You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@sling.apache.org by Apache Hudson Server <hu...@hudson.zones.apache.org> on 2009/07/02 16:13:24 UTC
Hudson build became unstable: sling-trunk-1.5 » Apache Sling Launchpad Testing #156
See http://hudson.zones.apache.org/hudson/job/sling-trunk-1.5/org.apache.sling$org.apache.sling.launchpad.testing/156/
Re: Hudson build is still unstable: sling-trunk-1.5 » Apache Sling Launchpad Testing #157
Posted by Bertrand Delacretaz <bd...@apache.org>.
On Mon, Jul 6, 2009 at 12:18 PM, Felix Meschberger<fm...@gmail.com> wrote:
> ...My option is to allow sling.properties style overwrite of the default or
> configured value....
That would work for integration testing IMHO.
-Bertrand
Re: Hudson build is still unstable: sling-trunk-1.5 » Apache Sling Launchpad Testing #157
Posted by Felix Meschberger <fm...@gmail.com>.
Hi,
Bertrand Delacretaz schrieb:
> On Mon, Jul 6, 2009 at 11:17 AM, Felix Meschberger<fm...@gmail.com> wrote:
>> This is most probably caused by a recent change in the default value for
>> the JSP "development" mode flag:...
>
>> ...I am currently investigating what to do....
>
> Could we set the JSP "development mode" via an HTTP request? The tests
> could then do that.
That would be conceivable as a short-term workaround. But in fact this
would be kind of a "breach of contract": Whether a request is handled by
a JSP or not is an implementation detail. How the JSP is called is
another implementation detail. "Publicizing" such details by supporting
a "use JSP compiler in development mode" flag is IMHO wrong.
My option is to allow sling.properties style overwrite of the default or
configured value. Not sure, whether this is really the best of all
options. Maybe we even have to go other routes....
Regards
Felix
Re: Hudson build is still unstable: sling-trunk-1.5 » Apache Sling Launchpad Testing #157
Posted by Bertrand Delacretaz <bd...@apache.org>.
On Mon, Jul 6, 2009 at 11:17 AM, Felix Meschberger<fm...@gmail.com> wrote:
> This is most probably caused by a recent change in the default value for
> the JSP "development" mode flag:...
> ...I am currently investigating what to do....
Could we set the JSP "development mode" via an HTTP request? The tests
could then do that.
-Bertrand
Re: Hudson build is still unstable: sling-trunk-1.5 » Apache Sling Launchpad Testing #157
Posted by Carsten Ziegeler <cz...@apache.org>.
Alexander Klimetschek wrote:
> On Mon, Jul 6, 2009 at 5:13 PM, Carsten Ziegeler<cz...@apache.org> wrote:
>> If development mode is turned on, jsp files are checked for
>> recompilation every 4 seconds - this value can be changed through
>> configuration.
>
> So this doesn't happen based on the modification event but by checking
> upon request for the compiled jsp, if 4 seconds are gone after the
> last check? That would explain why I sometimes need a few reloads...
> ;-)
>
Exactly - just work a little bit slower :)
Carsten
--
Carsten Ziegeler
cziegeler@apache.org
Re: Hudson build is still unstable: sling-trunk-1.5 » Apache Sling Launchpad Testing #157
Posted by Alexander Klimetschek <ak...@day.com>.
On Mon, Jul 6, 2009 at 5:13 PM, Carsten Ziegeler<cz...@apache.org> wrote:
> If development mode is turned on, jsp files are checked for
> recompilation every 4 seconds - this value can be changed through
> configuration.
So this doesn't happen based on the modification event but by checking
upon request for the compiled jsp, if 4 seconds are gone after the
last check? That would explain why I sometimes need a few reloads...
;-)
Regards,
Alex
--
Alexander Klimetschek
alexander.klimetschek@day.com
Re: Hudson build is still unstable: sling-trunk-1.5 » Apache Sling Launchpad Testing #157
Posted by Carsten Ziegeler <cz...@apache.org>.
Ok, just as a clarification :) (It took me a while to find this out
myself...)
If development mode is turned off, the jsp is compiled once - and never
checked again; even removing the generated class file or removing the
jsp file itself has no effect! So turning this off in development makes
no sense at all :) I even doubt that this is very usefull for production
in the Sling context.
If development mode is turned on, jsp files are checked for
recompilation every 4 seconds - this value can be changed through
configuration.
So I turned development mode back on and clarified the properties in the
metatype description a little bit :)
As a final node, consider this recompilation check if you're doing
performance tests :) You definitly don't want to hit a recheck phase
during your performance test....
Regards
Carsten
Carsten Ziegeler wrote:
> Felix Meschberger wrote:
>> Hi,
>>
>> This is most probably caused by a recent change in the default value for
>> the JSP "development" mode flag: Previously the default value was
>> "true", that is on each access to a JSP script, the last modification
>> time has been checked. Now, the default value is false and hence the
>> last modification time of the script is only checked every 4 seconds (by
>> default).
>>
>> Now, our integration tests replace JSP scripts but since the replacement
>> and call happens within 4 seconds before the last call, the old contents
>> is still used and hence the tests fail.
>>
>> I am currently investigating what to do.
>>
> It seems there is a bug in the jsp script handler - as soon as
> development mode is off
> it seems to not check anymore at all. At least some quick tests did show
> this behaviour.
>
> Carsten
>
>
--
Carsten Ziegeler
cziegeler@apache.org
Re: Hudson build is still unstable: sling-trunk-1.5 » Apache Sling Launchpad Testing #157
Posted by Carsten Ziegeler <cz...@apache.org>.
Felix Meschberger wrote:
> Hi,
>
> This is most probably caused by a recent change in the default value for
> the JSP "development" mode flag: Previously the default value was
> "true", that is on each access to a JSP script, the last modification
> time has been checked. Now, the default value is false and hence the
> last modification time of the script is only checked every 4 seconds (by
> default).
>
> Now, our integration tests replace JSP scripts but since the replacement
> and call happens within 4 seconds before the last call, the old contents
> is still used and hence the tests fail.
>
> I am currently investigating what to do.
>
It seems there is a bug in the jsp script handler - as soon as
development mode is off
it seems to not check anymore at all. At least some quick tests did show
this behaviour.
Carsten
--
Carsten Ziegeler
cziegeler@apache.org
Re: Hudson build is still unstable: sling-trunk-1.5 » Apache Sling Launchpad Testing #157
Posted by Felix Meschberger <fm...@gmail.com>.
Hi,
This is most probably caused by a recent change in the default value for
the JSP "development" mode flag: Previously the default value was
"true", that is on each access to a JSP script, the last modification
time has been checked. Now, the default value is false and hence the
last modification time of the script is only checked every 4 seconds (by
default).
Now, our integration tests replace JSP scripts but since the replacement
and call happens within 4 seconds before the last call, the old contents
is still used and hence the tests fail.
I am currently investigating what to do.
Regards
Felix
Apache Hudson Server schrieb:
> See http://hudson.zones.apache.org/hudson/job/sling-trunk-1.5/org.apache.sling$org.apache.sling.launchpad.testing/157/
>
>
>
Hudson build is still unstable: sling-trunk-1.5 » Apache Sling Launchpad Testing #157
Posted by Apache Hudson Server <hu...@hudson.zones.apache.org>.
See http://hudson.zones.apache.org/hudson/job/sling-trunk-1.5/org.apache.sling$org.apache.sling.launchpad.testing/157/