replace stax API with Geronimo-stax+Woodstox

classic Classic list List threaded Threaded
11 messages Options
Reply | Threaded
Open this post in threaded view
|

replace stax API with Geronimo-stax+Woodstox

Yonik Seeley-2
Someone brought up an issue with the StAX licensem, and after digging
through legal-discuss archives it seems the easiest way forward is to
perhaps replace it with the geronimo stax API and the woodstox
implementation.

I brought it up on legal-discuss again, but I'm not hopeful of an
unequivocal answer.
http://markmail.org/search/?q=StAX+(JSR+173)+License

Thoughts/Objections?

-Yonik
Reply | Threaded
Open this post in threaded view
|

Re: replace stax API with Geronimo-stax+Woodstox

Grant Ingersoll-2
Arg.  I'm torn.  We've released it before, which of course doesn't  
make it right.  However, if you check out the Stax code, it is all  
there and licensed as ASL.  You would think we could take their word  
for it.  However, the problem is that now someone has cast doubt on  
whether that code was truly donated.   It's further complicated by the  
fact that Oracle has now bought BEA, and much of the resources for  
JIRA, etc. listed on the mailing list page are dead and the stuff is  
virtually non-existent on Oracle's pages.

In the end, it is probably easier to just make the switch.  These last  
minute changes are not fun.

-Grant

On Sep 8, 2008, at 5:27 PM, Yonik Seeley wrote:

> Someone brought up an issue with the StAX licensem, and after digging
> through legal-discuss archives it seems the easiest way forward is to
> perhaps replace it with the geronimo stax API and the woodstox
> implementation.
>
> I brought it up on legal-discuss again, but I'm not hopeful of an
> unequivocal answer.
> http://markmail.org/search/?q=StAX+(JSR+173)+License
>
> Thoughts/Objections?
>
> -Yonik


Reply | Threaded
Open this post in threaded view
|

Re: replace stax API with Geronimo-stax+Woodstox

Shalin Shekhar Mangar
We have released with StaX earlier. We have never tested in production
environments with woodstox. Even though I doubt there will be a problem but
we can never know unless we use it. Let's not delay 1.3 anymore. We can make
this change for 1.4

On Tue, Sep 9, 2008 at 7:11 PM, Grant Ingersoll <[hidden email]> wrote:

> Arg.  I'm torn.  We've released it before, which of course doesn't make it
> right.  However, if you check out the Stax code, it is all there and
> licensed as ASL.  You would think we could take their word for it.  However,
> the problem is that now someone has cast doubt on whether that code was
> truly donated.   It's further complicated by the fact that Oracle has now
> bought BEA, and much of the resources for JIRA, etc. listed on the mailing
> list page are dead and the stuff is virtually non-existent on Oracle's
> pages.
>
> In the end, it is probably easier to just make the switch.  These last
> minute changes are not fun.
>
> -Grant
>
>
> On Sep 8, 2008, at 5:27 PM, Yonik Seeley wrote:
>
>  Someone brought up an issue with the StAX licensem, and after digging
>> through legal-discuss archives it seems the easiest way forward is to
>> perhaps replace it with the geronimo stax API and the woodstox
>> implementation.
>>
>> I brought it up on legal-discuss again, but I'm not hopeful of an
>> unequivocal answer.
>> http://markmail.org/search/?q=StAX+(JSR+173)+License<http://markmail.org/search/?q=StAX+%28JSR+173%29+License>
>>
>> Thoughts/Objections?
>>
>> -Yonik
>>
>
>
>


--
Regards,
Shalin Shekhar Mangar.
Reply | Threaded
Open this post in threaded view
|

Re: replace stax API with Geronimo-stax+Woodstox

Yonik Seeley-2
On Tue, Sep 9, 2008 at 10:12 AM, Shalin Shekhar Mangar
<[hidden email]> wrote:
> We have released with StaX earlier. We have never tested in production
> environments with woodstox. Even though I doubt there will be a problem but
> we can never know unless we use it. Let's not delay 1.3 anymore. We can make
> this change for 1.4

Careful tracking and handling of IP rights is paramount at Apache.
Knowingly releasing something with questionable IP, when a drop-in
replacement exists, is something that could bring the ASF Board
crashing down on us (rightfully so).  So unless some things are
cleared up on legal-discuss, I'm afraid I'm -1 on releasing in current
form.

(hey, is that my first -1 at Apache after all these years?)

-Yonik
Reply | Threaded
Open this post in threaded view
|

Re: replace stax API with Geronimo-stax+Woodstox

Yonik Seeley-2
In reply to this post by Shalin Shekhar Mangar
FYI, I'm testing Solr with woodstox now and will probably do some ad
hoc stress testing too.
But woodstox is a quality parser.  I expect less problems then we had
with the reference implementation (and it may even be faster too)

-Yonik
Reply | Threaded
Open this post in threaded view
|

Re: replace stax API with Geronimo-stax+Woodstox

Walter Underwood, Netflix
We've been using woodstox in production for over a year.
No problems.

wunder

On 9/9/08 8:07 AM, "Yonik Seeley" <[hidden email]> wrote:

> FYI, I'm testing Solr with woodstox now and will probably do some ad
> hoc stress testing too.
> But woodstox is a quality parser.  I expect less problems then we had
> with the reference implementation (and it may even be faster too)
>
> -Yonik

Reply | Threaded
Open this post in threaded view
|

Re: replace stax API with Geronimo-stax+Woodstox

Grant Ingersoll-2
In reply to this post by Yonik Seeley-2
Yeah, unfortunately, we can't release with IP issues like this.  
However, Woodstox actually looks to be quite nice from a quick  
glance.  It's ASL.  It's actively maintained (Stax is not).  The  
downside, of course, is we have virtually no running time w/ it.

