headers
Wolfgang Laun | 22 Jun 2012 20:36
Picon

[rules-dev] Bugs

Lately, I have occasionally tried to raise an alert for something I
consider a bug, but there was no comment.

Here's another one.

rule "Faults coincide"
when
    f1 : MyEvent( key == "faultType1", value == "ALARM" ) over
window:length( 1 )
    f2 : MyEvent( key == "faultType2", value == "ALARM" ) over
window:length( 1 )
then
    System.out.println( "************ both faults are active! f1.key =
" + f1.getKey() + " f2.key = " + f2.getKey() );
end

This rule fires after the insertion of a single fact matching the
event bound to f1.

Detailed code can be found on the user's list.

-W
_______________________________________________
rules-dev mailing list
rules-dev <at> lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-dev
Permalink | Reply |
headers
Geoffrey De Smet | 22 Jun 2012 22:19
Picon
Gravatar

Re: [rules-dev] Bugs

If you're sure it's a bug, it's probably a good idea to raise a jira too 
(and link it in any mail correspondence about it).
That way it has a status (is it still open or is it already resolved for 
the next release?).

But to really get the ball rolling sooner rather then later for an 
issue: create a pull request with an isolated regression test, for 
example on MiscTest.
It takes more time on your end, but it assures the developer he isn't 
spending time in invalid bugs
and it makes it really easy for him to see the code quickly on the github.
Just take a look on the track record of pull requests for bugs (not 
features), they are handled swiftly.

Note: I am not saying we shouldn't pick up on bugs reported on the 
mailing lists (we try to),
but pull requests naturally get priority because they are far more accurate.

Op 22-06-12 20:36, Wolfgang Laun schreef:
> Lately, I have occasionally tried to raise an alert for something I
> consider a bug, but there was no comment.
>
> Here's another one.
>
> rule "Faults coincide"
> when
>      f1 : MyEvent( key == "faultType1", value == "ALARM" ) over
> window:length( 1 )
>      f2 : MyEvent( key == "faultType2", value == "ALARM" ) over
> window:length( 1 )
(Continue reading)

Permalink | Reply |
headers
Wolfgang Laun | 23 Jun 2012 11:29
Picon

Re: [rules-dev] Bugs

On 22/06/2012, Geoffrey De Smet <ge0ffrey.spam <at> gmail.com> wrote:
> If you're sure it's a bug, it's probably a good idea to raise a jira too
> (and link it in any mail correspondence about it).
> That way it has a status (is it still open or is it already resolved for
> the next release?).

The number of my JIRAs that remain open is considerable.

>
> But to really get the ball rolling sooner rather then later for an
> issue: create a pull request with an isolated regression test, for
> example on MiscTest.

I invest enough work to identify the bug, and isolate it in a small
stand-alone application that can be run from the command line.
Currently I just don't have the resources to run git to stay
up-to-date and create pull requests. My small apps can be run just as
easily and more swiftly as a test in some framework.

If there were an easy way to upload such a stand-alone app, I'd do it.

> It takes more time on your end, but it assures the developer he isn't
> spending time in invalid bugs

I'm not saying that I never raised an invalid bug, but that has
happened very, very rarely. And I consider pointing out an issue on
the mailing list as an invitation to comment or review, just to get a
"second opinion". But if this is ignored, then I don't bother any
more.
(Continue reading)

Permalink | Reply |
headers
Geoffrey De Smet | 24 Jun 2012 09:27
Picon
Gravatar

Re: [rules-dev] Bugs

You're doing a great job, reporting these issues, don't get me wrong :) 
Keep it up.

Op 23-06-12 11:29, Wolfgang Laun schreef:
> On 22/06/2012, Geoffrey De Smet<ge0ffrey.spam <at> gmail.com>  wrote:
>> If you're sure it's a bug, it's probably a good idea to raise a jira too
>> (and link it in any mail correspondence about it).
>> That way it has a status (is it still open or is it already resolved for
>> the next release?).
> The number of my JIRAs that remain open is considerable.
>
>> But to really get the ball rolling sooner rather then later for an
>> issue: create a pull request with an isolated regression test, for
>> example on MiscTest.
> I invest enough work to identify the bug, and isolate it in a small
> stand-alone application that can be run from the command line.
> Currently I just don't have the resources to run git to stay
> up-to-date and create pull requests. My small apps can be run just as
> easily and more swiftly as a test in some framework.
>
> If there were an easy way to upload such a stand-alone app, I'd do it.
>
>> It takes more time on your end, but it assures the developer he isn't
>> spending time in invalid bugs
> I'm not saying that I never raised an invalid bug, but that has
> happened very, very rarely. And I consider pointing out an issue on
> the mailing list as an invitation to comment or review, just to get a
> "second opinion". But if this is ignored, then I don't bother any
> more.
>
(Continue reading)

Permalink | Reply |
headers
Greg Barton | 24 Jun 2012 18:42
Picon
Favicon

Re: [rules-dev] Bugs


--- On Sat, 6/23/12, Wolfgang Laun <wolfgang.laun <at> gmail.com> wrote:

> If there were an easy way to upload such a stand-alone app,
> I'd do it.

You can attach it to a JIRA ticket. 

_______________________________________________
rules-dev mailing list
rules-dev <at> lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-dev
Permalink | Reply |

Gmane