You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@wicket.apache.org by Korbinian Bachl <ko...@whiskyworld.de> on 2006/11/11 17:21:14 UTC
Severe Bug in Wicket 2 / WICKET-42
Hi,
I issued a severe BUG
http://issues.apache.org/jira/browse/WICKET-42
Best Regards,
Korbinian
Re: Severe Bug in Wicket 2 / WICKET-42
Posted by Igor Vaynberg <ig...@gmail.com>.
i dont think this is a very common usecase, at least it has never been
requested before. so what i am tempted to tell you is to implement your own
IRequestCodingStrategy to handle this. and if you would like you can provide
it as a patch and we might put it into extensions or core.
-igor
On 11/13/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
>
> Hi Johan,
>
> yes i know its not component, and its evil :O
>
> the *.html would mean a unique identifier where a filter could be attached
> to and so static resources wouldnt be touched.
>
> The one class - well, i know its a hard discussion, but as we also have
> wicket-apps where whole app is just 1 page that swaps around components in
> it this doenst sound too exotic for me.
>
> regards
>
>
> > -----Ursprüngliche Nachricht-----
> > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > Gesendet: Montag, 13. November 2006 20:29
> > An: wicket-dev@incubator.apache.org
> > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> >
> > you can have .html just fine.
> > Only you want to have xxx/param.html?
> > so like:
> >
> > /product/110203.html?
> > or
> > /product/id/110203.html?
> >
> > Why that html there?
> > What does that bring?
> > I don't find it nicer that is just taste.
> > And i guess you could build it just fine. Just make another
> > url encoder.
> >
> > And having one class handling everything.. a really really
> > big YUK or ARRGGH
> >
> > thats horrible. We are in a component oriented framework
> > here. Where small pieces do there job Not one big page where
> > you parse the page parameters.
> >
> > You could do this by the way just fine now.
> > Only have have to mount those pages not on / (thats homepage)
> > but on /mybigentry
> >
> >
> > johan
> >
> >
> >
> > On 11/13/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > >
> > > Yeah similar to that. I mean we currently even cant employ
> > file-endings.
> > >
> > > e.g: you can have /page/param/param
> > > but not /page/param/param.html wich would be even more nice
> > and logic
> > > (IMHO)
> > >
> > > imagine a mount where you could specify "/" as begin and ".htm" as
> > > end. -> you could have 1 handling-class for all data as well as
> > > natural URLs (if you use IndexedURLParams).
> > > if / is solely you could return the start page, if /.htm is called
> > > wicket knows that it would be the handling-class with empty
> > > params...so we also wouldnt crash with any other wicket
> > functionality.
> > >
> > > e.g: globalmount(".htm", new
> > > IndexedUrlParamStrategy("/",handler.class));
> > > so you also would centralise URLs to just 1 class in app
> > wich then can
> > > delegate to the resulting pages.
> > >
> > > was just a thought from what im in now as i need nice URLs and the
> > > problems i stumbled in regarding static content, root etc.
> > >
> > > what do you think about this? - i'm interested in any
> > thoughts about it.
> > >
> > > Best Regards,
> > >
> > > Korbinian
> > >
> > >
> > >
> > >
> > >
> > > > -----Ursprüngliche Nachricht-----
> > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > Gesendet: Montag, 13. November 2006 10:20
> > > > An: wicket-dev@incubator.apache.org
> > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > >
> > > > So this way?
> > > >
> > > > / == home page
> > > > /?p=a == homepage with page params.
> > > > /?interface:xxx == an callback to an existing page
> > > > /?bookmarkablePage=xxx == an none mounted bookmarkable
> > page request
> > > > /input == mounted bookmarkable page
> > > >
> > > > and then
> > > >
> > > > /a == this is not mounted like input above.. Then it
> > should go to a
> > > > page that was mounted on /* ??
> > > > So you want to have a fallback to a page?
> > > >
> > > > johan
> > > >
> > > >
> > > >
> > > > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > >
> > > > > Yeah, that is what i mean -
> > > > >
> > > > > e.g: mount to /*.html -> would mean it goes for
> > /[everything].html
> > > > > wich would introduce even more flexibility...