On Sep 9, 2008, at 10:57 AM, Yonik Seeley wrote:

> On Tue, Sep 9, 2008 at 10:12 AM, Shalin Shekhar Mangar
> <[hidden email]> wrote:
>> We have released with StaX earlier. We have never tested in  
>> production
>> environments with woodstox. Even though I doubt there will be a  
>> problem but
>> we can never know unless we use it. Let's not delay 1.3 anymore. We  
>> can make
>> this change for 1.4
>
> Careful tracking and handling of IP rights is paramount at Apache.
> Knowingly releasing something with questionable IP, when a drop-in
> replacement exists, is something that could bring the ASF Board
> crashing down on us (rightfully so).  So unless some things are
> cleared up on legal-discuss, I'm afraid I'm -1 on releasing in current
> form.
>
> (hey, is that my first -1 at Apache after all these years?)
>
> -Yonik


Reply | Threaded
Open this post in threaded view
|

Re: replace stax API with Geronimo-stax+Woodstox

Shalin Shekhar Mangar
In reply to this post by Yonik Seeley-2
On Tue, Sep 9, 2008 at 8:27 PM, Yonik Seeley <[hidden email]> wrote:

>
> Careful tracking and handling of IP rights is paramount at Apache.
> Knowingly releasing something with questionable IP, when a drop-in
> replacement exists, is something that could bring the ASF Board
> crashing down on us (rightfully so).  So unless some things are
> cleared up on legal-discuss, I'm afraid I'm -1 on releasing in current
> form.
>
> (hey, is that my first -1 at Apache after all these years?)
>

Well, alright then. I hope we will be able to release 1.3 this week. I'd
really really like that.

Grant, if there is anything I can do to help, please let me know.

--
Regards,
Shalin Shekhar Mangar.
Reply | Threaded
Open this post in threaded view
|

Re: replace stax API with Geronimo-stax+Woodstox

Yonik Seeley-2
In reply to this post by Yonik Seeley-2
I've updated trunk and all tests pass.  Other people can start testing
via trunk now if they want.
I'm in the process of updating NOTICE.txt

-Yonik

On Tue, Sep 9, 2008 at 11:07 AM, Yonik Seeley <[hidden email]> wrote:
> FYI, I'm testing Solr with woodstox now and will probably do some ad
> hoc stress testing too.
> But woodstox is a quality parser.  I expect less problems then we had
> with the reference implementation (and it may even be faster too)
>
> -Yonik
>
Reply | Threaded
Open this post in threaded view
|

Re: replace stax API with Geronimo-stax+Woodstox

thorsten
In reply to this post by Grant Ingersoll-2
On Tue, 2008-09-09 at 09:41 -0400, Grant Ingersoll wrote:
> Arg.  I'm torn.  We've released it before, which of course doesn't  
> make it right.  However, if you check out the Stax code, it is all  
> there and licensed as ASL.  You would think we could take their word  
> for it.  However, the problem is that now someone has cast doubt on  
> whether that code was truly donated.   It's further complicated by the  
> fact that Oracle has now bought BEA, and much of the resources for  
> JIRA, etc. listed on the mailing list page are dead and the stuff is  
> virtually non-existent on Oracle's pages.

The problematic part was (thanks to Yonik who fixed this so quick) the
API. The API is from Sun and not from BEA and does not have an ASL.

However using woodstox and the geronimo rewrite of the stax api is IP
wise the best move.

salu2

>
> In the end, it is probably easier to just make the switch.  These last  
> minute changes are not fun.
>
> -Grant
>
> On Sep 8, 2008, at 5:27 PM, Yonik Seeley wrote:
>
> > Someone brought up an issue with the StAX licensem, and after digging
> > through legal-discuss archives it seems the easiest way forward is to
> > perhaps replace it with the geronimo stax API and the woodstox
> > implementation.
> >
> > I brought it up on legal-discuss again, but I'm not hopeful of an
> > unequivocal answer.
> > http://markmail.org/search/?q=StAX+(JSR+173)+License
> >
> > Thoughts/Objections?
> >
> > -Yonik
>
>
--
Thorsten Scherler                                 thorsten.at.apache.org
Open Source Java                      consulting, training and solutions

Reply | Threaded
Open this post in threaded view
|

Re: replace stax API with Geronimo-stax+Woodstox

Otis Gospodnetic-2
In reply to this post by Yonik Seeley-2
Did this ("may even be faster") turn out to be true?


Otis


----- Original Message ----

> From: Yonik Seeley <[hidden email]>
> To: [hidden email]
> Sent: Tuesday, September 9, 2008 11:07:29 AM
> Subject: Re: replace stax API with Geronimo-stax+Woodstox
>
> FYI, I'm testing Solr with woodstox now and will probably do some ad
> hoc stress testing too.
> But woodstox is a quality parser.  I expect less problems then we had
> with the reference implementation (and it may even be faster too)
>
> -Yonik