[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: Async action with helper in KAuth
From:       Konstantinos Smanis <konstantinos.smanis () gmail ! com>
Date:       2012-12-06 15:16:13
Message-ID: CANJf_cAp_ng6exzyZChKD7ZRm6xwBcPQKwu9UBtD-t1tBciFRg () mail ! gmail ! com
[Download RAW message or body]

On Thu, Nov 15, 2012 at 5:08 PM, Dario Freddi <drf54321@gmail.com> wrote:
>
> 2012/11/15 Yichao Yu <yyc1992@gmail.com>
>>
>> Filed a bug here[1], hopefully this will help you tracking the bug and
>> to cherrypick/fix it in other branches.
>>
>> [1] https://bugs.kde.org/show_bug.cgi?id=310149
>
>
> Thanks!
>
>>
>>
>>
>> On Thu, Nov 15, 2012 at 8:45 AM, Dario Freddi <drf54321@gmail.com> wrote:
>> > Hello Yichao,
>> >
>> > 2012/11/12 Yichao Yu <yyc1992@gmail.com>
>> >>
>> >> I checkout out the code for KAuth and found out the following lines
>> >> in` ActionReply Action::execute(const QString &helperID) const`
>> >>
>> >>     if (d->async) {
>> >>         if (hasHelper()) {
>> >>             // It makes no sense
>> >>             return ActionReply::InvalidActionReply;
>> >>         }
>> >>
>> >> which gives the right error code and can explain why only the async
>> >> action fails. The comment says "makes no sense", but doesn't say
>> >> why.... can anyone explain the reason??
>> >
>> >
>> > Looks like there is a logical error in the code that I forgot to fix in
>> > the
>> > 4.x branch. if (!hasHelper()) is the correct line, and if the action has
>> > no
>> > helper it doesn't make sense to execute it asynchronously since there is
>> > actually nothing to do, but only to authorize. Sorry about that, I will
>> > provide a fix ASAP - my main focus lately has been the framework branch,
>> > where most of the logic actually changed.
>> >
>> >>
>> >> And I cannot found the
>> >> description for this anywhere in the tutorial (e.g. some backend may
>> >> not support async action with helper at all.....), if it really
>> >> doesn't make sense to do this (which I really doubt....) I guess it
>> >> would be better to add it...
>> >
>> >
>> > What you are doing is perfectly supported and the bug is in the library.
>> >
>> >>
>> >>
>> >> THX.
>> >>
>> >>
>> >> [1] http://forum.kde.org/viewtopic.php?f=43&t=94801
>> >>
>> >> >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to
>> >> >> unsubscribe <<
>> >
>> >
>> >
>> >
>> >>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to
>> >>> unsubscribe
>> >>> <<
>> >
>
>
>
>
>>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe
>>> <<
>

Why is this taking so long to fix? Does it go deeper than the logical error?

Kind regards
-- 
Konstantinos Smanis

>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
[prev in list] [next in list] [prev in thread] [next in thread]