OK, thanks for clarifying. So, when you asked the
question<https://groups.google.com/d/msg/web2py/lwdZ5vEMGdM/8WYviZxIWfAJ>about
using HTML5 "data-" attributes, you were asking about how they would
be supported in web3py, not how they are currently supported in web2py,
correct? I think Massimo was answering for web2py.
Anyway, there's nothing stopping us from making a change in web2py as well.
Maintaining backward compatibility doesn't prevent additions to the API --
we just can't change any old behavior (i.e., old code should continue to
work the same).
Anthony
On Thursday, December 20, 2012 7:23:23 PM UTC-5, Arnon Marcus wrote:
>
> LOL.. You got me all wrong! :)
>
> I meant no disrespect, I adore Massimo!
> Massimo - You Da Man!
>
> I was just teasing a little, is all...
> You know, the way he said that it could have sounded as if he was
> frustrated from not being able to change his own creation - it's like a
> clay-sculptor staring at his clay, saying "I wish I could mold that...
> A shame it's made out of clay..." :P
>
> And this is a web3py thread, so I think I got the context right...
>
> I know about the "-" problem not being able to get used within a
> function-argument-name, while being required as a tag-attribute-name by the
> HTML5 standard for full compliance.
>
> I'm just saying: "So what?" Who's to say that the expected arguments in
> the HTML-helpers must be regular arguments? Client-side frameworks are
> already overloading/abusing/nesting this attribute like crazy anyway, so
> why not make a special class for it to begin with? It IS a special kind of
> argument, even by HTML5's standards, why not treat it as such? I mean, it's
> not like we're stuck with the helpers the way they are, Massimo just said
> here that he's going to add context-managers to them, and even back-port
> that to web2py... My proposal doesn't even have to necessarily break
> backward compatibility...
>
>
> On Thursday, December 20, 2012 2:47:27 PM UTC-8, Anthony wrote:
>>
>> On Thursday, December 20, 2012 5:02:45 PM UTC-5, Arnon Marcus wrote:
>>
>>> Pfff...
>>>
>>> The creator of web2py says he wishes he could do web3py some other way,
>>> but can't...
>>>
>>
>> I don't think he said that. He wasn't talking about web3py at all -- this
>> is in reference to web2py. He was just noting the Python limitation
>> regarding the naming of function arguments (which the HTML helpers use as
>> attribute names). If we want attribute names that are not legal Python
>> function arguments, we either need to use the **{'attribute': 'value'}
>> syntax or some other alternative way to pass in the attributes. You have
>> proposed an alternative, but probably could have done so without the
>> attitude.
>>
>> Anthony
>>
>>>
--
