RE: [RESULT][VOTE] Re: XMLBeans 3.0.0 release vote

2018-07-03 Thread Murphy, Mark
It was my understanding that we now own XMLBeans, and no board discussion is 
necessary to release under http://www.apache.org/dist/xmlbeans/release

I think it was just some of the attic PMC that was bringing up the issue, not 
the board. In fact, I believe the board's response was JFDI.

So we own the distribution channel, the web site, Jira, and everything else 
associated with XMLBeans. So we should just keep things simple and continue 
distributing, and updating as if we were the XMLBeans PMC (which we are).

-Original Message-
From: Andreas Beeker [mailto:kiwiwi...@apache.org] 
Sent: Monday, July 02, 2018 7:07 PM
To: dev@poi.apache.org
Subject: Re: [RESULT][VOTE] Re: XMLBeans 3.0.0 release vote

> Would it make sense to change the links ... ?

Yes, that makes sense - please change it.

I had to install Forrest 0.8 and needed to start the site build via 
//bin/forrest site plus adding the permissions to download 
forrest plugins.
I.e. committing xmlbeans/build/site will publish to the website directly

I didn't want to push the announcement before the website and release directory 
is on a current state.

There are still some (2.4.0...) leftovers on the XMLBeans website which I need 
to deal with tomorrow ...
but feel free to make any changes earlier - it would be good if all (POI) PMCs 
have a look at the old/new/modified XMLBeans website, so we don't relaunch with 
too many old links

I've put the release artifacts now under 
http://www.apache.org/dist/poi/xmlbeans/release/
as I currently don't want to go the full way of discussing further XMLBeans 
community steps (including a separate release directory under the dist root) 
with the board.

Andi




Re: [RESULT][VOTE] Re: XMLBeans 3.0.0 release vote

2018-07-02 Thread Andreas Beeker
> Would it make sense to change the links ... ?

Yes, that makes sense - please change it.

I had to install Forrest 0.8 and needed to start the site build via 
//bin/forrest site
plus adding the permissions to download forrest plugins.
I.e. committing xmlbeans/build/site will publish to the website directly

I didn't want to push the announcement before the website and release directory 
is on a current state.

There are still some (2.4.0...) leftovers on the XMLBeans website which I need 
to deal with tomorrow ...
but feel free to make any changes earlier - it would be good if all (POI) PMCs 
have a look at the
old/new/modified XMLBeans website, so we don't relaunch with too many old links

I've put the release artifacts now under 
http://www.apache.org/dist/poi/xmlbeans/release/
as I currently don't want to go the full way of discussing further XMLBeans 
community steps
(including a separate release directory under the dist root) with the board.

Andi




signature.asc
Description: OpenPGP digital signature


Re: [RESULT][VOTE] Re: XMLBeans 3.0.0 release vote

2018-07-02 Thread pj.fanning
Thanks Andi for updating the xmlbeans.apache.org site.

Would it make sense to change the links on
https://xmlbeans.apache.org/community/index.html to point to the POI mailing
lists - the XMLBeans lists haven't been reactivated?



--
Sent from: http://apache-poi.1045710.n5.nabble.com/POI-Dev-f2312866.html

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org



[RESULT][VOTE] Re: XMLBeans 3.0.0 release vote

2018-07-02 Thread Andreas Beeker
The vote has passed with 3x +1 from POI / XmlBeans PMCs.

This thread can be used as a reference for the already released artifact ...
I'll now write the public announcement mail.

Andi.


On 7/1/18 9:09 PM, Dominik Stadler wrote:
> Uups, forgot to vote, I am +1 for XmlBeans 3.0.0 as no big problem related
> to XmlBeans showed up in the regression runs for Apache POI
>
> Dominik
>
> On Sat, Jun 30, 2018 at 1:48 PM, Andreas Beeker 
> wrote:
>
>> Hi PJ,
>>
>> before I update the website there is something missing:
>>
>> - a vote thread with 3x +1 from PMCs (so you are missing another one)
>>
>> Parallel to the website announcement an announcement mail should be also
>> sent.
>> just check the archives for the POI announcement.
>>
>> Can I support you on this?
>>
>> Andi
>>
>>
>>




signature.asc
Description: OpenPGP digital signature


Re: XMLBeans 3.0.0 release vote

2018-07-01 Thread Dominik Stadler
Uups, forgot to vote, I am +1 for XmlBeans 3.0.0 as no big problem related
to XmlBeans showed up in the regression runs for Apache POI

Dominik

On Sat, Jun 30, 2018 at 1:48 PM, Andreas Beeker 
wrote:

> Hi PJ,
>
> before I update the website there is something missing:
>
> - a vote thread with 3x +1 from PMCs (so you are missing another one)
>
> Parallel to the website announcement an announcement mail should be also
> sent.
> just check the archives for the POI announcement.
>
> Can I support you on this?
>
> Andi
>
>
>


Re: XMLBeans 3.0.0 release vote

2018-06-30 Thread Andreas Beeker
Hi PJ,

before I update the website there is something missing:

- a vote thread with 3x +1 from PMCs (so you are missing another one)

Parallel to the website announcement an announcement mail should be also sent.
just check the archives for the POI announcement.

