ignite-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ivan Pavlukhin <vololo...@gmail.com>
Subject Re: JavaDoc for Event's subjectId methods
Date Mon, 30 Sep 2019 07:59:44 GMT
Hi,

Do we allow commits to master without a ticket? I can imagine only
reverts as an exception.

Otherwise a ticket is a primary process item. Work description,
review, CI checks (we have a job checking javadocs).

ср, 25 сент. 2019 г. в 01:15, Denis Magda <dmagda@apache.org>:
>
> Denis, please feel free to go and edit the JavaDocs in place without a
> ticket. The changes suggested by you are reasonable.
>
> -
> Denis
>
>
> On Tue, Sep 24, 2019 at 3:55 AM Denis Garus <garus.d.g@gmail.com> wrote:
>
> > Hello, Igniters!
> >
> > Some events contain the subjectId method, for example, TaskEvent#subjectId.
> > The JavaDoc for this method is:
> > "Gets security subject ID initiated this task event, if available.
> > This property is not available for GridEventType#EVT_TASK_SESSION_ATTR_SET
> > task event.
> > Subject ID will be set either to node ID or client ID initiated task
> > execution."
> >
> > I think It's wrong. The main point is a subject id doesn't have any sense
> > if IgniteSecurity is disabled.
> > However, if IgniteSecurity is enabled, the method must return the subject
> > id from the current security context.
> > Thus, the description (and behavior) of the method should be the following:
> > Gets security subject ID initiated this task event if IgniteSecurity is
> > enabled, otherwise returns null.
> >
> > The same is actual for CacheEvent, CacheQueryExecutedEvent and
> > CacheQueryReadEvent.
> >
> > If there are no objections, I am going to create a relevant issue in Jira.
> >



-- 
Best regards,
Ivan Pavlukhin

Mime
View raw message