3 ms·

As a software engineer working in a data interoperability role (not that I would claim authority, but pragmatic experience): I'm not sure of the specifics but

by fsloth 6y ago

As a software engineer working in a data interoperability role (not that I would claim authority, but pragmatic experience): I'm not sure of the specifics but personally I prefer formats that don't evolve over time. So not changing a spec for over a decade should not be considered pathological but actually commendable, if the nature of spec is complete enough for it's purpose. I know vanilla Markdown is too limited for some use cases. But that is no reason to "overwrite" it.

szhu 6y ago
> I know vanilla Markdown is too limited for some use cases. CommonMark does not make Markdown less vanilla -- it's not like GFM or one of the other standards that adds support for tables or other features. From Jeff Atwood (who is one of the CommonMark creators): > The goal of CommonMark is not to redefine what Markdown is, or change the syntax, but make it parseable and predictable. https://talk.commonmark.org/t/what-changed-in-commonmark/15 https://talk.commonmark.org/t/what-changed-in-commonmark/15 Here's CommonMark's statement on why it exists: https://spec.commonmark.org/0.29/#why-is-a-spec-needed- https://spec.commonmark.org/0.29/#why-is-a-spec-needed- CommonMark's entire purpose is to fix data interoperability.
sago 6y ago
> So not changing a spec for over a decade The problem was there was no _specification_. It was a 'how to use' summary, And each implementation could be (and was) different in subtle edge cases. The point of CommonMark was to define the specification and stick to it. I agree with GP, I thought it rather sucky that he objected to using the name. Various markdowns have extension mechanisms, they always have. That's not what the GP was talking about.
lucideer 6y ago
The general idea behind your points are sound and correct. However, the problem is you seem to be generally ignorant of widely known points of knowledge about Markdown. > not changing a spec for over a decade should not be considered pathological but actually commendable, if the nature of spec is complete enough for it's purpose 100% agree in theory, but Markdown's creator never wrote any spec. when creating it. Initiatives like CommonMark are efforts to specify and unspecified language, not to evolve nor replace any existing spec.
privong 6y ago
> So not changing a spec for over a decade should not be considered pathological but actually commendable, if the nature of spec is complete enough for it's purpose. The "nature of the spec is complete enough for its purpose" is the part that's not met, though (at least in many people's minds). The Markdown "spec" (either the description written by Gruber or the `Markdown.pl` file) has ambiguities and inconsistent behavior. My understanding is that there were many requests from the community for this to be clarified, but it never was. So I think a decade of inattention is not commendable in this instance. The CommonMark landing page[0] has some more about this issue. [0] https://commonmark.org/#why https://commonmark.org/#why

by fsloth 6y ago

szhu 6y ago

> I know vanilla Markdown is too limited for some use cases. CommonMark does not make Markdown less vanilla -- it's not like GFM or one of the other standards that adds support for tables or other features. From Jeff Atwood (who is one of the CommonMark creators): > The goal of CommonMark is not to redefine what Markdown is, or change the syntax, but make it parseable and predictable. https://talk.commonmark.org/t/what-changed-in-commonmark/15 https://talk.commonmark.org/t/what-changed-in-commonmark/15 Here's CommonMark's statement on why it exists: https://spec.commonmark.org/0.29/#why-is-a-spec-needed- https://spec.commonmark.org/0.29/#why-is-a-spec-needed- CommonMark's entire purpose is to fix data interoperability.

sago 6y ago

> So not changing a spec for over a decade The problem was there was no _specification_. It was a 'how to use' summary, And each implementation could be (and was) different in subtle edge cases. The point of CommonMark was to define the specification and stick to it. I agree with GP, I thought it rather sucky that he objected to using the name. Various markdowns have extension mechanisms, they always have. That's not what the GP was talking about.

lucideer 6y ago

The general idea behind your points are sound and correct. However, the problem is you seem to be generally ignorant of widely known points of knowledge about Markdown. > not changing a spec for over a decade should not be considered pathological but actually commendable, if the nature of spec is complete enough for it's purpose 100% agree in theory, but Markdown's creator never wrote any spec. when creating it. Initiatives like CommonMark are efforts to specify and unspecified language, not to evolve nor replace any existing spec.

privong 6y ago

> So not changing a spec for over a decade should not be considered pathological but actually commendable, if the nature of spec is complete enough for it's purpose. The "nature of the spec is complete enough for its purpose" is the part that's not met, though (at least in many people's minds). The Markdown "spec" (either the description written by Gruber or the `Markdown.pl` file) has ambiguities and inconsistent behavior. My understanding is that there were many requests from the community for this to be clarified, but it never was. So I think a decade of inattention is not commendable in this instance. The CommonMark landing page[0] has some more about this issue. [0] https://commonmark.org/#why https://commonmark.org/#why