Can I support you on this?

Andi




signature.asc
Description: OpenPGP digital signature


Re: XMLBeans 3.0.0 release vote

2018-06-27 Thread Dominik Stadler
Here are the results of a regression run on 1258881 documents.

Overall it looks fairly good, I see no big problems with either the
replaced zip-classes or XmlBeans 3.0.

There are 1332 suspicious new errors with the following breakdown:

* 664 times java.lang.NoClassDefFoundError:
org/openxmlformats/schemas/presentationml/x2006/main/CTHeaderFooter, which
indicates a missing unit test for this CT-class and should be easy to fix
(actually already in the works in my local repo)
* 192 strange NullPointerExceptions in my code which I cannot fully explain
* a number of "Element styles@
http://schemas.openxmlformats.org/wordprocessingml/2006/main is not a valid
styleSheet@http://schemas.openxmlformats.org/spreadsheetml/2006/main
document or a valid substitution.", which I think we saw before as well, so
probably just a flakiness where we had TIMEOUT or OOM before
* 26 java.lang.ArrayIndexOutOfBoundsException in XSLF handling somewhere,
see report for details
* and finally a number of new ZipBombExceptions, which sounds like the
limit for reporting was lowered by the zip-handling-changes?

Full list of errors is at
http://people.apache.org/~centic/poi_regression/reportsAll/index317to400SNAPSHOT.html
Comparison of 3.17 vs. 4.0.0-SNAPSHOT is at
http://people.apache.org/~centic/poi_regression/reports/index317to400SNAPSHOT.html

Thanks... Dominik.

On Tue, Jun 26, 2018 at 9:30 PM, Dominik Stadler 
wrote:

> Hi,
>
> Unfortunately I have to restart the run, commons-compress was missing
> which caused 70k failures out of 1.2mio documents. Thus XmlBeans was not
> properly tested and I need to re-trigger.
>
>
> However I already noticed the following failures that look new (however
> these are all very rare, not related to XmlBeans at all, mostly in
> HSSF/HSLF!):
>
> 4
> ERROR
> java.lang.ClassCastException: o.a.p.hssf.record.StyleRecord cannot be cast
> to o.a.p.hssf.record.ExtendedFormatRecord
>
> java.lang.ClassCastException: o.a.p.hssf.record.StyleRecord cannot be cast to 
> o.a.p.hssf.record.ExtendedFormatRecord
>   at 
> o.a.p.hssf.model.InternalWorkbook.getExFormatAt(InternalWorkbook.java:870)
>   at o.a.p.hssf.usermodel.HSSFCell.getCellStyle(HSSFCell.java:943)
>   at o.a.p.hssf.usermodel.HSSFCell.getCellStyle(HSSFCell.java:71)
>   at o.a.p.stress.HSSFFileHandler.handleFile(HSSFFileHandler.java:63)
>   at 
> org.dstadler.commoncrawl.FileHandlingRunnable.run(FileHandlingRunnable.java:64)
>   at 
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
>   at 
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
>   at java.lang.Thread.run(Thread.java:748)
>
> Download
> 
>
>
>
> 1
> ERROR
> java.lang.NullPointerException
>
> java.lang.NullPointerException
>   at 
> o.a.p.hslf.model.textproperties.HSLFTabStopPropCollection.writeTabStops(HSLFTabStopPropCollection.java:83)
>   at o.a.p.hslf.record.TextRulerAtom.writeIf(TextRulerAtom.java:150)
>   at o.a.p.hslf.record.TextRulerAtom.writeOut(TextRulerAtom.java:126)
>   at 
> o.a.p.hslf.record.EscherTextboxWrapper.writeOut(EscherTextboxWrapper.java:90)
>   at 
> o.a.p.hslf.usermodel.HSLFTextParagraph.refreshRecords(HSLFTextParagraph.java:1158)
>   at 
> o.a.p.hslf.usermodel.HSLFTextParagraph.storeText(HSLFTextParagraph.java:969)
>   at o.a.p.hslf.usermodel.HSLFTextShape.storeText(HSLFTextShape.java:851)
>   at 
> o.a.p.hslf.usermodel.HSLFSlideShow.writeDirtyParagraphs(HSLFSlideShow.java:485)
>   at 
> o.a.p.hslf.usermodel.HSLFSlideShow.writeDirtyParagraphs(HSLFSlideShow.java:477)
>   at o.a.p.hslf.usermodel.HSLFSlideShow.write(HSLFSlideShow.java:451)
>   at o.a.p.stress.SlideShowHandler.writeToArray(SlideShowHandler.java:63)
>   at 
> o.a.p.stress.SlideShowHandler.handleSlideShow(SlideShowHandler.java:49)
>   at o.a.p.stress.HSLFFileHandler.handleFile(HSLFFileHandler.java:49)
>   at 
> org.dstadler.commoncrawl.FileHandlingRunnable.run(FileHandlingRunnable.java:64)
>   at 
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
>   at 
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
>   at java.lang.Thread.run(Thread.java:748)
>
> Download
> 
>
>
>
> 1
> ERROR
> java.lang.ArrayIndexOutOfBoundsException: *
>
> java.lang.ArrayIndexOutOfBoundsException: *
>   at 
> o.a.p.hssf.usermodel.HSSFOptimiser.optimiseCellStyles(HSSFOptimiser.java:229)
>   at o.a.p.stress.HSSFFileHandler.handleFile(HSSFFileHandler.java:59)
>   at 
> org.dstadler.commoncrawl.FileHandlingRunnable.run(FileHandlingRunnable.java:64)
>   at 
> 

