11:15 a.m.
Hi,
Not a biggie, but I think it would be nice to use this nicer syntax for exception
messages, e.g.
throw new ConfigurationException("you 're not allowed to use %s and %s
together", that, other);
wdyt?
Cheers,
Mircea
12:13 p.m.
Yeah, I currently do:
throw new ConfigurationException(String.format("you 're not allowed to use %s
and %s together", that, other));
but we could overload the constructor of CacheException ...
On 15 Mar 2011, at 16:15, Mircea Markus wrote:
Hi,
Not a biggie, but I think it would be nice to use this nicer syntax for exception
messages, e.g.
throw new ConfigurationException("you 're not allowed to use %s and %s
together", that, other);
wdyt?
Cheers,
Mircea
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Manik Surtani
manik(a)jboss.org
twitter.com/maniksurtani
Lead, Infinispan
http://www.infinispan.org
12:17 p.m.
As David pointed out %1$ is superior to %s for internationalization.
On 15 mars 2011, at 18:13, Manik Surtani wrote:
Yeah, I currently do:
throw new ConfigurationException(String.format("you 're not allowed to use %s
and %s together", that, other));
but we could overload the constructor of CacheException ...
On 15 Mar 2011, at 16:15, Mircea Markus wrote:
> Hi,
>
> Not a biggie, but I think it would be nice to use this nicer syntax for exception
messages, e.g.
> throw new ConfigurationException("you 're not allowed to use %s and %s
together", that, other);
> wdyt?
>
> Cheers,
> Mircea
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Manik Surtani
manik(a)jboss.org
twitter.com/maniksurtani
Lead, Infinispan
http://www.infinispan.org
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
12:21 p.m.
On 15 Mar 2011, at 17:17, Emmanuel Bernard wrote:
As David pointed out %1$ is superior to %s for internationalization.
Agreed.
Not sure we want i18n for logs and error messages (?)
On 15 mars 2011, at 18:13, Manik Surtani wrote:
> Yeah, I currently do:
>
> throw new ConfigurationException(String.format("you 're not allowed to use
%s and %s together", that, other));
>
> but we could overload the constructor of CacheException ...
>
> On 15 Mar 2011, at 16:15, Mircea Markus wrote:
>
>> Hi,
>>
>> Not a biggie, but I think it would be nice to use this nicer syntax for exception
messages, e.g.
>> throw new ConfigurationException("you 're not allowed to use %s and %s
together", that, other);
>> wdyt?
>>
>> Cheers,
>> Mircea
>>
>> _______________________________________________
>> infinispan-dev mailing list
>> infinispan-dev(a)lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
> --
> Manik Surtani
> manik(a)jboss.org
> twitter.com/maniksurtani
>
> Lead, Infinispan
> http://www.infinispan.org
>
>
>
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
12:29 p.m.
On 15 Mar 2011, at 17:21, Mircea Markus wrote:
On 15 Mar 2011, at 17:17, Emmanuel Bernard wrote:
> As David pointed out %1$ is superior to %s for internationalization.
Agreed.
Not sure we want i18n for logs and error messages (?)
It's on our roadmap and a requirement for a few downstream projects/products.
--
Manik Surtani
manik(a)jboss.org
twitter.com/maniksurtani
Lead, Infinispan
http://www.infinispan.org
3:09 a.m.
On Mar 15, 2011, at 6:29 PM, Manik Surtani wrote:
On 15 Mar 2011, at 17:21, Mircea Markus wrote:
>
> On 15 Mar 2011, at 17:17, Emmanuel Bernard wrote:
>
>> As David pointed out %1$ is superior to %s for internationalization.
> Agreed.
> Not sure we want i18n for logs and error messages (?)
It's on our roadmap and a requirement for a few downstream projects/products.
Yup: https://issues.jboss.org/browse/ISPN-380
--
Manik Surtani
manik(a)jboss.org
twitter.com/maniksurtani
Lead, Infinispan
http://www.infinispan.org
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Galder Zamarreño
Sr. Software Engineer
Infinispan, JBoss Cache
3:42 a.m.
