> On 03 May 2015, at 06:25, Sean P. DeNigris <[email protected]> wrote:
> 
> Questions first:
> - TxModel class comment - "I don't provide a direct interface for
> mutating/editing my data (and this is a very important point). Instead I am
> modified using position(s) (TxTextPosition) and/or selection(s)
> (TxInterval/TxSelection), providing a rich protocol for various operations
> over text." That sounds intriguing, but why is that so significant?

I think it’s related to TxText focus on locality: since you want that every 
operation takes constant time you cannot provide operations that take indexes.
Instead you use positions/cursors and intervals/selections because they know to 
which spans they correspond. 

> - TxForeColorAttribute - Why not TxFontColorAttribute? Is the fore/font
> distinction important?

Maybe it’s just to be symmetric with TxBackColorAttribute. 
IMO you’re right, TxBackgroundColorAttribute and TxFontColorAttribute are more 
intention revealing.

> - TxXyzAttribute - can we remove the 'Attribute' from all but the base
> class? e.g. TxFontAttribute -> TxFont

Maybe. It would be more concise, but a TxFontAttibute is not a font, it just 
refers to a font. 

> - TxStyle - why both #at: and #get:?

I guess #get: is what you want almost all the time. 
We should check where #at: #at:ifAbsent: is used.

> 
> Then a bit of cleaning...
> Issue 15475: TxText Cleanup for Pharo 5.0 #2
> https://pharo.fogbugz.com/default.asp?15475
> - TxXyzMorphs
>    - Push up #openInWindowWithString:, #openInWindowWithText:, and #text,
> which each had two materially-identical implementations
>    - Clean examples
> - TxModel - enhance class comment
> - TxStyle
>    - Enhance class comment
>    - Clean #removeAll comment
> - TxTextStyle
>    - Enhance class comment
>    - Clean #addStyle: comment
>    - Clean #extend:with: comment

Great!!

> 
> 
> 
> -----
> Cheers,
> Sean
> --
> View this message in context: 
> http://forum.world.st/TxText-More-Cleaning-and-Questions-tp4823894.html
> Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.
> 


Reply via email to