On 3/25/2014 4:19 AM, Peter Klügl wrote:
> I would personally argue for a new RC, but I can also live if this one
> is released.
>
> Marshall, what do you think?
I'm -0 on the release with it not building from source on linux or mac OS (due
only to back-slash in some paths, that's my understanding - please advise if not
correct).

If this were less high-profile (not a top level link from the top page, etc.), I
would be +0 for release :-) .

Also, since the release manager wants a new RC, I don't think anyone will object
(although we'll have to re-do the release checking work...)

-Marshall
>
> Richard, can you test the unit tests in the current trunk, if the
> problem still occurs on OS X?
>
> Peter
>
>
> Am 24.03.2014 14:08, schrieb Peter Klügl:
>> Hi,
>>
>> Am 23.03.2014 12:32, schrieb Richard Eckart de Castilho:
>>> - Checking staging repository
>>> -- org/apache/uima/ruta-core/2.2.0/ruta-core-2.2.0.pom contains some 
>>> backslashes instead of slashes in paths
>> I will fix that.
>>
>>> -- how about excluding the docbook module from the Maven deploy - it does 
>>> not contain any artifacts other than the POM
>> Good idea in my opinion. I suppose the indented way is to remove the
>> module in the top-level pom?
>>
>>
>>> -- I believe we had previously discussed changing the release procedure to 
>>> always use a tag that does not contain "-rcX", e.g. the file 
>>> org/apache/uima/ruta-ep-engine/2.2.0/ruta-ep-engine-2.2.0.pom contains the 
>>> reference to 
>>> http://svn.apache.org/viewvc/uima/ruta/tags/ruta-2.2.0-rc3/ruta-ep-engine 
>>> which will be invalid if we rename the tag after the release. However, the 
>>> release instructions on the UIMA website still mention adding the "-rcX" 
>>> and I do not find the old mails. Did we overlook updating the instructions 
>>> or do I remember the outcome of our conversation incorrectly?
>> I also remember now that you mentioned this issue, but I cannot recall
>> the result of the discussion.
>>
>>> - Source release ZIP
>>> -- build with clean repo on Java 1.7.0_21 on OS X: test failure in 
>>> org.apache.uima.ruta.engine.UimafitTest
>>>    looks like a problem with line endings: 
>>>    --expected--
>>>    original: <null> -------- View _InitialView end 
>>> ---------------------------------- ^M]
>>>    --actual--
>>>    original: <null>
>>> -------- View _InitialView end ----------------------------------
>>> ]
>> I think I already debugged the problem sometime and modified the
>> casdump.txt. Actually, I added svn:eol-style native.
>> I am tempted to simply change the test by removing the external file.
>>
>>> - SVN tag
>>> -- build with clean repo on Java 1.7.0_21 on OS X: OK
>>>
>>> - Eclipse update site
>>> -- Update existing Eclipse with 
>>> http://people.apache.org/~pkluegl/uima-releases/ruta-2.2.0-rc3/eclipse-update-site/ruta
>>> -- Appears to install ok
>>>
>>> - Checked several artifacts for license/notice files and general issues
>>> -- Looks ok
>>>
>>> Overall: nothing really critical, but some issues that could be addressed 
>>> for the next release.
>>>
>>> The worst problem seems to be that building from the source release does 
>>> not work on OS X (probably Linux).  What are others opinion on this? I 
>>> believe you all checked with Windows, so you didn't hit this problem. 
>>> Should we consider this release critical?
>> It is at least not trivial.
>>
>> Marshall?
>>
>>
>>
>> Best,
>>
>> Peter
>>
>>
>>> -- Richard
>>>
>>> On 12.03.2014, at 13:08, Peter Klügl <[email protected]> wrote:
>>>
>>>> Hi,
>>>>
>>>> the third release candidate of Apache UIMA Ruta v2.2.0 is ready for voting.
>>>>
>>>> Changes in RC3 since RC2:
>>>> (UIMA-3627) Replace old screenshot in Ruta documentation
>>>> (UIMA-3569) Allow extensions for complete block constructs in Ruta
>>>> (UIMA-3628) Loading scripts/descriptors in Ruta with incorrect paths
>>>> (UIMA-3622) Formatter in Ruta editor duplicates declare keywords
>>>> (UIMA-3621) Improve license/notice files in Ruta
>>>> (UIMA-3585) Ruta: Multi-Tree Wordlists with Multi-Token Entries
>>>> (UIMA-3229) Investigate rule performance dependent of previous rules
>>>> (UIMA-3651) Imports by name not resolved by ResourceManager in Ruta
>>>> Workbench
>>>> (UIMA-3377) Introduce inference exceptions in Ruta
>>>> (UIMA-3656) Nested disjunctive composed rule elements cause NPE
>>>> (UIMA-3666) Ruta: Allow an additional argument in INLIST condition for
>>>> the check
>>>> (UIMA-3469) (reopened) Ruta: Annotation Browser View Extensions
>>>> (UIMA-3670) Ruta action SETFEATURE does not work with feature expressions
>>>> (UIMA-3669) Problem with Annotation Declaration, Error parsing XCAS or
>>>> XMI-CAS ...
>>>>
>>>> Staging repository:
>>>> https://repository.apache.org/content/repositories/orgapacheuima-1009/
>>>>
>>>> SVN tag:
>>>> https://svn.apache.org/repos/asf/uima/ruta/tags/ruta-2.2.0-rc3
>>>>
>>>> Update site:
>>>> http://people.apache.org/~pkluegl/uima-releases/ruta-2.2.0-rc3/eclipse-update-site/ruta
>>>>
>>>> Archive with all sources:
>>>> http://people.apache.org/~pkluegl/uima-releases/ruta-2.2.0-rc3/ruta-2.2.0-source-release.zip
>>>>
>>>> Overall 67 issues have been fixed for this release.
>>>> They can be found in the RELEASE_NOTES.html and here:
>>>> https://issues.apache.org/jira/issues/?jql=project%20%3D%20UIMA%20AND%20fixVersion%20%3D%202.2.0ruta%20AND%20component%20%3D%20ruta%20ORDER%20BY%20priority%20DESC
>>>>
>>>> Major Changes in this Release
>>>>
>>>> UIMA Ruta Language and Analysis Engine:
>>>> - Major performance improvements
>>>> - Improved import type functionality and handling of ambiguous short names
>>>> - Options to determine where the next match should start
>>>> - Requires at least Java 6
>>>> - Many bug fixes
>>>>
>>>> UIMA Ruta Workbench:
>>>> - Smaller improvements in many views
>>>> - Support of mixin Java/Ruta projects
>>>> - Many bug fixes
>>>>
>>>>
>>>> ONLY FOR REVIEWING:
>>>>
>>>> Documentation (pdf file):
>>>> http://people.apache.org/~pkluegl/uima-releases/ruta-2.2.0-rc3/tools.ruta.book.pdf
>>>>
>>>>
>>>> Please vote on release:
>>>>
>>>> [ ] +1 OK to release
>>>> [ ]  0 Don't care
>>>> [ ] -1 Not OK to release, because ...
>>>>
>>>> Thanks.
>>>>
>>>> Peter
>
>

Reply via email to