The only problem I have with this is that if all error / logging messages
are i18n'ed, they will be difficult to search for on the web / forums /
wikis.
For this reason all error messages should include a locale-independent error
code. Numeric is ok, but I'd rather have a "readable" error code.
For example the string "Error pushing tx log" in InvertedLeaveTask.java
would become "TX_LOG_PUSH_ERR: Error pushing tx log".
Any thoughts on this ?
Tristan
On Wed, Mar 16, 2011 at 09:09, Galder Zamarreño <galder(a)redhat.com> wrote:
On Mar 15, 2011, at 6:29 PM, Manik Surtani wrote:
>
> On 15 Mar 2011, at 17:21, Mircea Markus wrote:
>
>>
>> On 15 Mar 2011, at 17:17, Emmanuel Bernard wrote:
>>
>>> As David pointed out %1$ is superior to %s for internationalization.
>> Agreed.
>> Not sure we want i18n for logs and error messages (?)
>
> It's on our roadmap and a requirement for a few downstream
projects/products.
Yup: https://issues.jboss.org/browse/ISPN-380
>
> --
> Manik Surtani
> manik(a)jboss.org
> twitter.com/maniksurtani
>
> Lead, Infinispan
> http://www.infinispan.org
>
>
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Galder Zamarreño
Sr. Software Engineer
Infinispan, JBoss Cache
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
4:08 a.m.
The road taken by JBoss Logging is to:
- associate an alphanumerical code that you can search on the internet. And of course by
default messages are in English.
- use methods to generate the log call
in Infinispan code
if ( txLog == null ) {
LOG.txLogPushError(txId);
}
@MessageLogger( projectCode = "ISPN" )
public interfaceInfinispanLogger extends BasicLogger {
@LogMessage( level = ERROR )
@Message( value = "Error pushing tx log (tx id=%1$)", id = 345 )
void txLogPushError(String txId);
in your app log
ISPN000345 Error pushing tx log (tx id=3232)
later on, on a wiki page
ISPN-000345 Error pushing tx log
When this happens, you're screwed unless you use JBoss Transaction and then...
I could not really find the doc for JBoss Logging but David will know where it is.
On 16 mars 2011, at 09:42, Tristan Tarrant wrote:
The only problem I have with this is that if all error / logging
messages are i18n'ed, they will be difficult to search for on the web / forums /
wikis.
For this reason all error messages should include a locale-independent error code.
Numeric is ok, but I'd rather have a "readable" error code.
For example the string "Error pushing tx log" in InvertedLeaveTask.java would
become "TX_LOG_PUSH_ERR: Error pushing tx log".
Any thoughts on this ?
Tristan
On Wed, Mar 16, 2011 at 09:09, Galder Zamarreño <galder(a)redhat.com> wrote:
On Mar 15, 2011, at 6:29 PM, Manik Surtani wrote:
>
> On 15 Mar 2011, at 17:21, Mircea Markus wrote:
>
>>
>> On 15 Mar 2011, at 17:17, Emmanuel Bernard wrote:
>>
>>> As David pointed out %1$ is superior to %s for internationalization.
>> Agreed.
>> Not sure we want i18n for logs and error messages (?)
>
> It's on our roadmap and a requirement for a few downstream projects/products.
Yup: https://issues.jboss.org/browse/ISPN-380
>
> --
> Manik Surtani
> manik(a)jboss.org
> twitter.com/maniksurtani
>
> Lead, Infinispan
> http://www.infinispan.org
>
>
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Galder Zamarreño
Sr. Software Engineer
Infinispan, JBoss Cache
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
5:39 a.m.
Yup. Essentially if anyone's ever worked with Oracle DB, plugging an ORA-12345 error
code into Google usually gives you details of the error regardless of the i18n error
message you find in the logs.