Re: XMLBeans 3.0.0 release vote

2018-06-26 Thread Dominik Stadler
Hi,

Unfortunately I have to restart the run, commons-compress was missing which
caused 70k failures out of 1.2mio documents. Thus XmlBeans was not properly
tested and I need to re-trigger.


However I already noticed the following failures that look new (however
these are all very rare, not related to XmlBeans at all, mostly in
HSSF/HSLF!):

4
ERROR
java.lang.ClassCastException: o.a.p.hssf.record.StyleRecord cannot be cast
to o.a.p.hssf.record.ExtendedFormatRecord

java.lang.ClassCastException: o.a.p.hssf.record.StyleRecord cannot be
cast to o.a.p.hssf.record.ExtendedFormatRecord
at 
o.a.p.hssf.model.InternalWorkbook.getExFormatAt(InternalWorkbook.java:870)
at o.a.p.hssf.usermodel.HSSFCell.getCellStyle(HSSFCell.java:943)
at o.a.p.hssf.usermodel.HSSFCell.getCellStyle(HSSFCell.java:71)
at o.a.p.stress.HSSFFileHandler.handleFile(HSSFFileHandler.java:63)
at 
org.dstadler.commoncrawl.FileHandlingRunnable.run(FileHandlingRunnable.java:64)
at 
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at 
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)

Download




1
ERROR
java.lang.NullPointerException

java.lang.NullPointerException
at 
o.a.p.hslf.model.textproperties.HSLFTabStopPropCollection.writeTabStops(HSLFTabStopPropCollection.java:83)
at o.a.p.hslf.record.TextRulerAtom.writeIf(TextRulerAtom.java:150)
at o.a.p.hslf.record.TextRulerAtom.writeOut(TextRulerAtom.java:126)
at 
o.a.p.hslf.record.EscherTextboxWrapper.writeOut(EscherTextboxWrapper.java:90)
at 
o.a.p.hslf.usermodel.HSLFTextParagraph.refreshRecords(HSLFTextParagraph.java:1158)
at 
o.a.p.hslf.usermodel.HSLFTextParagraph.storeText(HSLFTextParagraph.java:969)
at o.a.p.hslf.usermodel.HSLFTextShape.storeText(HSLFTextShape.java:851)
at 
o.a.p.hslf.usermodel.HSLFSlideShow.writeDirtyParagraphs(HSLFSlideShow.java:485)
at 
o.a.p.hslf.usermodel.HSLFSlideShow.writeDirtyParagraphs(HSLFSlideShow.java:477)
at o.a.p.hslf.usermodel.HSLFSlideShow.write(HSLFSlideShow.java:451)
at o.a.p.stress.SlideShowHandler.writeToArray(SlideShowHandler.java:63)
at 
o.a.p.stress.SlideShowHandler.handleSlideShow(SlideShowHandler.java:49)
at o.a.p.stress.HSLFFileHandler.handleFile(HSLFFileHandler.java:49)
at 
org.dstadler.commoncrawl.FileHandlingRunnable.run(FileHandlingRunnable.java:64)
at 
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at 
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)

Download




1
ERROR
java.lang.ArrayIndexOutOfBoundsException: *

java.lang.ArrayIndexOutOfBoundsException: *
at 
o.a.p.hssf.usermodel.HSSFOptimiser.optimiseCellStyles(HSSFOptimiser.java:229)
at o.a.p.stress.HSSFFileHandler.handleFile(HSSFFileHandler.java:59)
at 
org.dstadler.commoncrawl.FileHandlingRunnable.run(FileHandlingRunnable.java:64)
at 
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at 
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)

Download


Dominik.

