0.24.2 rc2

[Luke Kanies]

Hi all,

I think I've fixed all of the first round of problems people have
found, and I think we're ready for 0.24.2, but please make another run
through with the current 0.24.x HEAD if you can.

And yes, I know there are some tests failing when you run them all at
once, complaining about missing providers -- I'm going to migrate
these tests to rspec once the release is out, but I don't want to
delay the release to fix this problem.

--
Someday I want to be rich. Some people get so rich they lose all
respect for humanity. That's how rich I want to be. --Rita Rudner
---------------------------------------------------------------------
Luke Kanies | http://reductivelabs.com | http://madstop.com

[David Schmitt]

Luke Kanies schrieb:

> Hi all,
>
> I think I've fixed all of the first round of problems people have
> found, and I think we're ready for 0.24.2, but please make another run
> through with the current 0.24.x HEAD if you can.
>
> And yes, I know there are some tests failing when you run them all at
> once, complaining about missing providers -- I'm going to migrate
> these tests to rspec once the release is out, but I don't want to
> delay the release to fix this problem.

I've upgraded my puppetmaster (almost) successfully to current 0.24.x's
HEAD and it runs my configuration (most clients still on 0.24.1, I
tested a few against HEAD though).

After upgrading I still experienced #1083 though. Looking at my local
puppet repo, I noticed that I forgot to submit my last incarnation of
the patch. So I did that:

http://reductivelabs.com/trac/puppet/ticket/1083#comment:6

The change seems innocent enough: use mount.file_path() to expand the
path for the Metadata, instead of relying on the Metadata itself to
build the path.

Another thing that should probably be mentioned in the release notes (if
it is not a bug), is this:

> notice: //.../File[/path/file]/checksum: checksum changed '{time}Sat
Oct 27 16:13:41 +0200 2007' to '{mtime}Sat Oct 27 16:13:41 +0200 2007'

Which seems to happen for all files with checksum=>mtime, triggering of
course all kinds of updates in my manifest.

Regards, DavidS

[Luke Kanies]

On Mar 1, 2008, at 4:29 AM, David Schmitt wrote:

>
> Luke Kanies schrieb:
>> Hi all,
>>
>> I think I've fixed all of the first round of problems people have
>> found, and I think we're ready for 0.24.2, but please make another
>> run
>> through with the current 0.24.x HEAD if you can.
>>
>> And yes, I know there are some tests failing when you run them all at
>> once, complaining about missing providers -- I'm going to migrate
>> these tests to rspec once the release is out, but I don't want to
>> delay the release to fix this problem.
>
> I've upgraded my puppetmaster (almost) successfully to current
> 0.24.x's
> HEAD and it runs my configuration (most clients still on 0.24.1, I
> tested a few against HEAD though).
>
> After upgrading I still experienced #1083 though. Looking at my local
> puppet repo, I noticed that I forgot to submit my last incarnation of
> the patch. So I did that:
>
> http://reductivelabs.com/trac/puppet/ticket/1083#comment:6
>
> The change seems innocent enough: use mount.file_path() to expand the
> path for the Metadata, instead of relying on the Metadata itself to
> build the path.

Okay, this is done in the release.

>
> Another thing that should probably be mentioned in the release notes
> (if
> it is not a bug), is this:
>
>> notice: //.../File[/path/file]/checksum: checksum changed '{time}Sat
> Oct 27 16:13:41 +0200 2007' to '{mtime}Sat Oct 27 16:13:41 +0200 2007'
>
> Which seems to happen for all files with checksum=>mtime, triggering
> of
> course all kinds of updates in my manifest.

But I missed this. :/

Guess I should have read the whole email.

What should I do about the fact that I missed it?

--
Somewhere on this globe, every ten seconds, there is a woman giving
birth to a child. She must be found and stopped. -- Sam Levenson