[Cuis] assignment operator questions

[Casey Ransberger]

Top post. I personally prefer the arrows. I think they're beautiful and Cuis should be as beautiful as we can manage. I'd prefer (again, personally) to see them in core code, and I have a habit of replacing := in core code when I see it. 

Of course, I'm in favor of preferences for supporting := in spite of a disinterest in back (middle?) compatibility. It would be impractical not to. 

That said, I think that the arrows are a lovely part of our semiotic history, and smile whenever I see them. 

Of course: I may be fighting a losing battle on this one. On a mailing list, not long ago, Alan said something to the effect of "culture has a tendency to deposit its customs on the traveller." I can't remember exactly the context, but it was at least superficially similar to this one. 

Incidentally, I'm curious if this was one of his quips or if he was referencing something specific, so I've CC'd him on the off chance that I might find out. 

</2cents>

C

On May 23, 2012, at 11:29 AM, Daniel Lyons <fusion at storytotell.org> wrote:

> Thought I should ask since I don't think an official position has been
> taken, and now seems like a better time to ask than after having
> written a bunch of code.
> 
> In Cuis, _ is translated into ←, a left arrow, and this works as well
> for assignments as :=. Similarly, ^ is translated into ↑, an up
> arrow. I believe these behaviors originate in Smalltalk-80 and come to
> Cuis via Squeak.
> 
> Pharo took the position that the underscore would be deprecated for
> assignment, and typing it in a Pharo image produces an underscore
> rather than a left arrow, and you must use := for assignment. 
> Similarly, the caret produces a caret, though it works for 
> returning values.
> 
> My questions:
> 
> - Is Cuis taking a position on this issue?
> - Is Pharo compatibility affected by allowing this? Is that something
>   we are concerned about?
> - Is it worth considering actual translation of these characters into
>   their Unicode equivalents and enabling those instead? This raises
>   other questions about Unicode compatibility as well.
> 
> Thanks,
> 
> -- 
> Daniel Lyons
> 
> _______________________________________________
> Cuis mailing list
> Cuis at jvuletich.org
> http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org