Posted by
jonwimp on
URL: http://ofbiz.116.s1.nabble.com/release4-0-OFBIZ-1106-in-or-out-tp185421p185448.html
Call me reckless, but I don't think that bug should block binary release. As mentioned before,
obsolete a bad release if we need to. I'll be testing the framework a lot over the coming month,
so we might see another binary release if I happen to dig up enough critical bugs.
As the release branch is now, it is functional. No show-stopping bugs. Well, it runs. We'll let
the world-wide community catch the bugs in the binary release.
> As for the bug you mentioned... that is on my list of suspicious things
> to look into. It may not really be a bug. The XML Jonathon presented was
> not valid and some of what he wrote seemed inconsistent, and I don't get
> the feeling from what Al wrote that he actually tested it (I may be
> wrong on this, I haven't finished looking into it, as I mentioned above).
That particular bug in HtmlWidget.java should be a bug. However, the use case is quite rare, so
that may be why nobody has caught it yet.
The XML I wrote is conceptual only. I didn't test if that worked.
The concept is like this.
A decorator includes 2 sections, "section1" and "section2".
Section "section1" is included inside of a <html-template-decorator>, aka a decorator ftl (as
opposed to a decorator screen widget).
Section "section2" is included in the context of the top-level decorator (the one first mentioned,
above).
The bug is in the rendering of the <html>, I think. It is obvious, and the fix is simple. There is
a stack push without a corresponding stack pop. Simple as that.
Hey, why am I describing this again? The detailed test case was described in another thread already.
Jonathon
David E Jones wrote:
>
> The release4.0 branch needs testing, and that is the point of this
> thread. Of course there are bugs or issues, finding them and the nature
> of them is the point of doing this. You're right that these exist and
> need attention. Whether they should block a binary release is another
> question altogether.
>
> As for the bug you mentioned... that is on my list of suspicious things
> to look into. It may not really be a bug. The XML Jonathon presented was
> not valid and some of what he wrote seemed inconsistent, and I don't get
> the feeling from what Al wrote that he actually tested it (I may be
> wrong on this, I haven't finished looking into it, as I mentioned above).
>
> -David
>
>
> On Nov 26, 2007, at 1:15 PM, BJ Freeman wrote:
>
>> as per the bug found today on
>> Re: HtmlWidget missing a MapStack pop?
>> there are bugs and ver 40 needs a good test.
>> per the commit on this bug
>> Ver 4.0 was not updated.
>>
>> :(
>>
>>
>> Jacopo Cappellato sent the following on 11/26/2007 11:52 AM:
>>> Frankly speaking, my interest for the release branch is low, I've not
>>> tested it too much and I usually suggest to clients to build their
>>> fortune on the trunk.
>>> That said it would be great to release it, if there is consensus from
>>> the community.
>>>
>>> Jacopo
>>>
>>> David E Jones wrote:
>>>>
>>>> On Nov 26, 2007, at 9:19 AM, Jacques Le Roux wrote:
>>>>
>>>>> BTW I think the time is coming to answer questions like in
>>>>>
http://docs.ofbiz.org/display/OFBADMIN/Demo+and+Test+Setup+Guide?focusedCommentId=2604#comment-2604
>>>>>
>>>>>
>>>>>
>>>>> What to you think, you developpers ?
>>>>
>>>> I think first things first...
>>>>
>>>> The first question for the release4.0 branch is: is it ready from a
>>>> code, etc perspective to be released?
>>>>
>>>> I asked a question a few weeks ago to try to determine how many people
>>>> are using the release branch and have found it sufficient for even a
>>>> "beta" label (which technically even the trunk SHOULD have, ie no one
>>>> should commit anything that isn't at least point tested)?
>>>>
>>>> It only takes a couple of hours to build the release and get it
>>>> uploaded and such. I pretty much have to do that as I'm the one who
>>>> has been signing the releases and such (it is my signature in the KEYS
>>>> file, etc).
>>>>
>>>> Before that happens we need to make sure we're ready for a release as
>>>> a community, and then the PMC needs to vote on a candidate revision in
>>>> the branch for a binary release.
>>>>
>>>> Right now I personally haven't tested it much, and I realistically
>>>> won't be able to, but I am willing to vote for it if there is enough
>>>> community feedback that it is in a good state for release. In fact,
>>>> I'd be ecstatic to see this happen! Each PMC member needs to consider
>>>> their own criteria for the binary release being ready, and right now
>>>> this is mine.
>>>>
>>>> So, that gets us back to the first things first thingy mentioned
>>>> above...
>>>>
>>>> Please comment everyone so we can get this moving forward!
>>>>
>>>> I'll leave this on the dev list for now and we can start something in
>>>> a bit on the user list if there isn't enough feedback here.
>>>>
>>>> -David
>>>>
>>>>
>>>
>>>
>>>
>>>
>