Thanks for the heads up, Jeff!
-- Sriram
On 1/11/12, Jeffrey Johnson n3...@mac.com wrote:
The @rpm5.org project uses Launchpad for ROADMAP planning
(and bug reporting and release purposes) here:
http://launchpad.net/rpm
While its anyone's guess whether Mandriva will close
its doors on January 16th (I think not), Launchpad
is also in use for planning the deployment of RPM
features over the Mandriva/ROSA 2012 release schedule.
So both the @rpm5.org and ROSA teams are using Launchpad
while sorting out what features should be included in rpm-5.4.x
over the next 6 months at the moment. The activities SHOULD be
mostly in place by (nominally) January 20th, but realistically the
planning and resource allocation and priority/importance
scheduling will likely take the rest of January to complete.
Anyone here interested in having input into the feature set
for rpm-5.4.x is invited to participate.
The work flow at launchpad is a bit different than what
is usually done by linux distress using bugzilla. In bugzilla
there is a business service - client and implied support
with a simple work flow like
bugs - implementations
Using launchpad (as in use by @rpm5.org) the work flow is
bugs - blueprints - milestones - implementations - releases
The advantage(s) over bugzilla (or trac or wiki) are
1) blueprints permit rational planning of bugs that are in fact
complex RFE's
2) milestones and priority/importance facilities in launchpad can be
adapted
rapidly to changing needs.
With monthly releases, and continuous integration like here (currently
idle)
http://harwich.jbj.org:8010
its quite feasible to make progress towards desirable goals in a modern
software development process.
My own proposals to the immediate process are 3 flaws and 2 features
The flaws in need of fixing are
1) (mandriva) adding Disttag/Disttag has introduced an extra '-' in
*.rpm file names
2) (mandriva) the Nvra index in an rpmdb doesn't include
Epoch/Disttag/Distepoch
3) revisiting RPM ACID cost - benefit wrto legacy compatibility
The features bing proposed have been underway for several years:
1) commit to a MANDATORY *.rpm signing policy (using non-repudiable
signatures
and origin authentication)
2) continue making RPM multithreaded
I expect (from ROSA) some input on ways to improve the ISPRAS shallow
testing already in use @rpm5.org, and various other improvements in
automated testing (the ROSA people are in fact from ISPRAS with a huge
amount
of subject matter expertise and skill, just ask LSB ;-)
There will likely be benefits from rescusitating URPMI and tools
and build systems that will have impact on RPM itself.
If you wish more details, or wish to participate by prioritizing your
favorite bug, or proposing your favorite feature, then go examine
the planning that is currently (and through this month) active at
http://launchpad.net/rpm
All bugs will be added to blueprints, and there will be an attempt to
fit any additional proposals into the rpm-5.4.x release planning somehow,
depending mostly on resources and interest.
hth
73 de Jeff
__
RPM Package Managerhttp://rpm5.org
Developer Communication Listrpm-devel@rpm5.org
--
Sent from my mobile device
Belenix: www.belenix.org
Twitter: @sriramnrn
__
RPM Package Managerhttp://rpm5.org
Developer Communication Listrpm-devel@rpm5.org