continuum-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jesse McConnell" <jesse.mcconn...@gmail.com>
Subject Re: Preparing for continuum-1.1-alpha-1
Date Tue, 13 Mar 2007 18:02:39 GMT
On 3/13/07, Carlos Sanchez <carlos@apache.org> wrote:
> A bit of (constructive) criticism ahead

> These are the numbers I see in jira right now
> 1.1-alpha-1     11
> 1.1-alpha-2     72
> 1.1-alpha-#     156

that is correct, and I frankly don't think there is a hill of beans
chance of resolving all of those.  The alpha-1 numbers are mostly
trivial issues that just need to be looked at, the alpha-2 are ones
that I figure we really ought to be addressing since they (largely)
address functionality that ought to be fixed or exist.  I see the
alpha-2 ones getting resolved on a hopefully weekly/fortnight
schedule, pushing those issues not resolved forward to alpha-3 until
they are either resolved, or moved into that alpha-# version.  If
anyone has an issue in alpha-# they really want resolved they can fix
it and put it in the alpha-2 or 3 or live with it not making it into
release.

perhaps I should rename the alpha-# to 'bucket', stuff that someone
had a problem with and ought to be fixed, that we'll get to if we can,
but its not a blocker or anything.

but you are correct, no way all that will get done

> Am I the only one that thinks we are lost?
> 230+ issues before having even a beta?
> do we want to release the final 1.1 in 2050?

I was thinking 2039, just in time for us to run out of long timestamps
or something



> it shouldn't take us long to get dependencies released, this is our
> process, if the process needs to be improved we can talk about it, but
> apply it only sometimes doesn't sound right
>

I would like us to get out of the habit of forcing releases of
alpha-foo all over the place just to release something that
is...alpha.

so how do we fix _that_? :)


> if nobody takes care of an issue or there's no patch that issue should
> be pushed, period.
> This is a community and if people don't help on their own issues
> and/or they can't convince somebody to help too bad for them.

this is where you hit upon a problem I have with doing issue
triage...I have a real hard time just closing something as 'Can't
Reproduce' or 'As Designed' or whatever...so I just shoved it into
alpha-#...Future was easy, those were new things to design or
whatever...


> i would get alpha 1 out of the door right now, as long as there is no
> blocker issues, I think that's the hint, if there are no blocker
> issues the release can't be blocked, and many people have already
> asked about making a release.
>
>
> This is how it looks to me from outside, as I didn't have time lately
> to work on Continuum.

to be perfectly honest, I think there are a number of smelly parts
inside of continuum right now, I have worked out designs and written
mails on the subject a number of times to this list, but I haven't had
the time to just sit down and fix the things I find most annoying, I
tried with rahul back around xmas but I wasn't able to focus on it
like I had hoped...so one of my prime motivations in pushing to get
this released as an alpha is to, as trygve said, just get something
out there since its been so long.  There are a lot of nice
improvements in continuum, but there are a chunk of nasties that I
think with the proper time and some motivation we can get ironed out.

anyway, thats enough for me...

jesse

p.s. and no one has mentioned capturing all the changes into
documentation yet...omg that will be a pain...

Mime
View raw message