[Cuis-dev] Naming conventions

[Luciano Notarfrancesco]

Yeah, we talked about this last time namespaces came up in this list, I
like it, and I’d be happy with a simple implementation without special
tools. Perhaps instead of aiming for a solution for the base image we
should try to implement Environments or something similar as a package.
What would be nice is that we stopped using the global Smalltalk
everywhere, specially in the tools, and started using something like
‘aClass environment’ that defaults to Smalltalk.

On Sat, 2 May 2020 at 6:28 AM, Hernan Wilkinson <
hernan.wilkinson at 10pines.com> wrote:

> Hi all,
>  interesting thread!!
>  Class name prefix are not nice but it is very easily to get used to them.
> When you got use to them when you read something like WAHtmlCanvas, your
> mind disregards the WA and focus on HtmlCanvas. Prefix do not look nice,
> but they solve the problem of name clashing in a simple way and without
> adding anything special... but namespaces look as a better "engineering"
> solution :-)
>  Many details have to be take care when implementing namespaces, it is not
> that simple. The implementation I liked about this was Geman Leiva's one,
> showed at ESUG 2010. Here is a video showing it:
> https://www.youtube.com/watch?v=n4I7fSVNX2A
>  It was a clean, Smalltalk like solution and, if I recall correctly, it
> did not have a performance hit.
>
> Cheers!
> Hernan.
>
>
> On Thu, Apr 30, 2020 at 3:56 PM Juan Vuletich via Cuis-dev <
> cuis-dev at lists.cuis.st> wrote:
>
>> On 4/29/2020 12:57 AM, Luciano Notarfrancesco via Cuis-dev wrote:
>>
>>
>>
>> On Tue, 28 Apr 2020 at 8:52 PM, Juan Vuletich <juan at jvuletich.org> wrote:
>>
>>>
>>> Without backticks, we'd need to add packages as globals, so they can be
>>> resolved at runtime. With the backticks we can avoid that, because we
>>> can give the Compiler the context to resolve package names if in a
>>> required package, just like class names, at compile time.
>>>
>>
>> But that would work regardless of back ticks, right? Erik doesn’t like
>> backticks very much, I also avoid them in my code.
>>
>> I just did some experiments changing Class>>bindingOf:. Currently it
>> first looks in class var names and SharedPools calling localBindingOf:,
>> then it looks in globals by calling Smalltalk bindingOf:. I added a
>> packageBindingOf: after localBindingOf:, and if self package is not nil it
>> tries to resolve the binding within the package. We can put required
>> packages here and they will look like globals, right? And also resolve
>> classes of the local package.
>>
>>
>> You are right! Yes, the backticks are optional. There is still the
>> message send to get the class. So there is some performance penalty, and
>> some users in critical code might still prefer to use them, so the literal
>> is the class and not the module. But it is a small detail.
>>
>> I tried one of the ideas: keeping all classes global, but adding a prefix
>> like ‘PackageName.’ to the binding key in the Smalltalk system dictionary.
>> I added a Class>>globalName too. Seems to almost work, but tools break all
>> over the place.
>>
>>
>>>
>> Cheers,
>>
>> --
>> Juan Vuletichwww.cuis-smalltalk.orghttps://github.com/Cuis-Smalltalk/Cuis-Smalltalk-Devhttps://github.com/jvuletichhttps://www.linkedin.com/in/juan-vuletich-75611b3
>> @JuanVuletich
>>
>> --
>>
> Cuis-dev mailing list
>> Cuis-dev at lists.cuis.st
>> https://lists.cuis.st/mailman/listinfo/cuis-dev
>>
>
>
> --
>
> *Hernán WilkinsonAgile Software Development, Teaching & Coaching*
> *Phone: +54-011*-4893-2057
> *Twitter: @HernanWilkinson*
> *site: http://www.10Pines.com <http://www.10pines.com/>*
> Address: Alem 896, Floor 6, Buenos Aires, Argentina
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cuis.st/mailman/archives/cuis-dev/attachments/20200502/8450db92/attachment.htm>