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

List:       kfm-devel
Subject:    Re: KIOSlave errors an konqueror's handling of them
From:       Hamish Rodda <meddie () yoyo ! cc ! monash ! edu ! au>
Date:       2002-01-25 15:37:34
[Download RAW message or body]

>On Friday 25 January 2002 15:56, Hamish Rodda wrote:
>> >Yup, but you also need to add support for error:// in khtml, and call=
 it
>> > in KonqRun::handleError()  (as I see it)
>>
>> I was just going to hack away at KHTMLPart... pseudo code:
>>
>> void KHTMLPart::showError( KIO::Job* job )
>> if user wants html error pages
>> call and write data from buildHTMLErrorString
>> otherwise call and write data from buildErrorString
>
>Yes, that's the necessary good start, although it would help if you woul=
d
> make that code another protected method, called by showError, so that i=
t
> can take the errorcode, errorstring and URL as argument. Those would co=
me
> from the job in the above case, but they could come from the error:// U=
RL
> in case the error was detected by KonqRun.

Ah, I see... I think...

>I was just saying that this isn't enough to fix the "get() after a stat(=
)
> that gave an error" problem, for that the error:// support has to be ad=
ded
> (in khtml it's simply about retrieving the 3 params from the url, and
> calling the method I'm talking about above). And in Konq it's simply ab=
out
> setting m_strURL to that URL, in KonqRun::handleError.

I see... got it. Would you like to review this before it goes in?

>> Now, is there any way I could include active links to KDE services, su=
ch
>> as control panel options, without being validly criticised for introdu=
cing
>> security concerns?
>
>I think there should be no problem, you can simply add a bool that
> authorizes execution of programs and scripts w/o confirmation, and sett=
ing
> this bool to true in showError().

Good, this is getting pretty interesting...

>> I asked on #kde if anyone wanted to help... they asked if I would pay
>> them:)
>
>Lazy bast...s ;)

They were, however, happy to add {constructive, funny} criticisms...

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