> > > > >
> > > > > currently, it goes to "/[neededCharacter]" and this catches
> > > > > /[neededCharacter][anythingBehind]
> > > > > e.g: /a ->
> > > > > /a/b is valid
> > > > > /aa/b is valid
> > > > > /a is valid
> > > > >
> > > > > as / is mounted for the homepage and you cant mount a
> > > > wildcard so far,
> > > > > you still need a fixed point you can hook on.
> > > > >
> > > > > Regards
> > > > >
> > > > >
> > > > > > -----Ursprüngliche Nachricht-----
> > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > Gesendet: Sonntag, 12. November 2006 20:15
> > > > > > An: wicket-dev@incubator.apache.org
> > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > >
> > > > > > what does a page mounted on /* means in your eyes?
> > > > > > /[everything]??
> > > > > >
> > > > > > there is no such a thing like wild card mounting.
> > > > > >
> > > > > > johan
> > > > > >
> > > > > >
> > > > > > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > > > >
> > > > > > > Hi Igor,
> > > > > > >
> > > > > > > yes - but its not mounted to "/*" but to "/" - is it
> > > > > > possible to have
> > > > > > > the homepage mounted on "/*" ? means that this page gets
> > > > > > all path as
> > > > > > > indexedURLparams??
> > > > > > >
> > > > > > > if i mount it to /* he doesnt see it as wildcard but as
> > > > fixed "*"
> > > > > > >
> > > > > > > Regards
> > > > > > >
> > > > > > >
> > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > Von: Igor Vaynberg [mailto:igor.vaynberg@gmail.com]
> > > > > > > > Gesendet: Sonntag, 12. November 2006 18:39
> > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > >
> > > > > > > > the homepage is always mounted on /
> > > > > > > >
> > > > > > > > -igor
> > > > > > > >
> > > > > > > >
> > > > > > > > On 11/12/06, Korbinian Bachl
> > <ko...@whiskyworld.de> wrote:
> > > > > > > > >
> > > > > > > > > Johan,
> > > > > > > > >
> > > > > > > > > i checked all - everything is now behaving as expected.
> > > > > > Great Work!
> > > > > > > > >
> > > > > > > > > One question however is still to me - why cant i mount a
> > > > > > > > page to "/" ?
> > > > > > > > >
> > > > > > > > > so i could mount a PageDistribution class there
> > which then
> > > > > > > > handles the
> > > > > > > > > PageParams logic... or would this get in conflict with
> > > > > > the rest
> > > > > > > > > of wicket ?
> > > > > > > > > if so, would it be possible to have a page
> > responsible for
> > > > > > > > every path
> > > > > > > > > behind / but for /wicket/ where the internal could
> > > > be handled ??
> > > > > > > > >
> > > > > > > > > i know this aproach is a bit away from what wicket was
> > > > > > made for...
> > > > > > > > >
> > > > > > > > > Rgards
> > > > > > > > >
> > > > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > > > Von: Korbinian Bachl [mailto:korbinian@whiskyworld.de]
> > > > > > > > > > Gesendet: Sonntag, 12. November 2006 10:32
> > > > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > > > Betreff: AW: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > > > >
> > > > > > > > > > Really fast :)
> > > > > > > > > >
> > > > > > > > > > i check it out soon and report,
> > > > > > > > > >
> > > > > > > > > > Regards
> > > > > > > > > >
> > > > > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > > > > > > Gesendet: Samstag, 11. November 2006 23:17
> > > > > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > > > > >
> > > > > > > > > > > should be fixed now.
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > On 11/11/06, Korbinian Bachl
> > > > > > <ko...@whiskyworld.de> wrote:
> > > > > > > > > > > >
> > > > > > > > > > > > well, it works if it is:
> > > > > > > > > > > > a, the first access of the WebApp with no provious
> > > > > > > > > > ReDeploy b, the
> > > > > > > > > > > > root is accessed then
> > > > > > > > > > > >
> > > > > > > > > > > > if i redeploy just once, i cant access the
> > > > root anymore
> > > > > > > > > > > > -
> > > > > > > > > > > however the
> > > > > > > > > > > > direct path always works...
> > > > > > > > > > > > if i redeploy it and access the path before, it
> > > > > > > > > > > > chains
> > > > > > > > > > > >
> > > > > > > > > > > > another danger thing is that following works:
> > > > > > > > > > > > assume you have IndexedPageParams and index.class
> > > > > > > > > > > > mounted
> > > > > > > > > > to /Index:
> > > > > > > > > > > >
> > > > > > > > > > > > if you then access it via
> > /Index/foo/bar/Index/foo2
> > > > > > > > > > > >
> > > > > > > > > > > > and print out pageparams.toString you only
> > > > get 0=foo2 ,
> > > > > > > > > > meaning he
> > > > > > > > > > > > access it at last point, and not at first!
