Escape some code parts in manual.

[01mf02]

This is needed to make the Markdown -> HTML conversion work with Pandoc, see #3183.

[wader]

I also always forget the man page 😬

[itchyny]

Why do you leave older versions (v1.6, v1.5...)? If you will change all the manuals to Markdown, older versions should be converted as well.

[01mf02]

Why do you leave older versions (v1.6, v1.5...)? If you will change all the manuals to Markdown, older versions should be converted as well.

I originally converted the v1.7 manual during my tests, but I then noticed that the dev manual had some new content compared to v1.7, so I converted that as well.
However, as I've tried to motivate in #3183 (comment), I think the best solution would be to replace the current version-specific manuals by a single manual that documents all versions. That way, all versions of jq can benefit from an up-to-date manual, and maintenance becomes much easier. I'm motivated to do this work myself, starting from the dev manual.

[pkoppstein]

@01mf02 wrote:

all versions of jq can benefit from an up-to-date manual

Do you mean going all the way back to v1.3? That would be an enormous amount of work, and might well make the manual very difficult to understand. One of the nice things about the manual, even now, is that it is quite succinct.

Perhaps with hyperlinks it would be possible to make the result readable, but the risk of introducing inaccuracies would be substantial. There's also the risk that all the work that you put into this project might not reach a critical mass of acceptance.

Since I don't understand all the issues involved, let me just ask whether there isn't a less onerous and risky alternative.

[01mf02]

Do you mean going all the way back to v1.3? That would be an enormous amount of work, and might well make the manual very difficult to understand. One of the nice things about the manual, even now, is that it is quite succinct.

Perhaps with hyperlinks it would be possible to make the result readable, but the risk of introducing inaccuracies would be substantial. There's also the risk that all the work that you put into this project might not reach a critical mass of acceptance.

Since I don't understand all the issues involved, let me just ask whether there isn't a less onerous and risky alternative.

Indeed, I think I found such an alternative. See #3186.

[pkoppstein]

@01mf02 wrote:

Indeed, I think I found such an alternative. See #3186.

Excellent! Thank you.