On 16 Mar 2011, at 09:08, Emmanuel Bernard wrote:
The road taken by JBoss Logging is to:
- associate an alphanumerical code that you can search on the internet. And of course by
default messages are in English.
- use methods to generate the log call
in Infinispan code
if ( txLog == null ) {
LOG.txLogPushError(txId);
}
@MessageLogger( projectCode = "ISPN" )
public interfaceInfinispanLogger extends BasicLogger {
@LogMessage( level = ERROR )
@Message( value = "Error pushing tx log (tx id=%1$)", id = 345 )
void txLogPushError(String txId);
in your app log
ISPN000345 Error pushing tx log (tx id=3232)
later on, on a wiki page
ISPN-000345 Error pushing tx log
When this happens, you're screwed unless you use JBoss Transaction and then...
I could not really find the doc for JBoss Logging but David will know where it is.
On 16 mars 2011, at 09:42, Tristan Tarrant wrote:
> The only problem I have with this is that if all error / logging messages are
i18n'ed, they will be difficult to search for on the web / forums / wikis.
> For this reason all error messages should include a locale-independent error code.
Numeric is ok, but I'd rather have a "readable" error code.
> For example the string "Error pushing tx log" in InvertedLeaveTask.java
would become "TX_LOG_PUSH_ERR: Error pushing tx log".
>
> Any thoughts on this ?
>
> Tristan
>
> On Wed, Mar 16, 2011 at 09:09, Galder Zamarreño <galder(a)redhat.com> wrote:
>
> On Mar 15, 2011, at 6:29 PM, Manik Surtani wrote:
>
> >
> > On 15 Mar 2011, at 17:21, Mircea Markus wrote:
> >
> >>
> >> On 15 Mar 2011, at 17:17, Emmanuel Bernard wrote:
> >>
> >>> As David pointed out %1$ is superior to %s for internationalization.
> >> Agreed.
> >> Not sure we want i18n for logs and error messages (?)
> >
> > It's on our roadmap and a requirement for a few downstream
projects/products.
>
> Yup: https://issues.jboss.org/browse/ISPN-380
>
> >
> > --
> > Manik Surtani
> > manik(a)jboss.org
> > twitter.com/maniksurtani
> >
> > Lead, Infinispan
> > http://www.infinispan.org
> >
> >
> >
> > _______________________________________________
> > infinispan-dev mailing list
> > infinispan-dev(a)lists.jboss.org
> > https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
> --
> Galder Zamarreño
> Sr. Software Engineer
> Infinispan, JBoss Cache
>
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Manik Surtani
manik(a)jboss.org
twitter.com/maniksurtani
Lead, Infinispan
http://www.infinispan.org
9:18 a.m.
Not only Oracle but our own Weld has this. It is awesome. I googled
error codes and fixed the problem immediately. Fantastic idea!
On 11-03-16 6:39 AM, Manik Surtani wrote:
Yup. Essentially if anyone's ever worked with Oracle DB,
plugging an
ORA-12345 error code into Google usually gives you details of the
error regardless of the i18n error message you find in the logs.
On 16 Mar 2011, at 09:08, Emmanuel Bernard wrote:
> The road taken by JBoss Logging is to:
> - associate an alphanumerical code that you can search on the
> internet. And of course by default messages are in English.
> - use methods to generate the log call
>
> *in Infinispan code*
> if ( txLog == null ) {
> LOG.txLogPushError(txId);
> }
>
> @MessageLogger( projectCode = "ISPN" )
> public interfaceInfinispanLogger extends BasicLogger {
> @LogMessage( level = ERROR )
> @Message( value = "Error pushing tx log (tx id=%1$)", id = 345 )
> void txLogPushError(String txId);
>
> *in your app log*
> ISPN000345 Error pushing tx log (tx id=3232)
>
>
> *later on, on a wiki page*
>
> ISPN-000345 Error pushing tx log
> When this happens, you're screwed unless you use JBoss Transaction
> and then...
>
>
> I could not really find the doc for JBoss Logging but David will know
> where it is.