> > > > > > > > > > > > This violates JEE security !!!
> > > > > > > > > > > >
> > > > > > > > > > > > i also noticed that if you use a servlet and
> > > > > > > > > > > BookMarkablePageLinks the
> > > > > > > > > > > > URLs look like these
> > > > > > 127.0.0.1/<servlet>?param=value - and
> > > > > > > > > > > mounting
> > > > > > > > > > > > using a servlet was not possible at all
> > > > > > > > > > > >
> > > > > > > > > > > > Best Regards
> > > > > > > > > > > >
> > > > > > > > > > > > Korbinian
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > > > > > > Von: Johan Compagner
> > [mailto:jcompagner@gmail.com]
> > > > > > > > > > > > > Gesendet: Samstag, 11. November 2006 17:30
> > > > > > > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > > > > > > >
> > > > > > > > > > > > > does it always just work if always your
> > > > first access
> > > > > > > > > > > > > is
> > > > > > > > > > > the root of
> > > > > > > > > > > > > your webapp?
> > > > > > > > > > > > > So if you redeploy or restart then your
> > > > first request
> > > > > > > > > > > should be the
> > > > > > > > > > > > > root of your webapp.
> > > > > > > > > > > > > Then afterwards should everything just work.
> > > > > > > > > > > > >
> > > > > > > > > > > > > This is a bug yes i have to rollback code for
> > > > > > this because
> > > > > > > > > > > > > getServletPath() can't be depended on
> > in a filter
> > > > > > > > > > configuration.
> > > > > > > > > > > > > Which we do now.
> > > > > > > > > > > > >
> > > > > > > > > > > > > johan
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > On 11/11/06, Korbinian Bachl
> > > > > > > > <ko...@whiskyworld.de> wrote:
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Hi,
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > I issued a severe BUG
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > http://issues.apache.org/jira/browse/WICKET-42
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Best Regards,
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Korbinian
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > >
> > >
> >
>
>
AW: Severe Bug in Wicket 2 / WICKET-42
Posted by Korbinian Bachl <ko...@whiskyworld.de>.
Hi Johan,
yes i know its not component, and its evil :O
the *.html would mean a unique identifier where a filter could be attached
to and so static resources wouldnt be touched.
The one class - well, i know its a hard discussion, but as we also have
wicket-apps where whole app is just 1 page that swaps around components in
it this doenst sound too exotic for me.
regards
> -----Ursprüngliche Nachricht-----
> Von: Johan Compagner [mailto:jcompagner@gmail.com]
> Gesendet: Montag, 13. November 2006 20:29
> An: wicket-dev@incubator.apache.org
> Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
>
> you can have .html just fine.
> Only you want to have xxx/param.html?
> so like:
>
> /product/110203.html?
> or
> /product/id/110203.html?
>
> Why that html there?
> What does that bring?
> I don't find it nicer that is just taste.
> And i guess you could build it just fine. Just make another
> url encoder.
>
> And having one class handling everything.. a really really
> big YUK or ARRGGH
>
> thats horrible. We are in a component oriented framework
> here. Where small pieces do there job Not one big page where
> you parse the page parameters.
>
> You could do this by the way just fine now.
> Only have have to mount those pages not on / (thats homepage)
> but on /mybigentry
>
>
> johan
>
>
>
> On 11/13/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> >
> > Yeah similar to that. I mean we currently even cant employ
> file-endings.
> >
> > e.g: you can have /page/param/param
> > but not /page/param/param.html wich would be even more nice
> and logic
> > (IMHO)
> >
> > imagine a mount where you could specify "/" as begin and ".htm" as
> > end. -> you could have 1 handling-class for all data as well as
> > natural URLs (if you use IndexedURLParams).
> > if / is solely you could return the start page, if /.htm is called
> > wicket knows that it would be the handling-class with empty
> > params...so we also wouldnt crash with any other wicket
> functionality.
