OK.  I'm still +1 then as well.

I'd say we add this to the bug and to the next user guide and call 3.1 done.

--a.

----- Original Message ----- From: "Allen Gilliland" <[EMAIL PROTECTED]>
To: <dev@roller.apache.org>
Sent: Wednesday, April 18, 2007 9:29 AM
Subject: Re: Release Apache Roller 3.1 RC8


Yes, documentation is always a good thing, and what would probably be better is to improve the on screen description of what that "aggregated" property is really meant for.

So to answer Anil's question, 1349 is expected behavior. Basically, what Matt and I did was to check the "aggregated" box when we shouldn't have. What that checkbox effectively does is it changes the scope of the frontpage blog from being just a normal weblog into being a representation of all content from all blogs, i.e. aggregated. The reason that feeds (or pages) by category should not work on an aggregated frontpage is because we don't expect categories to span multiple blogs, so a query for all entries in the category "/Java" across multiple blogs would not be appropriate.

When Matt setup his installation he was only going to host 1 blog as his frontpage, so there is no reason to check that 'aggregated' box. I did the same thing in my test.

The other thing that would probably be the most appropriate fix to this situation is that we should really be offering 404 responses for feed urls by category on an aggregated frontpage blog to indicate that the request is to an invalid url, rather than just sending back an unexpected set of content.

I think that the usability of this feature could be improved, but I don't think we need another release candidate for that. From a functional point of view things are working properly.

I don't about 1347, it sounds like Dave submitted a fix a while ago to actually fix that one.

In any case, I am still +1 on the release.

-- Allen


Matt Raible wrote:
I think it needs to be heavily documented if we're going to release
R8. For example:

If you set a blog as site-wide, make sure you don't set it as "aggregated".

In the future, we should prevent this from happening, or make it
possible for this to work.

Matt

On 4/18/07, Anil Gangolli <[EMAIL PROTECTED]> wrote:

Allen, I'm still confused here. Are you saying that 1349/1347 are expected
behaviors or non-critical bugs?

--a.


----- Original Message -----
From: "Allen Gilliland" <[EMAIL PROTECTED]>
To: <dev@roller.apache.org>
Sent: Tuesday, April 17, 2007 12:55 PM
Subject: Re: Release Apache Roller 3.1 RC8


> okay, after reading through that jira issue i see that i was running
into
> the same problem that Matt did.  i created a test blog and set it as
the
> site-wide blog, plus set it as "aggregated", which alters its > behavior.
> so this is not a bug, just me getting confused :(
>
> the rest of my tests went fine, so i am +1 on RC8.
>
> -- Allen
>
>
> Matt Raible wrote:
>> Here's the bug for it:
>>
>> http://opensource.atlassian.com/projects/roller/browse/ROL-1349
>>
>> I believe the solution was to turn off aggregated front-page feeds.
>>
>> http://opensource.atlassian.com/projects/roller/browse/ROL-1347
>>
>> Also, this happened for a blog I had set to the front-page, don't >> know
>> if that helps.
>>
>> Matt
>>
>> On 4/17/07, Dave <[EMAIL PROTECTED]> wrote:
>>> On 4/17/07, Anil Gangolli <[EMAIL PROTECTED]> wrote:
>>> > Matt (Raible) reported a similar problem, I think on 3.0 or an
>>> earlier 3.1
>>> > candidate?  Was that ever resolved?
>>>
>>> Yes. Matt reported that but I was never able to duplicate the
problem.
>>>
>>> - Dave
>>>
>>>
>>>
>>> >
>>> > --a.
>>> >
>>> >
>>> > ----- Original Message -----
>>> > From: "Allen Gilliland" <[EMAIL PROTECTED]>
>>> > To: <dev@roller.apache.org>
>>> > Sent: Monday, April 16, 2007 9:32 AM
>>> > Subject: Re: Release Apache Roller 3.1 RC8
>>> >
>>> >
>>> > >
>>> > >
>>> > > Allen Gilliland wrote:
>>> > >> I just did some quick testing and unfortunately I have to
vote -1
>>> > >> :(
>>> > >>
>>> > >> There is a bug in the feed model which ignores category
>>> constraints, so
>>> > >> all the feeds by category are just returning all entries. >>> > >> we'll
>>> need to
>>> > >> fix that and apply it to another RC.
>>> > >>
>>> > >> I'll get the fix tested and committed in a couple minutes.
>>> > >
>>> > > Actually, I take that back, I can't seem to figure why the
category
>>> > > constraints on feeds aren't working in the 3.1 branch :/
>>> > >
>>> > > I checked the trunk and found that there was a problem there
which I
>>> > > fixed, but the 3.1 code is different and I can't seem to find >>> > > the
>>> problem.
>>> > > Can someone else verify that they see the same problem with >>> > > their
>>> > > deployment of 3.1 rc8?
>>> > >
>>> > > -- Allen
>>> > >
>>> > >
>>> > >>
>>> > >> -- Allen
>>> > >>
>>> > >>
>>> > >> Anil Gangolli wrote:
>>> > >>> +1 (3.1 RC8 only)
>>> > >>>
>>> > >>> My superficial manual smoke tests passed.
>>> > >>>
>>> > >>> - Verified sigs on all packages.
>>> > >>> - Used the tar.gz binary package and did a clean install with >>> > >>> a
>>> > >>> from-scratch db build.
>>> > >>> - Went through the initial registration and blog setup steps,
>>> posted an
>>> > >>> entry and a comment.
>>> > >>>
>>> > >>>
>>> > >>> ----- Original Message ----- From: "Dave" >>> > >>> <[EMAIL PROTECTED]>
>>> > >>> To: <dev@roller.apache.org>
>>> > >>> Sent: Friday, April 13, 2007 1:34 PM
>>> > >>> Subject: VOTE: Release Apache Roller 3.1 RC8
>>> > >>>
>>> > >>>
>>> > >>>> A new release candidate RC8 for Roller 3.1 is ready.
>>> > >>>> Here are the changes since RC7:
>>> > >>>>
>>> > >>>> - In ./WEB-INF/velocity/weblog.vm instances of
>>> > >>>> "$utilities.escapeHTML($cform.name)" have been replaced with
>>> > >>>> "$utils.escapeHTML($cform.name)"
>>> > >>>>
>>> > >>>> - The comment escaping changes have been applied to the >>> > >>>> legacy
>>> macro
>>> > >>>> file ./WEB-INF/velocity/deprecated/comments.vm
>>> > >>>>
>>> > >>>> You can get the new Roller 3.1 RC8 files here:
>>> > >>>> http://people.apache.org/~snoopdave/apache-roller-3.1
>>> > >>>>
>>> > >>>>
>>> > >>>> Committers: please test and vote on this release.
>>> > >>>>
>>> > >>>> Thanks,
>>> > >>>> Dave
>>> > >>>>
>>> > >>>
>>> > >
>>> >
>>> >
>>>
>>
>>
>






Reply via email to