09 09:23:49 -0700
> Subject: Re: zip problems with 2.1.8
> From: musa...@gmail.com
> To: dev@struts.apache.org
>
> If we can't get the help from people in the user list, then I would
> say we do what you are suggesting, but we take it to the "extreme" a
> bit, mean
If we can't get the help from people in the user list, then I would
say we do what you are suggesting, but we take it to the "extreme" a
bit, meaning, that when we think the code is ready to be frozen for a
release, we cut a build(with tag, staging repo and all) and let people
in dev@ know about it
On Thu, Oct 1, 2009 at 10:50 PM, Wes Wannemacher wrote:
> On Fri, Oct 2, 2009 at 1:25 AM, Martin Cooper wrote:
>> On Thu, Oct 1, 2009 at 10:00 PM, Musachy Barroso wrote:
>>> I still don't understand why we don't let users know that there is a
>>> build that we are testing so we get more eyes on
I have always liked that option better, I don't see the point of
calling GA something that has been tested for 3(+/-) days, instead of
voting on something that has been in use for a couple of months and is
known to be good for sure. When you say it is a "policy", do you mean
an Apache policy? or a
I'm on Windows, and I did test it after you reported the failure. I
tested it on Win with 7-zip, and didn't notice any errors...
Nils-H
> Well, I did notice that the zips were malfunctioning (see my previous
> email), but I assumed it was a local problem (bad harddrive with many
> corrupted secto
On Fri, Oct 2, 2009 at 1:26 AM, Martin Cooper wrote:
> On Thu, Oct 1, 2009 at 10:44 AM, Musachy Barroso wrote:
>> that sounds good to me. If we just overwrite 2.1.8 then the mirrors
>> will also be updated right? This is technically speaking 2.1.8.
>
> The answer to that is "maybe". Some mirrors
On Fri, Oct 2, 2009 at 1:25 AM, Martin Cooper wrote:
> On Thu, Oct 1, 2009 at 10:00 PM, Musachy Barroso wrote:
>> I still don't understand why we don't let users know that there is a
>> build that we are testing so we get more eyes on it, before we call it
>> a GA. Is there any practical reason?
On Thu, Oct 1, 2009 at 10:00 PM, Musachy Barroso wrote:
> I still don't understand why we don't let users know that there is a
> build that we are testing so we get more eyes on it, before we call it
> a GA. Is there any practical reason? or is it just the way it has
> always been done?
It is sup
Musachy,
Policy is that the user list is only notified for releases. For a
release to occur, it needs at least 3 +1 binding votes.
Maybe one option is to introduce a graded release promotion. First,
eliminate the possibility to vote GA in the first round; it becomes
either Beta or the version is
I still don't understand why we don't let users know that there is a
build that we are testing so we get more eyes on it, before we call it
a GA. Is there any practical reason? or is it just the way it has
always been done?
musahcy
On Thu, Oct 1, 2009 at 8:19 PM, Wes Wannemacher wrote:
> I was s
I was sort of thinking the same thing... I know I'll check the docs
zip in the future, but I think it's a legitimate mistake that most of
us aren't looking in the docs zip (since we've all already read them
all, cover to cover, right?) :)
-Wes
On Thu, Oct 1, 2009 at 11:15 PM, Musachy Barroso wro
>. In fact, I
> would ask those who voted on 2.1.8 to look at how they tested before
> they voted, and perhaps think about ways in which they might change
> their testing so that we can catch something like this before it goes
> out in a release again.
Hey my windows partition is just for playing
On Thu, Oct 1, 2009 at 10:44 AM, Musachy Barroso wrote:
> that sounds good to me. If we just overwrite 2.1.8 then the mirrors
> will also be updated right? This is technically speaking 2.1.8.
The answer to that is "maybe". Some mirrors will update, others will
not. This is one of the main reasons
Wendy,
Your logic about repacking everything or nothing makes sense. Agreed.
Lastly, unless there needs to be changes within SVN (like correcting
Maven configuration to fix the build), I see no reason for a new
release. Were there commits to fix something? If so, that satisfies
me. Otherwise, it'
"The whole release" would be both the distribution zips and the Maven
artifacts. If we're going to retract the release, we should retract
*all* of it. I'm not in favor of that since the code is fine, we just
have a packaging problem with the documentation.
As I understand the proposed solution,
Wendy, I did say in a later email I wrongly believed we were talking
about Maven repos. If it is just the Apache distributions, and those
have bogus files, those should be deleted. Just can the whole release.
I think that's sensible.
Paul
On Thu, Oct 1, 2009 at 3:00 PM, Wes Wannemacher wrote:
>
On Thu, Oct 1, 2009 at 3:51 PM, Wendy Smoak wrote:
> On Thu, Oct 1, 2009 at 12:39 PM, Paul Benedict wrote:
>> I still advocate deleting the 2.1.8 binaries. Will that be done? Any
>> bad distribution should get the shovel.
>
> As I understand it, there's nothing wrong with the artifacts in the
> M
On Thu, Oct 1, 2009 at 12:39 PM, Paul Benedict wrote:
> I still advocate deleting the 2.1.8 binaries. Will that be done? Any
> bad distribution should get the shovel.
As I understand it, there's nothing wrong with the artifacts in the
Maven repo, and no reason to delete them.
This is a problem w
I still advocate deleting the 2.1.8 binaries. Will that be done? Any
bad distribution should get the shovel.
Paul
On Thu, Oct 1, 2009 at 2:05 PM, Wes Wannemacher wrote:
> I've already branched and made the changes to the poms (to make the
> whole thing build correctly). To me, I'd rather get it
I've already branched and made the changes to the poms (to make the
whole thing build correctly). To me, I'd rather get it done right than
to just push out a few updated zips, especially since this includes
the struts-2.1.8-all.zip. I'll have the build pushed out to the
staging repo in a bit and I
Oh, okay. I misunderstood. If only the Apache distribution packages
are messed up, go ahead and replace those. I don't think that calls
for a new build.
I had bad MD5 files before. I was asked to fix those on a release
(after the mirrors received them), and I did.
Paul
On Thu, Oct 1, 2009 at 12:
that sounds good to me. If we just overwrite 2.1.8 then the mirrors
will also be updated right? This is technically speaking 2.1.8.
musachy
On Thu, Oct 1, 2009 at 10:30 AM, Wes Wannemacher wrote:
> So, I was trying to figure out what is going on here because there are
> more problems with the zi
Really, this doesn't affect maven, it is the zips (which I don't think
are usable as maven dependencies). All of the jars are fine. I would
suggest signing a new copy of the struts-2.1.8-docs.zip file, but
since -all.zip also includes the contents, I figured that it's
probably best to just start ov
If the artifacts are bad, version 1.2.8.1 should be deleted from the
repository. What I've been reading on the Maven Developer's List, this
kind of issue is probably the one acceptable time to remove a version.
Paul
On Thu, Oct 1, 2009 at 12:30 PM, Wes Wannemacher wrote:
> So, I was trying to fi
Sorry, I meant 1.2.8 should be removed.
On Thu, Oct 1, 2009 at 12:35 PM, Paul Benedict wrote:
> If the artifacts are bad, version 1.2.8.1 should be deleted from the
> repository. What I've been reading on the Maven Developer's List, this
> kind of issue is probably the one acceptable time to remo
So, I was trying to figure out what is going on here because there are
more problems with the zip than mentioned on the user@ list. Since I
took site-deploy out of the release plugins default goals (to keep
other artifacts from stomping the main struts.apache.org site), it had
the unintended side e
26 matches
Mail list logo
