yup, looks like the new columns added as part of OGM-717 don't work with
cassandra. Which kinda begs the question - how did the branch get into
that state? Is the CI skipping tests?
Jonathan.
On 15/05/15 16:58, Emmanuel Bernard wrote:
> Hey Jonathan,
>
> I investigated a bit the Cassandra fa
On 18 May 2015 at 10:06, Jonathan Halliday wrote:
>
> yup, looks like the new columns added as part of OGM-717 don't work with
> cassandra. Which kinda begs the question - how did the branch get into
> that state? Is the CI skipping tests?
No, it was reported:
- http://ci.hibernate.org/view/O
Hey,
Yes, Davide has already sent a fix for this, by mapping these types to
String rather than Character. Indeed CI was skipping test execution for
Cassandra (as CASSANDRA_HOSTNAME wasn't set), I have changed this earlier
this morning. Just re-triggered the build for master, let's see how that
goe
Ok, sorry I must have checked the build right after you fixed it ..
Thanks,
Sanne
On 18 May 2015 at 10:12, Gunnar Morling wrote:
> Hey,
>
> Yes, Davide has already sent a fix for this, by mapping these types to
> String rather than Character. Indeed CI was skipping test execution for
> Cassandra
There are three classes AssociationKey, EnityKey, IdSourceKey, which all
contains methods:
public String[] getColumnNames()
public Object[] getColumnValues()
public String getTable()
and we have three methods in couchdb module:
Identifier.createAssociationId, Identifier.createEntityId,
CouchDBData
No. We had that abstract class before and that turned out to be a bad idea when
you factor forever backward compatibility into the mix.
> On 18 mai 2015, at 13:08, Ewa Stawicka wrote:
>
> There are three classes AssociationKey, EnityKey, IdSourceKey, which all
> contains methods:
> public Stri
Note that custom checkstyle rule implementations does not fare well with IDE
checkstyle rules enforcement.
I can't make it reliably work on search and OGM.
> On 16 mai 2015, at 16:11, Sanne Grinovero wrote:
>
> One nice thing of checkstyle is that it's easily extended, for example
> in Search
On 16 May 2015 at 19:43, Steve Ebersole wrote:
>
> Sanne, can you give me the low-down on this line?
>
>
I'm not sure of that one; it's "documented" here under TreeWalker section:
- http://checkstyle.sourceforge.net/config.html
That specific value we use is what is suggested to use in combinat
On 18 May 2015 at 15:23, Emmanuel Bernard wrote:
> Note that custom checkstyle rule implementations does not fare well with IDE
> checkstyle rules enforcement.
> I can't make it reliably work on search and OGM.
As far as I remember the problem was related with having the
customisations as part o
Sure. Personally I think the "trick" is to keep the number of failures
small. In fact I initially listed just 9, but am already thinking of
moving at least one from high to medium (and therefore not considered a
failure).
On Mon, May 18, 2015 at 9:23 AM, Emmanuel Bernard
wrote:
> Note that
As far as the file header, I have often thought of condensing our ORM one
so that is awesome information. For my own curiosity did Richard give you
a reason that the copyright statement is not necessary?
On Mon, May 18, 2015 at 9:58 AM, Sanne Grinovero
wrote:
> On 18 May 2015 at 15:23, Emmanuel
Sorry hit send too soon :(
Anyway.. I assume it is because we know this information from commit
history?
As far as the ORM one, I propose a similar change. Any objections?
On Mon, May 18, 2015 at 10:14 AM, Steve Ebersole
wrote:
> As far as the file header, I have often thought of condensing o
+1 I thought it was in already :)
> On 18 mai 2015, at 17:20, Steve Ebersole wrote:
>
> Sorry hit send too soon :(
>
> Anyway.. I assume it is because we know this information from commit
> history?
>
> As far as the ORM one, I propose a similar change. Any objections?
>
> On Mon, May 18, 20
And see, this[1] is exactly the kind of thing I always try to avoid... Why
did y'all decide on that approach?
[1] https://github.com/hibernate/hibernate-search/blob/master/copyright.txt
On Mon, May 18, 2015 at 10:33 AM, Emmanuel Bernard
wrote:
> +1 I thought it was in already :)
>
> > On 18 m
On 18 May 2015 at 19:44, Steve Ebersole wrote:
> And see, this[1] is exactly the kind of thing I always try to avoid... Why
> did y'all decide on that approach?
Personally I think the main reason is pride. Some contributors
explicitly asked me if it was ok to add their names, after they had
cont
Hardy, do you happen to still have your sed script for updating these file
headers?
On Mon, May 18, 2015 at 2:48 PM, Sanne Grinovero
wrote:
> On 18 May 2015 at 19:44, Steve Ebersole wrote:
> > And see, this[1] is exactly the kind of thing I always try to avoid...
> Why
> > did y'all decide on t
The recent discussion on Checkstyle got me thinking about best ways to
share/distribute code styles for IDE setup. Storing them in the project
itself is not very workable for IntelliJ at least. I started thinking
about a separate repository under the hibernate GitHub organization. Does
everyone
ORM master is now enforcing checkstyle. The CI builds will fail if you
commit anything that was deemed a severity of error/high. The only part
missing there is headers, which I will follow up with once I hear back from
Hardy.
On Mon, May 18, 2015 at 3:07 PM, Steve Ebersole wrote:
> Hardy, do y
18 matches
Mail list logo