>
> On 16 mars 2011, at 09:42, Tristan Tarrant wrote:
>
>> The only problem I have with this is that if all error / logging
>> messages are i18n'ed, they will be difficult to search for on the
>> web / forums / wikis.
>> For this reason all error messages should include a
>> locale-independent error code. Numeric is ok, but I'd rather have a
>> "readable" error code.
>> For example the string "Error pushing tx log"
>> in InvertedLeaveTask.java would become "TX_LOG_PUSH_ERR: Error
>> pushing tx log".
>>
>> Any thoughts on this ?
>>
>> Tristan
>>
>> On Wed, Mar 16, 2011 at 09:09, Galder Zamarreño <galder(a)redhat.com
>> <mailto:galder@redhat.com>> wrote:
>>
>>
>> On Mar 15, 2011, at 6:29 PM, Manik Surtani wrote:
>>
>> >
>> > On 15 Mar 2011, at 17:21, Mircea Markus wrote:
>> >
>> >>
>> >> On 15 Mar 2011, at 17:17, Emmanuel Bernard wrote:
>> >>
>> >>> As David pointed out %1$ is superior to %s for
>> internationalization.
>> >> Agreed.
>> >> Not sure we want i18n for logs and error messages (?)
>> >
>> > It's on our roadmap and a requirement for a few downstream
>> projects/products.
>>
>> Yup: https://issues.jboss.org/browse/ISPN-380
>>
>> >
>> > --
>> > Manik Surtani
>> > manik(a)jboss.org <mailto:manik@jboss.org>
>> > twitter.com/maniksurtani <http://twitter.com/maniksurtani>
>> >
>> > Lead, Infinispan
>> > http://www.infinispan.org <http://www.infinispan.org/>
>> >
>> >
>> >
>> > _______________________________________________
>> > infinispan-dev mailing list
>> > infinispan-dev(a)lists.jboss.org
>> <mailto:infinispan-dev@lists.jboss.org>
>> > https://lists.jboss.org/mailman/listinfo/infinispan-dev
>>
>> --
>> Galder Zamarreño
>> Sr. Software Engineer
>> Infinispan, JBoss Cache
>>
>>
>> _______________________________________________
>> infinispan-dev mailing list
>> infinispan-dev(a)lists.jboss.org
>> <mailto:infinispan-dev@lists.jboss.org>
>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>>
>>
>> _______________________________________________
>> infinispan-dev mailing list
>> infinispan-dev(a)lists.jboss.org <mailto:infinispan-dev@lists.jboss.org>
>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org <mailto:infinispan-dev@lists.jboss.org>
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Manik Surtani
manik(a)jboss.org <mailto:manik@jboss.org>
twitter.com/maniksurtani <http://twitter.com/maniksurtani>
Lead, Infinispan
http://www.infinispan.org
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
9:23 a.m.
in future we might have the logger output a link to a wiki page, one
page for each error, where people can write about issue context and
possible solutions, or links to JIRAs, fixed versions. Then we can
collect statistics on those page visits..
2011/3/16 Vladimir Blagojevic <vblagoje(a)redhat.com>:
Not only Oracle but our own Weld has this. It is awesome. I googled
error
codes and fixed the problem immediately. Fantastic idea!
On 11-03-16 6:39 AM, Manik Surtani wrote:
Yup. Essentially if anyone's ever worked with Oracle DB, plugging an
ORA-12345 error code into Google usually gives you details of the error
regardless of the i18n error message you find in the logs.
On 16 Mar 2011, at 09:08, Emmanuel Bernard wrote:
The road taken by JBoss Logging is to:
- associate an alphanumerical code that you can search on the internet. And
of course by default messages are in English.
- use methods to generate the log call
in Infinispan code
if ( txLog == null ) {
LOG.txLogPushError(txId);
}
@MessageLogger( projectCode = "ISPN" )
public interfaceInfinispanLogger extends BasicLogger {
@LogMessage( level = ERROR )
@Message( value = "Error pushing tx log (tx id=%1$)", id = 345 )
void txLogPushError(String txId);
in your app log
ISPN000345 Error pushing tx log (tx id=3232)
later on, on a wiki page
ISPN-000345 Error pushing tx log
When this happens, you're screwed unless you use JBoss Transaction and
then...
