Hi Sam,
On Fri, Jan 10, 2014 at 5:58 PM, Sam Adams ssadams@us.ibm.com wrote:
I've hacked Glamour in two places recently to great effect, and I wonder what the community thinks about these new features. Both involve allowing arbitrary morphs in place of assumed images or text.
- I found myself wanting to have act: icons that changed state, acting
like a toggle. In my case I wanted to use a PluggableThreePhaseButtonMorph. A one line mode to GLMMorphicRender>>renderAction: did the trick
renderAction: anAction ^(UITheme current newButtonIn: nil for: anAction getState: nil action: #morphicActOn: arguments: {} getEnabled: nil
label: (anAction icon isMorph ifFalse:[AlphaImageMorph new image:
anAction icon]ifTrue:[anAction icon]) <<<<<<<< help: (anAction title, Character tab asString, anAction shortcutAsString) trimBoth) valueOfProperty: #noBorder ifAbsentPut: [true]; "this is a hack to tell the GLMUITheme to not draw the border and the fill" valueOfProperty: #noFill ifAbsentPut: [true]; setProperty: #wantsKeyboardFocusNavigation toValue: false; "to disable the focus" yourself
Interesting. We certainly do need to extend Glamour to support preferences not just actions, and a toggle is one kind of preferences. With your approach, you would not have access from the browser to the value of the toggle, and this would make it less valuable. A better approach would be to subclass Action with a ToggleAction and model it explicitly. Would you like to give this a try?
- I also needed to embed morphs in table presentations to allow editing
of the cell text. Again, a single change to GLMTreeMorphNodeModel>>rowMorphForColumn:
rowMorphForColumn: aGlamourColumn
| content | content := self containerTree glamourPresentation column: aGlamourColumn valueFor: self item.
^content isString ifTrue:[StringMorph contents: content]
ifFalse:[content] <<<<<<<<
This lead me to customize a subclass of PluggableTextMorph to use blocks instead of the model>>selector approach for pluggable behavior, especially so it would work nicely in a Glamour script and all the context to be referenced on the getText/setText etc operations.
Nice. I integrated this one simply by calling asMorph to content.
Cheers, Doru
These are obviously hacks but since both cases are wanting a morph and seemingly don't really need a morph of the original type (Image or String) they should be generalized to allow arbitrary morphs. This also led me to imagine embedding full presentations or visualizations into table cells, but that's beyond my understanding of the framework at this point.
Reminds me of the old XEROX Analyst spreadsheet.
Regards, Sam
Sam S. Adams, CTO - Contextual Computinghttps://w3-connections.ibm.com/communities/service/html/communityview?communityUuid=29937c79-74c1-439c-b096-4988a115daaf IBM Distinguished Engineer, IBM Research Mobile: 919-696-6064, email: ssadams@us.ibm.com Assistant: Linda R. Morrison. (720) 395-0460 Fax: (845) 491-4318, Tie: 676-0460, linda.r.morrison@us.ibm.com <<Hebrews 11:6, Proverbs 3:5-6, Romans 1:16-17, 1 Corinthians 1:10>>
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev