You are viewing a plain text version of this content. The canonical link for it is here.
Posted to pluto-dev@portals.apache.org by "David H. DeWolf" <dd...@apache.org> on 2004/07/30 22:57:09 UTC

Driving towards a release. . .

All,

The incubator has us listed as a "successfully incubated" project.  I guess
that means we are now able to drive towards a release.  I  finally have a
few extra cycles and am ready to help the release get out the door.  

Below is a list of (potentially) critical bugs I think we should resolve (or
decide are invalid) before cutting our first release:

PLUTO-53 - Session Sharing
PLUTO-51 - Includes
PLUTO-44 - Parameters with . causing NullPointer
PLUTO-42 - getPIDValue never set
PLUTO-27 - IllegalStateException
PLUTO-20 - getPIDValue vs. getPortletID()
PLUTO-8  - Parameters with % causes loss of all parameters

Does anyone have any other issues they think we should attempt to get in the
release?  

Any patches and/or insights on the above would be greatly appreaciated.
Otherwise, I'll start chipping away at them one at a time.

David H. DeWolf 
ddewolf@apache.org
http://www.daviddewolf.com
http://www.vivare.com



RE: Driving towards a release. . .

Posted by "David H. DeWolf" <dd...@apache.org>.
> 
> There are conflicting reports as to whether Tomcat 5 fixes this. It's 
> possible that some versions do and some don't, or else there 
> maybe other 
> factors.

Isn't this fix proprietary though?  If I'm not mistaken, servlet containers
are not required to support cross context sessions.  Is that correct?

If it is correct, I wonder if we shouldn't go ahead and create a solution
that will work in any compliant container.

> 
> I did some superficial investigation and found that I could reproduce 
> the problem on Tomcat 5.0.19, but I haven't looked further. 
> There have 
> been mailing list reports of people fixing this by going to Tomcat 5, 
> though.

I have had other issues with 5.0.19 as well.  Specifically, with Pluto, I
was receiving the following exception with a clean installation of pluto
while attempting to simply execute our test portlet:

java.lang.NullPointerException
	at
org.apache.catalina.core.ApplicationHttpRequest.getSession(ApplicationHttpRe
quest.java:546)
	at
javax.servlet.http.HttpServletRequestWrapper.getSession(HttpServletRequestWr
apper.java:260)
	at
org.apache.pluto.core.impl.PortletRequestImpl.getPortletSession(PortletReque
stImpl.java:126)
	at
org.apache.pluto.portalImpl.portlet.TestPortlet.doDispatch(TestPortlet.java:
128)
	at javax.portlet.GenericPortlet.render(GenericPortlet.java:178)
	at
org.apache.pluto.core.PortletServlet.dispatch(PortletServlet.java:205)
	at
org.apache.pluto.core.PortletServlet.doGet(PortletServlet.java:145) 
      . . .

A simple upgrade to 5.0.27 resolved the issue.


David
> 
> Nick
> 


Re: Driving towards a release. . .

Posted by Nick Lothian <ni...@mackmo.com>.
David H. DeWolf wrote:

>All,
>
>The incubator has us listed as a "successfully incubated" project.  I guess
>that means we are now able to drive towards a release.  I  finally have a
>few extra cycles and am ready to help the release get out the door.  
>
>Below is a list of (potentially) critical bugs I think we should resolve (or
>decide are invalid) before cutting our first release:
>
>PLUTO-53 - Session Sharing
>  
>

There are conflicting reports as to whether Tomcat 5 fixes this. It's 
possible that some versions do and some don't, or else there maybe other 
factors.

I did some superficial investigation and found that I could reproduce 
the problem on Tomcat 5.0.19, but I haven't looked further. There have 
been mailing list reports of people fixing this by going to Tomcat 5, 
though.

Nick


RE: Driving towards a release. . .

Posted by Carsten Ziegeler <cz...@s-und-n.de>.
It seems that we are lucky: Noel just informed me (on the infrastructure
list),
that he will try a test run tonight!!!

Carsten

> -----Original Message-----
> From: David H. DeWolf [mailto:ddewolf@apache.org] 
> Sent: Monday, August 02, 2004 1:21 PM
> To: pluto-dev@portals.apache.org
> Subject: RE: Driving towards a release. . . 
> 
> I'm off for a week long vacation today, I'll take a look at 
> cvs2svn, when I get back, unless someone else beats me to it.
> 
> David
> 
> > -----Original Message-----
> > From: Carsten Ziegeler [mailto:cziegeler@s-und-n.de]
> > Sent: Sunday, August 01, 2004 6:10 AM
> > To: pluto-dev@portals.apache.org
> > Subject: RE: Driving towards a release. . . 
> > 
> > 
> > I think we could speed up the migration by providing the dump from 
> > running the cvs2svn tool. So, if someone of us here is 
> interested in 
> > trying that out and verifiying the dump, the move would happen very 
> > fast :)
> > 
> > Carsten
> > 
> > > -----Original Message-----
> > > From: David H. DeWolf [mailto:ddewolf@apache.org]
> > > Sent: Saturday, July 31, 2004 6:19 PM
> > > To: pluto-dev@portals.apache.org
> > > Subject: RE: Driving towards a release. . . 
> > > 
> > > > 
> > > > The only thing still missing is the move to subversion 
> :( It seems 
> > > > that noone is able to help us there.
> > > > So, I think we can either try to build up some pressure
> > or stick to
> > > > CVS for the next months. In this case we have to rename our CVS 
> > > > repository which should be trivial.
> > > 
> > > Personally, I'm all for waiting the couple of months 
> simply so that 
> > > the processes and infrastructure get hammered out.
> > > There seems to have been a lot of feedback so far on some of the 
> > > other mailing lists - both positive and negative - about 
> subversion.  
> > > Waiting a couple of months may give things enough time to 
> calm down 
> > > to the point where we won't have to deal with some of the issues 
> > > that have been brought up.
> > > 
> > > That said, I've been able to look into subversin a little 
> more since 
> > > our vote about it and I'm all for moving as soon as 
> everything is in 
> > > place and we can get converted in a
> > timely manner.
> > > 
> > > David
> > > 
> > > > 
> > > > Carsten
> > > 
> > 
> 


RE: Driving towards a release. . .

Posted by Carsten Ziegeler <cz...@s-und-n.de>.
Wow! Great, David! 

Have a nice vacation!!

Carsten

> -----Original Message-----
> From: David H. DeWolf [mailto:ddewolf@apache.org] 
> Sent: Monday, August 02, 2004 1:21 PM
> To: pluto-dev@portals.apache.org
> Subject: RE: Driving towards a release. . . 
> 
> I'm off for a week long vacation today, I'll take a look at 
> cvs2svn, when I get back, unless someone else beats me to it.
> 
> David
> 
> > -----Original Message-----
> > From: Carsten Ziegeler [mailto:cziegeler@s-und-n.de]
> > Sent: Sunday, August 01, 2004 6:10 AM
> > To: pluto-dev@portals.apache.org
> > Subject: RE: Driving towards a release. . . 
> > 
> > 
> > I think we could speed up the migration by providing the dump from 
> > running the cvs2svn tool. So, if someone of us here is 
> interested in 
> > trying that out and verifiying the dump, the move would happen very 
> > fast :)
> > 
> > Carsten
> > 
> > > -----Original Message-----
> > > From: David H. DeWolf [mailto:ddewolf@apache.org]
> > > Sent: Saturday, July 31, 2004 6:19 PM
> > > To: pluto-dev@portals.apache.org
> > > Subject: RE: Driving towards a release. . . 
> > > 
> > > > 
> > > > The only thing still missing is the move to subversion 
> :( It seems 
> > > > that noone is able to help us there.
> > > > So, I think we can either try to build up some pressure
> > or stick to
> > > > CVS for the next months. In this case we have to rename our CVS 
> > > > repository which should be trivial.
> > > 
> > > Personally, I'm all for waiting the couple of months 
> simply so that 
> > > the processes and infrastructure get hammered out.
> > > There seems to have been a lot of feedback so far on some of the 
> > > other mailing lists - both positive and negative - about 
> subversion.  
> > > Waiting a couple of months may give things enough time to 
> calm down 
> > > to the point where we won't have to deal with some of the issues 
> > > that have been brought up.
> > > 
> > > That said, I've been able to look into subversin a little 
> more since 
> > > our vote about it and I'm all for moving as soon as 
> everything is in 
> > > place and we can get converted in a
> > timely manner.
> > > 
> > > David
> > > 
> > > > 
> > > > Carsten
> > > 
> > 
> 


RE: Driving towards a release. . .

Posted by "David H. DeWolf" <dd...@apache.org>.
I'm off for a week long vacation today, I'll take a look at cvs2svn, when I
get back, unless someone else beats me to it.

