Hi all, Hi Levente,
I have a question regarding the further development of the regex package. As you might have noticed, I have proposed several changes to the regex in the past two years, many of them already merged into the Trunk by Levente, but many others still lying in the inbox for around one year -- several small bugfixes and convenience functions (mostly tested), but also some larger new features such as named/non-capturing groups, Unicode backslash atoms, etc. These new features also entail major design decisions regarding matcher/link communication, parser composition, etc. (Shameless self-plug: You can use Squeak Inbox Talk to survey all open proposals).
My question is: How should we process these proposals? While I have received a number of kind reviews by Levente, my impression is that I spammed all of you with too many proposals for reviewing them in a foreseeable future. Shall I just merge them as-is or would anyone like to commit to reviewing them in-depth? I don't want to exclude anyone from the continuation of this project, and I would gladly receive more reviews, but I do understand the limited reviewer resources. While I gave my best to question my decisions and test my code, nothing of it will be 100% perfectly thought-through and free of bugs. On the other hand, weighting quality over quantity too much in this process might lead to a near-standstill of the project.
tl,dr: Shall I just merge all my regex proposals to the trunk, risking possible decrease in quality, or would anyone advise differently? I do commit to "you break it, you fix it", and if someone discovers any design flaw, we can still try to revert that design later, of course. For future proposals, I could upload them to the inbox and move the Trunk one or two weeks later.
Thanks in advance!
Best, Christoph
--- Sent from Squeak Inbox Talk ["sit-regex-inbox.png"]