I could not really find the doc for JBoss Logging but David will know where
it is.
On 16 mars 2011, at 09:42, Tristan Tarrant wrote:
The only problem I have with this is that if all error / logging messages
are i18n'ed, they will be difficult to search for on the web / forums /
wikis.
For this reason all error messages should include a locale-independent error
code. Numeric is ok, but I'd rather have a "readable" error code.
For example the string "Error pushing tx log" in InvertedLeaveTask.java
would become "TX_LOG_PUSH_ERR: Error pushing tx log".
Any thoughts on this ?
Tristan
On Wed, Mar 16, 2011 at 09:09, Galder Zamarreño <galder(a)redhat.com> wrote:
>
> On Mar 15, 2011, at 6:29 PM, Manik Surtani wrote:
>
> >
> > On 15 Mar 2011, at 17:21, Mircea Markus wrote:
> >
> >>
> >> On 15 Mar 2011, at 17:17, Emmanuel Bernard wrote:
> >>
> >>> As David pointed out %1$ is superior to %s for internationalization.
> >> Agreed.
> >> Not sure we want i18n for logs and error messages (?)
> >
> > It's on our roadmap and a requirement for a few downstream
> > projects/products.
>
> Yup: https://issues.jboss.org/browse/ISPN-380
>
> >
> > --
> > Manik Surtani
> > manik(a)jboss.org
> > twitter.com/maniksurtani
> >
> > Lead, Infinispan
> > http://www.infinispan.org
> >
> >
> >
> > _______________________________________________
> > infinispan-dev mailing list
> > infinispan-dev(a)lists.jboss.org
> > https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
> --
> Galder Zamarreño
> Sr. Software Engineer
> Infinispan, JBoss Cache
>
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Manik Surtani
manik(a)jboss.org
twitter.com/maniksurtani
Lead, Infinispan
http://www.infinispan.org
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
11:18 a.m.
Also a good idea. The approach may well be to start with error codes though, and have a
error code reference on our wiki pointing to specific pages per error code - which would
be picked up by search engines, etc.
On 16 Mar 2011, at 14:23, Sanne Grinovero wrote:
in future we might have the logger output a link to a wiki page, one
page for each error, where people can write about issue context and
possible solutions, or links to JIRAs, fixed versions. Then we can
collect statistics on those page visits..
2011/3/16 Vladimir Blagojevic <vblagoje(a)redhat.com>:
> Not only Oracle but our own Weld has this. It is awesome. I googled error
> codes and fixed the problem immediately. Fantastic idea!
>
> On 11-03-16 6:39 AM, Manik Surtani wrote:
>
> Yup. Essentially if anyone's ever worked with Oracle DB, plugging an
> ORA-12345 error code into Google usually gives you details of the error
> regardless of the i18n error message you find in the logs.
> On 16 Mar 2011, at 09:08, Emmanuel Bernard wrote:
>
> The road taken by JBoss Logging is to:
> - associate an alphanumerical code that you can search on the internet. And
> of course by default messages are in English.
> - use methods to generate the log call
>
> in Infinispan code
> if ( txLog == null ) {
> LOG.txLogPushError(txId);
> }
> @MessageLogger( projectCode = "ISPN" )
> public interfaceInfinispanLogger extends BasicLogger {
> @LogMessage( level = ERROR )
> @Message( value = "Error pushing tx log (tx id=%1$)", id = 345 )
> void txLogPushError(String txId);
> in your app log
> ISPN000345 Error pushing tx log (tx id=3232)
>
> later on, on a wiki page
> ISPN-000345 Error pushing tx log
> When this happens, you're screwed unless you use JBoss Transaction and
> then...
