Very elegantly said, Lex.

having multiple implementations is a good thing
>

I absolutely agree. While I've certainly invested a lot of time and effort
in Asciidoctor, I care about AsciiDoc the language (as well as the content
written in it) just as much. And I believe a Python implementation is an
important part of that story. I have no intention of keeping AsciiDoc
coupled to a specific language platform (Ruby or otherwise). In fact,
multiple compatible implementations can keep the language vibrant and
evolving.

I'm very excited about the journey ahead. The AsciiDoc community has a long
and rich history, yet I hope it's just the beginning of a much longer one.
It's a community I'm proud to be a part of.

Best Regards,

-Dan

On Sat, Jan 5, 2019 at 2:49 AM Lex Trotman <[email protected]> wrote:

> So to summarise:
>
> The current situation is that there are two versions of Asciidoc
> Python, where the Python 3 is a port (ie a minimal set of changes)
> from Python 2 to allow the project to continue past the EOL of
> Python2.  The Python 3 port is the way forward and I would not expect
> much to change in the Python 2 implementation.  Asciidoc Python is not
> very "active", it is a volunteer project and contributions are welcome
> for:
>
> - testing (there are proposed PRs available to test, hint hint)
> - proposals for changes (such as adding functionality Asciidoctor has
> added thus improving compatibility), and
> - infrastructure (Eric has already kindly volunteered to help with the
> website)
>
> There is another personal Python 3 port as well which (IIUC) was made
> for the interest of the developer, who has noted that they can't offer
> long term support, so it isn't the way ahead in its current form.  It
> also started badly because initially it illegally changed the license
> of the files it copied from Asciidoc Python, which has made its
> relationship with the rest of the community rather fraught, but
> hopefully improving, as the developer has proposed adding their
> improvements to Asciidoc Python 3 so only one version needs to be
> supported.
>
> The most active Asciidoc implementation is Asciidoctor.  It is a
> separate greenfields implementation and not any sort of "port" of the
> Python implementation.  As such its internals are very different and
> its mechanisms for extensions and configurations are different.
>
> It is implemented in a number of environments (Ruby, JVM, JS being the
> main ones) and covers a wide range of use-cases and that means it
> meets the needs of many people, but as noted in the thread, it doesn't
> meet everybody's needs. Thats why having multiple implementations is a
> good thing.   Even though much more active than Asciidoc Python,
> Asciidoctor is still resource limited so its not going to get a new
> website every week.  But Dan has noted that a new site is imminent.
>
> Asciidoc has traditionally had only one, or at least a leading
> implementation, but the community acknowledges that to support
> multiple implementations an implementation independent standard needs
> to be defined for Asciidoc.  That is proposed to be hosted under the
> asciidoc.org URL with links to the various implementations which of
> course will have their own sites and documentation.  This work should
> start this year if resources allow.
>
> A standard will of course define the language and allow for future
> evolution in a controlled manner and allow for multiple
> implementations that address various different use-cases.
>
> It is proposed that the standard be under asciidoc.org so it is
> prominently the "reference" for the language, but that does mean some
> changes to the current infrastructure of the asciidoc python project
> that the asciidoc.org currently points to.  As Dan has noted the
> mechanism of creating the website for Asciidoc Python is not well
> understood, and as its about to have a change now is a perfect time
> for it to be updated and hopefully Erics expertise can extend to that
> as well as tidying up the site, or others can help him.  (on a
> pragmatic note, PRs to Asciidoc Python 3 please).
>
> Dan has also noted that he is about to provide a proposal for the way
> ahead that hopefully addresses this change process, so there will be
> something more concrete to discuss soon.
>
> Finally a philosophical point if I may, these are fairly seismic
> shifts, from one implementation to another leading to a standard with
> multiple implementations.  The Asciidoc community has so far traversed
> the shifting landscapes in a fairly measured and friendly manner.  We
> have different backgrounds and expertise, and don't agree on
> everything, and that might become more apparent as we move into the
> next major change. But the fact that everybody is talking about it and
> cooperating is a great example for other communities.
>
> Cheers
> Lex
>
> --
> You received this message because you are subscribed to the Google Groups
> "asciidoc" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To post to this group, send email to [email protected].
> Visit this group at https://groups.google.com/group/asciidoc.
> For more options, visit https://groups.google.com/d/optout.
>


-- 
Dan Allen | @mojavelinux | https://twitter.com/mojavelinux

-- 
You received this message because you are subscribed to the Google Groups 
"asciidoc" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/asciidoc.
For more options, visit https://groups.google.com/d/optout.

Reply via email to