Is backward compatibility an important goal of AsciiDoctor since 1.5.0?

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Is backward compatibility an important goal of AsciiDoctor since 1.5.0?

egalitarian
1.5.0 introduced some breaking changes and was announced as the first stable release.
By backward compatibility, I mean document backward compatibility, but not API backward compatibility.

Will there be breaking changes in the foreseeable future?

Is document backward compatibility an important goal for developers?
Reply | Threaded
Open this post in threaded view
|

Re: Is backward compatibility an important goal of AsciiDoctor since 1.5.0?

mojavelinux
Administrator
Compatibility has always been an important goal of Asciidoctor. (The three main goals are probably compatibility, speed, and extensibility). I weigh every change very carefully and try to avoid changes that would break existing documents if at all possible.

Having said that, we do want to continue to evolve the language and we're not afraid to do so. As the new home of the AsciiDoc syntax, we have a very important responsibility to do right by the syntax. That means keeping the syntax modern, exploring new ways to make writing fun and easy, and breaking new ground. That's a balance I work very hard to strike.

Are there going to be breaking changes? Saying anything but yes would be to deny reality. We live in a world that constantly changes. To stay exactly the same and never change would mean to become obsolete. Even if we tried not to change, the world around us would force us to do so.

But there's a right way of doing it and a wrong way. I'll be the first to admit we need to adopt some better practices like semantic versioning and a more robust inline parser. I'm also interested in pursuing a standard for AsciiDoc, I just haven't had the time or resources to embark on that journey yet. All those things will bring more stability to the syntax over time.

Of course, the old releases will always remain available, so if you ever do run into a breaking change, you can simply go back and use an older version to convert the document (perhaps even to migrate it).

That's how I see it. I'm open to hearing other opinions. My mind is always open to being changed.

Cheers,

-Dan

On Thu, Aug 3, 2017 at 9:46 PM, egalitarian [via Asciidoctor :: Discussion] <[hidden email]> wrote:
1.5.0 introduced some breaking changes and was announced as the first stable release.

Will there be breaking changes in the foreseeable future?

How important is backward compatibility to developers?


If you reply to this email, your message will be added to the discussion below:
http://discuss.asciidoctor.org/Is-backward-compatibility-an-important-goal-of-AsciiDoctor-since-1-5-0-tp5798.html
To start a new topic under Asciidoctor :: Discussion, email [hidden email]
To unsubscribe from Asciidoctor :: Discussion, click here.
NAML



--
Dan Allen | @mojavelinux | https://twitter.com/mojavelinux
Reply | Threaded
Open this post in threaded view
|

Re: Is backward compatibility an important goal of AsciiDoctor since 1.5.0?

egalitarian
My opinion is that the document format should be versioned.
There should also be at least one decent tool to migrate from older versions to the latest.

Migrating by hand is nearly impossible when there are many documents.