>
>
> I could not really find the doc for JBoss Logging but David will know where
> it is.
> On 16 mars 2011, at 09:42, Tristan Tarrant wrote:
>
> The only problem I have with this is that if all error / logging messages
> are i18n'ed, they will be difficult to search for on the web / forums /
> wikis.
> For this reason all error messages should include a locale-independent error
> code. Numeric is ok, but I'd rather have a "readable" error code.
> For example the string "Error pushing tx log" in InvertedLeaveTask.java
> would become "TX_LOG_PUSH_ERR: Error pushing tx log".
> Any thoughts on this ?
> Tristan
> On Wed, Mar 16, 2011 at 09:09, Galder Zamarreño <galder(a)redhat.com> wrote:
>>
>> On Mar 15, 2011, at 6:29 PM, Manik Surtani wrote:
>>
>>>
>>> On 15 Mar 2011, at 17:21, Mircea Markus wrote:
>>>
>>>>
>>>> On 15 Mar 2011, at 17:17, Emmanuel Bernard wrote:
>>>>
>>>>> As David pointed out %1$ is superior to %s for internationalization.
>>>> Agreed.
>>>> Not sure we want i18n for logs and error messages (?)
>>>
>>> It's on our roadmap and a requirement for a few downstream
>>> projects/products.
>>
>> Yup: https://issues.jboss.org/browse/ISPN-380
>>
>>>
>>> --
>>> Manik Surtani
>>> manik(a)jboss.org
>>> twitter.com/maniksurtani
>>>
>>> Lead, Infinispan
>>> http://www.infinispan.org
>>>
>>>
>>>
>>> _______________________________________________
>>> infinispan-dev mailing list
>>> infinispan-dev(a)lists.jboss.org
>>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>>
>> --
>> Galder Zamarreño
>> Sr. Software Engineer
>> Infinispan, JBoss Cache
>>
>>
>> _______________________________________________
>> infinispan-dev mailing list
>> infinispan-dev(a)lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
> --
> Manik Surtani
> manik(a)jboss.org
> twitter.com/maniksurtani
> Lead, Infinispan
> http://www.infinispan.org
>
>
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Manik Surtani
manik(a)jboss.org
twitter.com/maniksurtani
Lead, Infinispan
http://www.infinispan.org
9:09 a.m.
Yes I think people really like the "certainty" associated with error codes -
they are sure the error they are getting is the same as the other persons. Now we just
need to actually start using the codes in wiki pages ;-)
On 16 Mar 2011, at 14:18, Vladimir Blagojevic wrote:
Not only Oracle but our own Weld has this. It is awesome. I googled
error codes and fixed the problem immediately. Fantastic idea!
On 11-03-16 6:39 AM, Manik Surtani wrote:
> Yup. Essentially if anyone's ever worked with Oracle DB, plugging an ORA-12345
error code into Google usually gives you details of the error regardless of the i18n error
message you find in the logs.
>
> On 16 Mar 2011, at 09:08, Emmanuel Bernard wrote:
>
>> The road taken by JBoss Logging is to:
>> - associate an alphanumerical code that you can search on the internet. And of
course by default messages are in English.
>> - use methods to generate the log call
>>
>> in Infinispan code
>> if ( txLog == null ) {
>> LOG.txLogPushError(txId);
>> }
>>
>> @MessageLogger( projectCode = "ISPN" )
>> public interfaceInfinispanLogger extends BasicLogger {
>> @LogMessage( level = ERROR )
>> @Message( value = "Error pushing tx log (tx id=%1$)", id = 345 )
>> void txLogPushError(String txId);
>>
>> in your app log
>> ISPN000345 Error pushing tx log (tx id=3232)
>>
>>
>> later on, on a wiki page
>>
>> ISPN-000345 Error pushing tx log
>> When this happens, you're screwed unless you use JBoss Transaction and
then...
>>
>>
>> I could not really find the doc for JBoss Logging but David will know where it
is.
