColorPolicy
> > sends: #coloredTilesEnabled
> > which is defined in: nowhere
> > comment: relies on #doesNotUnderstand and #valueOfFlag:
> > solution: not needed
> >
> > #removeAlarm:at:
> > sends: #removeAlarm:at:for:
> > which is defined in: nowhere
> > comment: it shouldn't work
> > solution: welcome
> >
> > #showActions
> > sends: #actionSelector
> > which is defined in: various Morph subclasses
> > comment: checks if respondsTo: #actionSelector before sending
> > solution: not needed
> >
> > #specialNameInModel
> > sends: #slotName
> > which is defined in: MorphicModel and others
> > comment: checks for #isMorphicModel before sending
> > solution: move part to MorphicModel and call super
> >
> > #undoMove:redo:owner:bounds:predecessor:
> > sends: #activate
> > which is defined in: SystemWindow
> > comment: checks for #isSystemWindow before sending
> > solution: move part to SystemWindow and call super
> >
> >
> > Thanks in advance,
> >
> > German Morales



More information about the Squeak-dev mailing list