forking Markdown.pl?

[Tomas Doran]

On 28 Feb 2008, at 23:04, Yuri Takhteyev wrote:


>>  I'd like to get to a point where I'm a little more happy with the

>>  code, and then I'll start promoting this as a 'true' fork, or, if I

>>  can get John to agree and approve - I'd like to become the  

>> 'official'

>>  maintained version which is linked from daringfireball.

>

> I am sure Markdown.pm will make life much simpler for those using

> Markdown with Perl.  As a maintainer of a markdown module in a

> different language, however, I am not as excited about the idea of a

> new "official" implementation.


Can we change the word "official" for the word "reference", which is  
much more in line with what I actually meant.


> In fact, I think a new official

> implementation is the last thing we need, since in the absence of a

> clear spec, a new official implementation would be a source of great

> confusion, especially if the new implementation suddenly adds a large

> number of undocumented features.


Indeed.

1) I'm not adding features to the Text::Markdown language.
2) I do however support a number of options (most of the things that  
were hard coded globals in Markdown.pl can now be configured). These  
*are all documented*, and have unit tests.


>   Would this now make all of "Markdown

> Extra" official?  (And give us two under-defined specs instead of

> one.)  Or just those parts of Markdown Extra that Markdown.pm

> implements?


The latter, given that it doesn't implement *any* of them :)


>   And if Markdown.pm keeps evolving (which it should), does

> this mean that we would now be on the hook for diffing Markdown.pm

> code daily to find out what new features has become official?


Nope, I think that a 'formal' channel for discussing things like this  
is needed. That'd probably be this list. The thing that we're missing  
is a 'formal' / 'official' way to get a conclusion from these  
discussions reached and documented.


>   Note

> that I am not against the new features in Markdown Extra.  I added

> some of them in markdown.py and I would add more or even all if we

> could agree to make them official.  But I want to implement them

> against a spec, not against a perl module.  I also think some of those

> features should be discussed first.


Totally agree. However, the fact that Markdown.pl *is* the reference  
implementation (despite it's known bugs), is very sad.

I'd really like a spec (or set of specs, for different dialects) that  
we could sanely write code against, however I strongly feel that the  
reference implementation should stick like glue to the original  
Markdown feature set, and only fix bugs and inconsistencies.


> Perhaps there is a need for a better _perl_ implementation (or a few,

> competition is fun), but as far as "official" goes, we need a

> comprehensive and up-to-date spec and a test suite against which all

> implementations could be measured.


Here here, I *strongly* agree with this. One of the main things I've  
been trying to do is get a really good test suite together which can  
be used as a yardstick, and I'll happily jump on the bandwagon and  
contribute (and code to) any community effort for a better spec. :)

Cheers
Tom