Closed josephsl closed 8 years ago
This is on the agenda. There's a small API change too, so some kind of version bump has to happen. My preference is 3.0.0. Also, I believe there are plans to release an alpha version soon (unless I get lost in this code again). The impact on table authors is relative, since some of these opcodes either didn't exist before or existed but were broken to the point where few tables used them, instead using e.g. multipass rules. On the other hand there are quite some renames and under-the-hood changes, so some break-in time will be required.
Thanks Joseph. The idea is indeed to move to 3.0.0, starting with a 3.0.0-alpha (not 2.9.0).
But are you suggestion we keep making 2.6.x releases in parallel to the 3.0.x releases for a few iterations?
Hi Bert, I’d say for at least six months so people can transition to the new opcodes and such (maybe 2.6.7, announce 3.0.0, 2.6.8, 3.0.0 or a similar order). Thanks.
From: Bert Frees [mailto:notifications@github.com] Sent: Thursday, February 4, 2016 2:13 AM To: liblouis/liblouis liblouis@noreply.github.com Cc: josephsl joseph.lee22590@gmail.com Subject: Re: [liblouis] Set LibLouis version to 2.7.0 or 3.0.0 once emphasis overhaul and other work are implemented (#156)
Thanks Joseph. The idea is indeed to move to 3.0.0, starting with a 3.0.0-alpha (not 2.9.0).
But are you suggestion we keep making 2.6.x releases in parallel to the 3.0.x releases for a few iterations?
— Reply to this email directly or view it on GitHub https://github.com/liblouis/liblouis/issues/156#issuecomment-179749091 .
The plan is to release a 3.0 alpha once the ueb stuff has been integrated
Hi,
The work done by Mike, Davy and others on emphasis unification and other work seems promising. Since we're dealing with new opcodes and other work, let's bump LibLouis version to 2.7.0 or 3.0.0 (I'd vote for the latter).
Benefits and reasons:
Disadvantages and costs:
However, given the amount of work that's going under the hood in LibLouis to support UEB standard properly and since the new opcodes looks promising (ability to define emphasis class for additional items such as foreign language indicator and so on), as a table maintainer and a contributor to a screen reader project, I would vote for us to move to 3.x.y version number (starting with 3.0.0).
Implementation strategy:
Possible issues to think about:
Thanks.