> >
> > e.g: globalmount(".htm", new
> > IndexedUrlParamStrategy("/",handler.class));
> > so you also would centralise URLs to just 1 class in app
> wich then can
> > delegate to the resulting pages.
> >
> > was just a thought from what im in now as i need nice URLs and the
> > problems i stumbled in regarding static content, root etc.
> >
> > what do you think about this? - i'm interested in any
> thoughts about it.
> >
> > Best Regards,
> >
> > Korbinian
> >
> >
> >
> >
> >
> > > -----Ursprüngliche Nachricht-----
> > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > Gesendet: Montag, 13. November 2006 10:20
> > > An: wicket-dev@incubator.apache.org
> > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > >
> > > So this way?
> > >
> > > / == home page
> > > /?p=a == homepage with page params.
> > > /?interface:xxx == an callback to an existing page
> > > /?bookmarkablePage=xxx == an none mounted bookmarkable
> page request
> > > /input == mounted bookmarkable page
> > >
> > > and then
> > >
> > > /a == this is not mounted like input above.. Then it
> should go to a
> > > page that was mounted on /* ??
> > > So you want to have a fallback to a page?
> > >
> > > johan
> > >
> > >
> > >
> > > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > >
> > > > Yeah, that is what i mean -
> > > >
> > > > e.g: mount to /*.html -> would mean it goes for
> /[everything].html
> > > > wich would introduce even more flexibility...
> > > >
> > > > currently, it goes to "/[neededCharacter]" and this catches
> > > > /[neededCharacter][anythingBehind]
> > > > e.g: /a ->
> > > > /a/b is valid
> > > > /aa/b is valid
> > > > /a is valid
> > > >
> > > > as / is mounted for the homepage and you cant mount a
> > > wildcard so far,
> > > > you still need a fixed point you can hook on.
> > > >
> > > > Regards
> > > >
> > > >
> > > > > -----Ursprüngliche Nachricht-----
> > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > Gesendet: Sonntag, 12. November 2006 20:15
> > > > > An: wicket-dev@incubator.apache.org
> > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > >
> > > > > what does a page mounted on /* means in your eyes?
> > > > > /[everything]??
> > > > >
> > > > > there is no such a thing like wild card mounting.
> > > > >
> > > > > johan
> > > > >
> > > > >
> > > > > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > > >
> > > > > > Hi Igor,
> > > > > >
> > > > > > yes - but its not mounted to "/*" but to "/" - is it
> > > > > possible to have
> > > > > > the homepage mounted on "/*" ? means that this page gets
> > > > > all path as
> > > > > > indexedURLparams??
> > > > > >
> > > > > > if i mount it to /* he doesnt see it as wildcard but as
> > > fixed "*"
> > > > > >
> > > > > > Regards
> > > > > >
> > > > > >
> > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > Von: Igor Vaynberg [mailto:igor.vaynberg@gmail.com]
> > > > > > > Gesendet: Sonntag, 12. November 2006 18:39
> > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > >
> > > > > > > the homepage is always mounted on /
> > > > > > >
> > > > > > > -igor
> > > > > > >
> > > > > > >
> > > > > > > On 11/12/06, Korbinian Bachl
> <ko...@whiskyworld.de> wrote:
> > > > > > > >
> > > > > > > > Johan,
> > > > > > > >
> > > > > > > > i checked all - everything is now behaving as expected.
> > > > > Great Work!
> > > > > > > >
> > > > > > > > One question however is still to me - why cant i mount a
> > > > > > > page to "/" ?
> > > > > > > >
> > > > > > > > so i could mount a PageDistribution class there
> which then
> > > > > > > handles the
> > > > > > > > PageParams logic... or would this get in conflict with
> > > > > the rest
> > > > > > > > of wicket ?
> > > > > > > > if so, would it be possible to have a page
> responsible for
> > > > > > > every path
> > > > > > > > behind / but for /wicket/ where the internal could
> > > be handled ??
> > > > > > > >
> > > > > > > > i know this aproach is a bit away from what wicket was
> > > > > made for...
> > > > > > > >
> > > > > > > > Rgards
> > > > > > > >
> > > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > > Von: Korbinian Bachl [mailto:korbinian@whiskyworld.de]
> > > > > > > > > Gesendet: Sonntag, 12. November 2006 10:32
> > > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > > Betreff: AW: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > > >
> > > > > > > > > Really fast :)
> > > > > > > > >
> > > > > > > > > i check it out soon and report,
> > > > > > > > >
> > > > > > > > > Regards
> > > > > > > > >
> > > > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > > > > > Gesendet: Samstag, 11. November 2006 23:17
> > > > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > > > >
> > > > > > > > > > should be fixed now.
