Re: [ANNOUNCE] 2.2.0-rc only a few hours off, 2.2-stable branch soon

hi

> I'm thru with the points I had on my TODO / the feature set I use and rely on 
> for 2.2. My plan is to tag 2.2.0-rc when my ppc, ppc64, x86 and x86-64 builds 
> got testinstalled the next days (read weekend).
just for curiosity, what test builds are you doing? which libc-s? which 
targets? how many packages your reference build contains? how many of 
them fail on buildtime? how many on stage 9? how many on runtime?
> Given the extreme amount of time we are overdue with 2.2 I assume all of you 
> have flushed your TODO long ago and we tag 2.2 till the end of this month.
heading my TODO for 2.2 is the cross-libc building of perl and the 
uclibc based targets with 'install' image output. then the 
multiple-target installers, which use to work on 2.1 but not anylonger 
on 2.2. excluded modules and devices list and forced modules for 
hotplug++ which works on 2.1/rockplug but not on the current hotplug++. 
also the current stone modules need to be synced with the current T2 
reality, it still offer hwscan for example. PIE support is broken, and a 
few others things that work on 2.1 but not on 2.2 which i don't have 
time to write right now but i think, after fixing those mentioned above 
for rc1 we can fix on rc2.
> If you have something todo finish it up quickly - and remeber, no big stuff 
> that breaks more than it fixes or dublicate packages.
eterm worked over e17-only here, so it was not duplicated. i don't care 
about e16, and i don't want to install it's stuff for been able to have 
eterm. you decided unilaterally to delete eterm package, now can you 
please make eterm16 {e16,e17} aware on a single package as you seem to want?
> Given that we need more PR and user / developer base and proof the IT world 
> that T2 is stable and ready to deliver systems I suggest we concentrate less 
> on rewrites (of the build system in e.g. LUA or so) and more on usability 
> up-to-dateness and so on for the next time to not fail in the historic ROCK 
> traps.
of course we need a release, but with EDONTCARE and EDONTUSE we are 
getting quiet far from 'SDE' (imo)
> Btw. I would rather see mnemosyne and archivista not in the relase tarball and 
> suggest we remove the 1st's external and the 2nd into a own repository.
go for it. and Mr. "2.2 Stable Tree Maintainer" please take a look into 
the other targets present on target/ before releasing -rc1.
Alejandro Mery