>>
>> On 16 mars 2011, at 09:42, Tristan Tarrant wrote:
>>
>>> The only problem I have with this is that if all error / logging messages are
i18n'ed, they will be difficult to search for on the web / forums / wikis.
>>> For this reason all error messages should include a locale-independent error
code. Numeric is ok, but I'd rather have a "readable" error code.
>>> For example the string "Error pushing tx log" in
InvertedLeaveTask.java would become "TX_LOG_PUSH_ERR: Error pushing tx log".
>>>
>>> Any thoughts on this ?
>>>
>>> Tristan
>>>
>>> On Wed, Mar 16, 2011 at 09:09, Galder Zamarreño <galder(a)redhat.com>
wrote:
>>>
>>> On Mar 15, 2011, at 6:29 PM, Manik Surtani wrote:
>>>
>>> >
>>> > On 15 Mar 2011, at 17:21, Mircea Markus wrote:
>>> >
>>> >>
>>> >> On 15 Mar 2011, at 17:17, Emmanuel Bernard wrote:
>>> >>
>>> >>> As David pointed out %1$ is superior to %s for
internationalization.
>>> >> Agreed.
>>> >> Not sure we want i18n for logs and error messages (?)
>>> >
>>> > It's on our roadmap and a requirement for a few downstream
projects/products.
>>>
>>> Yup: https://issues.jboss.org/browse/ISPN-380
>>>
>>> >
>>> > --
>>> > Manik Surtani
>>> > manik(a)jboss.org
>>> > twitter.com/maniksurtani
>>> >
>>> > Lead, Infinispan
>>> > http://www.infinispan.org
>>> >
>>> >
>>> >
>>> > _______________________________________________
>>> > infinispan-dev mailing list
>>> > infinispan-dev(a)lists.jboss.org
>>> > https://lists.jboss.org/mailman/listinfo/infinispan-dev
>>>
>>> --
>>> Galder Zamarreño
>>> Sr. Software Engineer
>>> Infinispan, JBoss Cache
>>>
>>>
>>> _______________________________________________
>>> infinispan-dev mailing list
>>> infinispan-dev(a)lists.jboss.org
>>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>>>
>>> _______________________________________________
>>> infinispan-dev mailing list
>>> infinispan-dev(a)lists.jboss.org
>>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>>
>> _______________________________________________
>> infinispan-dev mailing list
>> infinispan-dev(a)lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
> --
> Manik Surtani
> manik(a)jboss.org
> twitter.com/maniksurtani
>
> Lead, Infinispan
> http://www.infinispan.org
>
>
>
>
> _______________________________________________
> infinispan-dev mailing list
>
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
12:20 p.m.
If you use the JBoss standard i18n library, you can put things like this
on your message bundle interface:
@Message("You're not allowed to use %s and %s together")
ConfigurationException cantUseTogether(String that, String other);
then you can just do:
throw cantUseTogether(that, other);
with the advantage that it can be translated into other languages.
On 03/15/2011 12:13 PM, Manik Surtani wrote:
Yeah, I currently do:
throw new ConfigurationException(String.format("you 're not allowed to use %s
and %s together", that, other));
but we could overload the constructor of CacheException ...
On 15 Mar 2011, at 16:15, Mircea Markus wrote:
> Hi,
>
> Not a biggie, but I think it would be nice to use this nicer syntax for exception
messages, e.g.
> throw new ConfigurationException("you 're not allowed to use %s and %s
together", that, other);
> wdyt?
>
> Cheers,
> Mircea
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Manik Surtani
manik(a)jboss.org
twitter.com/maniksurtani
Lead, Infinispan
http://www.infinispan.org
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
- DML
5028
days inactive
5031
days old
infinispan-dev@lists.jboss.org
13 comments
9 participants
participants (9)
-
David M. Lloyd -
Emmanuel Bernard -
Galder Zamarreño -
Manik Surtani -
Mircea Markus -
Pete Muir -
Sanne Grinovero -
Tristan Tarrant -
Vladimir Blagojevic