[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: ice-9/getopt-long.scm + 1.5.1 please? + time-based 1.5.x?
From: |
Neil Jerram |
Subject: |
Re: ice-9/getopt-long.scm + 1.5.1 please? + time-based 1.5.x? |
Date: |
02 Aug 2001 20:03:01 +0100 |
User-agent: |
Gnus/5.0808 (Gnus v5.8.8) Emacs/20.7 |
>>>>> "thi" == Thien-Thi Nguyen <address@hidden> writes:
thi> i'd like to urge 1.5.1 release soon so that we can get
thi> feedback for a 1.5.2 and so on. definitely, i will not be
thi> doing any more modifications in the immediate future. fwiw,
thi> i'm in favor of many 1.5.x releases and a single 1.6.0
thi> release as opposed to both 1.5.x and 1.6.x series. the
thi> latter is confusing to me.
I agree that mixing 1.5.x and 1.6.x is confusing, but I think the
right solution is to stick to 1.6.x for stable releases.
This is consistent with
- README, which says that versions with an odd middle number are
unstable development versions
- Emacs release numbers
- Linux kernel release numbers.
thi> for 1.5.x, perhaps we can try a time-based release process.
thi> this week 1.5.1, next week 1.5.2, etc. when bug rate /
thi> severity falls below a certain threshold for several weeks,
thi> we declare it 1.6.0, and then go back to feature- (or todo-)
thi> based release instead of time-based.
Time release sounds like an interesting idea for non-critical bugs,
but I think that, if we hit a bug that was critical for some OS, we'd
want to fix that and do a new release immediately. If we insist on
waiting a week, we just lose the interest of those OS users.
Neil