DIP1000

[Seb via Digitalmars-d]

On Tuesday, 30 August 2016 at 17:46:41 UTC, Andrei Alexandrescu 
wrote:
> On 08/30/2016 12:39 PM, Seb wrote:
>> On Tuesday, 30 August 2016 at 16:27:05 UTC, Lodovico Giaretta 
>> wrote:
>>> On Tuesday, 30 August 2016 at 16:12:19 UTC, Andrei 
>>> Alexandrescu wrote:
>>>> I'd like to initiate collaboration on an effort to do DIP1000
>>>> rigorously.
>>>>
>>>> [...]
>>>
>>> If I may suggest, a repository with some LaTeX code may be a 
>>> good
>>> idea, especially if the idea is to write things like << If 
>>> gamma
>>> derives a with lifetime l, then gamma derives a' with 
>>> lifetime l' >>
>>> (I mean, diagrams like the ones at page 9 of the paper). Then 
>>> you
>>> could use subscripts/superscripts to associate lifetime to 
>>> expressions
>>> in these graphs and upload on your site a nice readable pdf.
>>
>> Okay this is a bit unrelated to the original question, but 
>> it's 2016 and
>> we can do a lot better. It's quite easy to combine LaTeX and 
>> Markdowm to
>> get something easy to use and yielding pretty HTML. Have a 
>> look at the
>> sources of my blog post from last week:
>>
>> https://github.com/libmir/blog/blob/master/_posts/2016-08-19-transformed-density-rejection-sampling.md
>>
>>
>> (As I am quite familiar with the setup and would be happy to 
>> set it up -
>> including auto-deploy)
>
> Isn't it amazing how fast we got to a debate on choosing 
> tools... :o) -- Andrei

Sorry, please ignore, but it would still be nice to put it on 
Github, so that we can make pull requests. A couple of nits from 
a first pass:

- ArgumentList is not defined -> ParameterList
- The parameters is limited to 0,1,2 - on purpose? (opt -> * 
would allow arbitrary numbers of parameters)
- In contrast to FeatherweightJava the grammar allows a lot of 
weird stuff like:

ref int** foo {
    42 = 42;
    return null;
    *42 = &42;
}