David

> -----Original Message-----
> From: Carsten Ziegeler [mailto:cziegeler@s-und-n.de] 
> Sent: Sunday, August 01, 2004 6:10 AM
> To: pluto-dev@portals.apache.org
> Subject: RE: Driving towards a release. . . 
> 
> 
> I think we could speed up the migration by providing the dump 
> from running the cvs2svn tool. So, if someone of us here is 
> interested in trying that out and verifiying the dump, the 
> move would happen very fast :)
> 
> Carsten
> 
> > -----Original Message-----
> > From: David H. DeWolf [mailto:ddewolf@apache.org]
> > Sent: Saturday, July 31, 2004 6:19 PM
> > To: pluto-dev@portals.apache.org
> > Subject: RE: Driving towards a release. . . 
> > 
> > > 
> > > The only thing still missing is the move to subversion :( It seems
> > > that noone is able to help us there.
> > > So, I think we can either try to build up some pressure 
> or stick to 
> > > CVS for the next months. In this case we have to rename our CVS 
> > > repository which should be trivial.
> > 
> > Personally, I'm all for waiting the couple of months simply
> > so that the processes and infrastructure get hammered out.  
> > There seems to have been a lot of feedback so far on some of 
> > the other mailing lists - both positive and negative - about 
> > subversion.  Waiting a couple of months may give things 
> > enough time to calm down to the point where we won't have to 
> > deal with some of the issues that have been brought up.
> > 
> > That said, I've been able to look into subversin a little
> > more since our vote about it and I'm all for moving as soon 
> > as everything is in place and we can get converted in a 
> timely manner.
> > 
> > David
> > 
> > > 
> > > Carsten
> > 
> 


RE: Driving towards a release. . .

Posted by Carsten Ziegeler <cz...@s-und-n.de>.
I think we could speed up the migration by providing the dump from
running the cvs2svn tool. So, if someone of us here is interested
in trying that out and verifiying the dump, the move would happen
very fast :)

Carsten

> -----Original Message-----
> From: David H. DeWolf [mailto:ddewolf@apache.org] 
> Sent: Saturday, July 31, 2004 6:19 PM
> To: pluto-dev@portals.apache.org
> Subject: RE: Driving towards a release. . . 
> 
> > 
> > The only thing still missing is the move to subversion :( It seems 
> > that noone is able to help us there.
> > So, I think we can either try to build up some pressure or stick to 
> > CVS for the next months. In this case we have to rename our CVS 
> > repository which should be trivial.
> 
> Personally, I'm all for waiting the couple of months simply 
> so that the processes and infrastructure get hammered out.  
> There seems to have been a lot of feedback so far on some of 
> the other mailing lists - both positive and negative - about 
> subversion.  Waiting a couple of months may give things 
> enough time to calm down to the point where we won't have to 
> deal with some of the issues that have been brought up.
> 
> That said, I've been able to look into subversin a little 
> more since our vote about it and I'm all for moving as soon 
> as everything is in place and we can get converted in a timely manner.
> 
> David
> 
> > 
> > Carsten
> 


RE: Driving towards a release. . .

Posted by "David H. DeWolf" <dd...@apache.org>.
> 
> The only thing still missing is the move to subversion :(
> It seems that noone is able to help us there.
> So, I think we can either try to build up some pressure
> or stick to CVS for the next months. In this case we have
> to rename our CVS repository which should be trivial.

Personally, I'm all for waiting the couple of months simply so that the
processes and infrastructure get hammered out.  There seems to have been a
lot of feedback so far on some of the other mailing lists - both positive
and negative - about subversion.  Waiting a couple of months may give things
enough time to calm down to the point where we won't have to deal with some
of the issues that have been brought up.

That said, I've been able to look into subversin a little more since our
vote about it and I'm all for moving as soon as everything is in place and
we can get converted in a timely manner.

David

> 
> Carsten 


RE: Driving towards a release. . .

Posted by Carsten Ziegeler <cz...@s-und-n.de>.
The only thing still missing is the move to subversion :(
It seems that noone is able to help us there.
So, I think we can either try to build up some pressure
or stick to CVS for the next months. In this case we have
to rename our CVS repository which should be trivial.

Carsten 

> -----Original Message-----
> From: David H. DeWolf [mailto:ddewolf@apache.org] 
> Sent: Friday, July 30, 2004 10:57 PM
> To: pluto-dev@portals.apache.org
> Subject: Driving towards a release. . . 
> 
> All,
> 
> The incubator has us listed as a "successfully incubated" 
> project.  I guess that means we are now able to drive towards 
> a release.  I  finally have a few extra cycles and am ready 
> to help the release get out the door.  
> 
> Below is a list of (potentially) critical bugs I think we 
> should resolve (or decide are invalid) before cutting our 
> first release:
> 
> PLUTO-53 - Session Sharing
> PLUTO-51 - Includes
> PLUTO-44 - Parameters with . causing NullPointer
> PLUTO-42 - getPIDValue never set
> PLUTO-27 - IllegalStateException
> PLUTO-20 - getPIDValue vs. getPortletID()
> PLUTO-8  - Parameters with % causes loss of all parameters
> 
> Does anyone have any other issues they think we should 
> attempt to get in the release?  
> 
> Any patches and/or insights on the above would be greatly 
> appreaciated.
> Otherwise, I'll start chipping away at them one at a time.
> 
> David H. DeWolf
> ddewolf@apache.org
> http://www.daviddewolf.com
> http://www.vivare.com
> 
>