Panu Matilainen wrote:
Just an off-the-cuff idea: if one trusted, or two untrusted developers vote a package ready to move to "upwards" it should be moved ? (assuming that one PUBLISH vote from untrusted developer is enough to get a package into testing/unstable)
So what yer saying is... if my friend and I, see something we want sitting in QA, I can tell him, and we can make half-hearted attempts as untrusted people to do QA and both setting the +1 publish note without actually doing the QA checklist at all....and get it out and built? Seems there has to be an emergency stop button somewhere to prevent me from not making an effort to do the QA work. Seems in the untrusted space -1 publish comments need to matter too. Because i can certainly see user interest in short-cutting the process to get packages out of QA faster, tempting the use of +1 publish inappropriately by untrusted people. But I can't really see a compelling misuse of the -1 publish comment by untrusted people, holding up packages just to be an ass, isn't really a compelling interest in more than 1 or 2 people in the userbase( well 3 counting me).
-jef
On Friday 09 January 2004 11:25 am, Jef Spaleta wrote:
So what yer saying is... if my friend and I, see something we want sitting in QA, I can tell him, and we can make half-hearted attempts as untrusted people to do QA and both setting the +1 publish note without actually doing the QA checklist at all....and get it out and built? Seems there has to be an emergency stop button somewhere to prevent me from not making an effort to do the QA work. Seems in the untrusted space -1 publish comments need to matter too. Because i can certainly see user interest in short-cutting the process to get packages out of QA faster, tempting the use of +1 publish inappropriately by untrusted people. But I can't really see a compelling misuse of the -1 publish comment by untrusted people, holding up packages just to be an ass, isn't really a compelling interest in more than 1 or 2 people in the userbase( well 3 counting me).
So you're talking about something similar to how Slashdot moderation works with points. Then if someone wants to run untested code, they just drop their 'view level' to the appropriate place. If they want solid code, they raise the threshhold and don't see the cruft. In essence, at least. I like your idea.
I think that trusted people should be able to vote negatively too; a trusted person might +1 in error occassionally. Or maybe the trusted person has an interest in seeing a favorite package go through quickly (one would hope not, but facts are facts, and people sometime lose sight of the big picture when their pet idea or package is under fire).
On Fri, 9 Jan 2004, Jef Spaleta wrote:
Panu Matilainen wrote:
Just an off-the-cuff idea: if one trusted, or two untrusted developers vote a package ready to move to "upwards" it should be moved ? (assuming that one PUBLISH vote from untrusted developer is enough to get a package into testing/unstable)
So what yer saying is... if my friend and I, see something we want sitting in QA, I can tell him, and we can make half-hearted attempts as untrusted people to do QA and both setting the +1 publish note without actually doing the QA checklist at all....and get it out and built? Seems there has to be an emergency stop button somewhere to prevent me from not making an effort to do the QA work. Seems in the untrusted space -1 publish comments need to matter too. Because i can
Sure. Of course two trusted developers "teaming up" can get packages published very quickly as well, with half-hearted QA if they wish (and I sure hope everybody respects their "trusted" status not to do it - not that I'm claiming anybody doing this but the chance is there). There's the pending-state of course where anybody can verify/veto a package.. and then the release managers who can veto it if they suspect foul play. Far from foolproof, I agree - better ideas more than welcome.
certainly see user interest in short-cutting the process to get packages out of QA faster, tempting the use of +1 publish inappropriately by untrusted people. But I can't really see a compelling misuse of the -1
Well, you're signing "this matches upstream md5sum and seems ok" by your gpg signature, it *ought* to mean something to people. And again, I'm talking about getting the packages to testing/unstable by just one untrusted vote, which anybody can veto from being published.
publish comment by untrusted people, holding up packages just to be an ass, isn't really a compelling interest in more than 1 or 2 people in the userbase( well 3 counting me).
Holding up packages just to be an ass.. well, that's another story :)
- Panu -
On Fri, 09 Jan 2004 11:25:42 -0500, Jef Spaleta wrote:
Panu Matilainen wrote:
Just an off-the-cuff idea: if one trusted, or two untrusted developers vote a package ready to move to "upwards" it should be moved ? (assuming that one PUBLISH vote from untrusted developer is enough to get a package into testing/unstable)
So what yer saying is... if my friend and I, see something we want sitting in QA, I can tell him, and we can make half-hearted attempts as untrusted people to do QA and both setting the +1 publish note without actually doing the QA checklist at all....and get it out and built?
Yes. Note though that you never know who watches your activities. It may be that someone blocks the publish votes and points out problems. The people from the build/release team may refuse to build/publish a package due to concerns, too. Also, if you don't get the review right and in the current manual build system, the package fails to build, that slows down the approval once more.