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > On 11/11/06, Korbinian Bachl
> > > > > <ko...@whiskyworld.de> wrote:
> > > > > > > > > > >
> > > > > > > > > > > well, it works if it is:
> > > > > > > > > > > a, the first access of the WebApp with no provious
> > > > > > > > > ReDeploy b, the
> > > > > > > > > > > root is accessed then
> > > > > > > > > > >
> > > > > > > > > > > if i redeploy just once, i cant access the
> > > root anymore
> > > > > > > > > > > -
> > > > > > > > > > however the
> > > > > > > > > > > direct path always works...
> > > > > > > > > > > if i redeploy it and access the path before, it
> > > > > > > > > > > chains
> > > > > > > > > > >
> > > > > > > > > > > another danger thing is that following works:
> > > > > > > > > > > assume you have IndexedPageParams and index.class
> > > > > > > > > > > mounted
> > > > > > > > > to /Index:
> > > > > > > > > > >
> > > > > > > > > > > if you then access it via
> /Index/foo/bar/Index/foo2
> > > > > > > > > > >
> > > > > > > > > > > and print out pageparams.toString you only
> > > get 0=foo2 ,
> > > > > > > > > meaning he
> > > > > > > > > > > access it at last point, and not at first!
> > > > > > > > > > > This violates JEE security !!!
> > > > > > > > > > >
> > > > > > > > > > > i also noticed that if you use a servlet and
> > > > > > > > > > BookMarkablePageLinks the
> > > > > > > > > > > URLs look like these
> > > > > 127.0.0.1/<servlet>?param=value - and
> > > > > > > > > > mounting
> > > > > > > > > > > using a servlet was not possible at all
> > > > > > > > > > >
> > > > > > > > > > > Best Regards
> > > > > > > > > > >
> > > > > > > > > > > Korbinian
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > > > > > Von: Johan Compagner
> [mailto:jcompagner@gmail.com]
> > > > > > > > > > > > Gesendet: Samstag, 11. November 2006 17:30
> > > > > > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > > > > > >
> > > > > > > > > > > > does it always just work if always your
> > > first access
> > > > > > > > > > > > is
> > > > > > > > > > the root of
> > > > > > > > > > > > your webapp?
> > > > > > > > > > > > So if you redeploy or restart then your
> > > first request
> > > > > > > > > > should be the
> > > > > > > > > > > > root of your webapp.
> > > > > > > > > > > > Then afterwards should everything just work.
> > > > > > > > > > > >
> > > > > > > > > > > > This is a bug yes i have to rollback code for
> > > > > this because
> > > > > > > > > > > > getServletPath() can't be depended on
> in a filter
> > > > > > > > > configuration.
> > > > > > > > > > > > Which we do now.
> > > > > > > > > > > >
> > > > > > > > > > > > johan
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > On 11/11/06, Korbinian Bachl
> > > > > > > <ko...@whiskyworld.de> wrote:
> > > > > > > > > > > > >
> > > > > > > > > > > > > Hi,
> > > > > > > > > > > > >
> > > > > > > > > > > > > I issued a severe BUG
> > > > > > > > > > > > >
> > > > > > > > > > > > > http://issues.apache.org/jira/browse/WICKET-42
> > > > > > > > > > > > >
> > > > > > > > > > > > > Best Regards,
> > > > > > > > > > > > >
> > > > > > > > > > > > > Korbinian
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > >
> > > >
> > >
> >
> >
>
Re: Severe Bug in Wicket 2 / WICKET-42
Posted by Johan Compagner <jc...@gmail.com>.
you can have .html just fine.
Only you want to have xxx/param.html?
so like:
/product/110203.html?
or
/product/id/110203.html?
Why that html there?
What does that bring?
I don't find it nicer that is just taste.
And i guess you could build it just fine. Just make another url encoder.
And having one class handling everything.. a really really big YUK or ARRGGH
thats horrible. We are in a component oriented framework here. Where small
pieces do there job
Not one big page where you parse the page parameters.
You could do this by the way just fine now.