On Mon, Jun 25, 2018 at 7:30 AM, Dominik Stadler 
wrote:

> A regression run on the > 1mio commoncrawl documents is underway...
>
> Dominik
>
> On Sun, Jun 24, 2018, 22:50 Andreas Beeker  wrote:
>
>> I'm +1 too ... we can refactor the code in the next release.
>>
>> A govdocs run would be nice, to see if we have further OOM problems.
>>
>> Andi
>>
>>
>>


Re: XMLBeans 3.0.0 release vote

2018-06-24 Thread Dominik Stadler
A regression run on the > 1mio commoncrawl documents is underway...

Dominik

On Sun, Jun 24, 2018, 22:50 Andreas Beeker  wrote:

> I'm +1 too ... we can refactor the code in the next release.
>
> A govdocs run would be nice, to see if we have further OOM problems.
>
> Andi
>
>
>


Re: XMLBeans 3.0.0 release vote

2018-06-24 Thread Andreas Beeker
I'm +1 too ... we can refactor the code in the next release.

A govdocs run would be nice, to see if we have further OOM problems.

Andi




signature.asc
Description: OpenPGP digital signature


Re: XMLBeans 3.0.0 release vote

2018-06-22 Thread pj.fanning
Some of the github pull request builds failed and I wasn't sure what was
causing the issue. Conceptually, the POI build no longer needs any piccolo
manipulation because xmlbeans 3.0.0 removes the support for piccolo
entirely. 



--
Sent from: http://apache-poi.1045710.n5.nabble.com/POI-Dev-f2312866.html

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org



Re: XMLBeans 3.0.0 release vote

2018-06-22 Thread Andreas Beeker
Sorry for the confusion of pull request vs. github mirror - ... now it's on the 
SVN.

What was the problem with the piccolo removal? ... after removing the "*.orig" 
renamings it worked out of the box?

As we (POI) provide our own ContextTypeLoader (POIXMLTypeLoader), I'm just 
trying to find out, if the TypeLoader loaded via [1] is safely loading the xml 
sources.

Andi

[1] org.apache.xmlbeans.XmlBeans.getContextTypeLoader()



signature.asc
Description: OpenPGP digital signature


Re: XMLBeans 3.0.0 release vote

2018-06-22 Thread pj.fanning
github is a read only mirror of the poi svn - we use pull requests on github
to demonstrate proposed changes but those pull requests need to be merged
using svn commits (by POI committers)



--
Sent from: http://apache-poi.1045710.n5.nabble.com/POI-Dev-f2312866.html

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org



Re: XMLBeans 3.0.0 release vote

2018-06-22 Thread sebb
On 22 June 2018 at 21:49, Andreas Beeker  wrote:
> I've just realized, that you've done the changes to the github mirror.
> In the context of POI, I'm looking at github only as a source for patches, 
> but not as our project repo.
> So I try to integrate your changes locally and then to the svn repo ...

Huh?
If GitHub is set up as a proper mirror, it should not differ from the
project repo.

There should be only one master repo which holds all changes.
If that is not the case, it seems to me that something has gone badly wrong.

> Andi
>
> On 6/22/18 10:35 PM, Andreas Beeker wrote:
>> Before I vote, I'd like to see the XmlBeans 3.0.0 RC version in our POI build
>> and although you've removed and then reverted it, it would be nice, if
>> we don't need to do the piccolo modifications.
>>
>> So it looks like I need to do those modifications locally and find out why
>> the piccolo stuff makes problems ...
>>
>> Andi
>>
>>
>
>

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org



Re: XMLBeans 3.0.0 release vote

2018-06-22 Thread Andreas Beeker
I've just realized, that you've done the changes to the github mirror.
In the context of POI, I'm looking at github only as a source for patches, but 
not as our project repo.
So I try to integrate your changes locally and then to the svn repo ...

Andi

On 6/22/18 10:35 PM, Andreas Beeker wrote:
> Before I vote, I'd like to see the XmlBeans 3.0.0 RC version in our POI build
> and although you've removed and then reverted it, it would be nice, if
> we don't need to do the piccolo modifications.
>
> So it looks like I need to do those modifications locally and find out why
> the piccolo stuff makes problems ...
>
> Andi
>
>




signature.asc
Description: OpenPGP digital signature


Re: XMLBeans 3.0.0 release vote

2018-06-22 Thread Andreas Beeker
Before I vote, I'd like to see the XmlBeans 3.0.0 RC version in our POI build
and although you've removed and then reverted it, it would be nice, if
we don't need to do the piccolo modifications.

So it looks like I need to do those modifications locally and find out why
the piccolo stuff makes problems ...

Andi




signature.asc
Description: OpenPGP digital signature