Versioning of CSL and the CSL spec

Basically good for me. But then current master will be ahead of 1.0.1. Will that cause trouble if someone validates against that schema version?

My thought is if people want to validate against 1.0.1, they should validate against that tag?

1 Like

Main problem I see is the sequence of the commits would then constrain the tagging. As in, if I merged intext first, that would pretty much require it be part of same tag is the more minor changes.

I guess unless we rewrite the git history, which is possible, but maybe a little risky?

I merged the variable and type addition commits.

I believe I wrote the commit messages such that all linked issues are now closed.

Any other minor changes along these lines we want to add now?

On the intext PR, I had to do some git surgery to fix it. Can you two take another look?

@bwiernik @Denis_Maier

I will take a look at intext. There are a few more variables that I think we can add in the minor release. I will open a new PR.

1 Like

@Dan_Stillman - any input on this?

Not sure what other implementers are around here, but now would be the time to speak up for those folks.

To update, I’m waiting on @bwiernik’s PR, and then plan to merge that and then the new intext element PR.

I haven’t decided yet if I’ll bundle all of this in one 1.0.2 tag, or split the latter off to 1.0.3; or simply leave master untagged for now and we can decide later.

What do you think about also adding the split title feature to 1.0.2 or to 1.0.3 together with the intext element? I think we already have a pretty sound proposal.

I haven’t had the bandwidth to follow that discussion, but will likely defer to whatever you two settle on.

1 Like

It will be non-disruptive, like in text, so 1.0.x will be fine I think.

2 Likes

Norm Walsh has brought up some issues of relevance to this discussion. Here’s his conclusion, and my reply.

If we followed his suggestion, I think we’d do a 1.0.x release with the new variables, etc., and then a 1.1 release with the changes he suggests (and maybe intext?).

We could bundle 1.1 with a formal “processing” spec or supplement, that would probably necessarily (because of time) have to start off small?