At 04:33 AM 5/10/2005, Mladen Turk wrote: >Interesting is that it was spotted only when the release >was made, so this gives one reason more for making some >sort of releases and binaries to attract more users to >actually do the testing.
Agreed that development releases (early and often) are a very good thing. They need the same 3 +1 votes (more + than -) but have a much lower barrier of entry. (Is it tar'red complete? Does it build? Do some basic regressions work?) >We can retag and release 1.2.13, or just wait if something else >will arise and have some vacation in the mean time :). I'd suggest 1) tag 1.2.13 with any bug fixes in the past week or two. Give me this evening to commit a small tweak to the Win32 .dsp project compile flags, which emit much more legible user.dmp crash analysis output. (I found this out over the weekend.) 2) let it ride - skip adding any new features till it shakes out, at least to 1.2. If someone wants to add some new stuff, make a CVS branch to develop it on, or fork 1.3 already depending on the desired patches. 3) wait for something else to arise, put together the docs on the newest features, and in, say, 3 weeks after 1.2.13 is announced, bless 1.2.14 as the next stable release. And 4) pull down 1.2.11 which was never released (according to your judgement) and 1.2.12 (which is broke, and obviously confuses users who were used to the odds-evens system of dev and stable tarballs.) If they are needed by the occasional user, archive.apache.org/dist/ is the place to look. What is the thought, is 1.2.10 stable? 1.2.8? Or 1.2.6? I'm partial to 1.2.8 myself. >Bill will love this, for sure ;) Nope - I hoped 1.2.12 would pan out, I just wasn't planning to endorse it for another three weeks or so. Bill --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
