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

List:       kde-core-devel
Subject:    Re: kdDebug modifying value ?? [Re: kio_http bug, introduced by a feature yesterday.]
From:       David Faure <david () mandrakesoft ! com>
Date:       2000-10-17 22:57:01
[Download RAW message or body]

On Tuesday 17 October 2000 22:47, Dawit Alemayehu wrote :
>On Tue, 17 Oct 2000, Stephan Kulow wrote:
>> Dawit Alemayehu wrote:
>> > Hi,
>> >
>> > Okay, the debug statement just before openPassDlg in kio_http needs to be
>> > diabled!!  It modifies the value of the variable supplied can causes the
>> > open password dialog to make the username entry editable when it should
>> > not.
>> >
>> > To reporduce do:
>> >
>> > 1.) Type in username when you enter a URL (http://me@somehost)
>> > 2.) Enter the wrong password on purpose.
>> > 3.) Say "yes" when prompted to retry
>> >
>> > and wala now the name is editable.  This was caused by that kdDebug
>> > statement just before the openPassDlg(...).  Does kdDebug modify supplied
>> > values ??
>>
>> Of course it does not. In what way should it modify it? It is
>> given url.user(), right? And even if it would modify that QString,
>> it still shouldn't affect what user() returns next time as it's
>> a new QString. I think you have to search for another one to blame :)
>
>No need to blame anyone.  I was just as surprised by what I got, henece the "?" in my
>subject.  It seemed weird  to me, but that is exactly what happened to me.  I put that
>debug statement there to see if the user() field was set or not to begin with.  And it works
>perfectly.  However, if you follow the steps I have outlined above on the second iteration
>the statement that checks whether user() is empty fails !!  That is the
>(!m_request.url.user().isEmpty()) part in openPassDlg returns false.  Don't know why, but it
>does.  The weird thing is that if I comment the debug statement out, I get the correct result
>all the time!!!  This is why I asked this question.  Anyways,  I have commented the debug
>statement out since it is no longer needed, but it was just a weird problem for me...

In any case, using hasUser() seems better than .user().isEmpty() to me, but I suppose
the former calls the latter anyway :)

When a debug statement changes something, it means there is some memory corruption
somewhere - or you just confused by something....

PS: unrelated... Is it normal that I can't make a google query containing a '.', using gg:blah.blah ?
It resolves that into a short-url (something like http://ggblah.blah which looks wrong to me).

-- 
David FAURE, david@mandrakesoft.com, faure@kde.org
http://www.mandrakesoft.com/~david/, http://www.konqueror.org/
KDE, Making The Future of Computing Available Today
See http://www.kde.org/kde1-and-kde2.html for how to set up KDE 2

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