Only have have to mount those pages not on / (thats homepage) but on
/mybigentry
johan
On 11/13/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
>
> Yeah similar to that. I mean we currently even cant employ file-endings.
>
> e.g: you can have /page/param/param
> but not /page/param/param.html wich would be even more nice and logic
> (IMHO)
>
> imagine a mount where you could specify "/" as begin and ".htm" as end. ->
> you could have 1 handling-class for all data as well as natural URLs (if
> you
> use IndexedURLParams).
> if / is solely you could return the start page, if /.htm is called wicket
> knows that it would be the handling-class with empty params...so we also
> wouldnt crash with any other wicket functionality.
>
> e.g: globalmount(".htm", new IndexedUrlParamStrategy("/",handler.class));
> so you also would centralise URLs to just 1 class in app wich then can
> delegate to the resulting pages.
>
> was just a thought from what im in now as i need nice URLs and the
> problems
> i stumbled in regarding static content, root etc.
>
> what do you think about this? - i'm interested in any thoughts about it.
>
> Best Regards,
>
> Korbinian
>
>
>
>
>
> > -----Ursprüngliche Nachricht-----
> > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > Gesendet: Montag, 13. November 2006 10:20
> > An: wicket-dev@incubator.apache.org
> > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> >
> > So this way?
> >
> > / == home page
> > /?p=a == homepage with page params.
> > /?interface:xxx == an callback to an existing page
> > /?bookmarkablePage=xxx == an none mounted bookmarkable page request
> > /input == mounted bookmarkable page
> >
> > and then
> >
> > /a == this is not mounted like input above.. Then it should
> > go to a page that was mounted on /* ??
> > So you want to have a fallback to a page?
> >
> > johan
> >
> >
> >
> > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > >
> > > Yeah, that is what i mean -
> > >
> > > e.g: mount to /*.html -> would mean it goes for /[everything].html
> > > wich would introduce even more flexibility...
> > >
> > > currently, it goes to "/[neededCharacter]" and this catches
> > > /[neededCharacter][anythingBehind]
> > > e.g: /a ->
> > > /a/b is valid
> > > /aa/b is valid
> > > /a is valid
> > >
> > > as / is mounted for the homepage and you cant mount a
> > wildcard so far,
> > > you still need a fixed point you can hook on.
> > >
> > > Regards
> > >
> > >
> > > > -----Ursprüngliche Nachricht-----
> > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > Gesendet: Sonntag, 12. November 2006 20:15
> > > > An: wicket-dev@incubator.apache.org
> > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > >
> > > > what does a page mounted on /* means in your eyes?
> > > > /[everything]??
> > > >
> > > > there is no such a thing like wild card mounting.
> > > >
> > > > johan
> > > >
> > > >
> > > > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > >
> > > > > Hi Igor,
> > > > >
> > > > > yes - but its not mounted to "/*" but to "/" - is it
> > > > possible to have
> > > > > the homepage mounted on "/*" ? means that this page gets
> > > > all path as
> > > > > indexedURLparams??
> > > > >
> > > > > if i mount it to /* he doesnt see it as wildcard but as
> > fixed "*"
> > > > >
> > > > > Regards
> > > > >
> > > > >
> > > > > > -----Ursprüngliche Nachricht-----
> > > > > > Von: Igor Vaynberg [mailto:igor.vaynberg@gmail.com]
> > > > > > Gesendet: Sonntag, 12. November 2006 18:39
> > > > > > An: wicket-dev@incubator.apache.org
> > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > >
> > > > > > the homepage is always mounted on /
> > > > > >
> > > > > > -igor
> > > > > >
> > > > > >
> > > > > > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > > > >
> > > > > > > Johan,
> > > > > > >
> > > > > > > i checked all - everything is now behaving as expected.
> > > > Great Work!
> > > > > > >
> > > > > > > One question however is still to me - why cant i mount a
> > > > > > page to "/" ?
> > > > > > >
> > > > > > > so i could mount a PageDistribution class there which then
> > > > > > handles the
> > > > > > > PageParams logic... or would this get in conflict with
> > > > the rest
> > > > > > > of wicket ?
> > > > > > > if so, would it be possible to have a page responsible for
> > > > > > every path
> > > > > > > behind / but for /wicket/ where the internal could
> > be handled ??
> > > > > > >
> > > > > > > i know this aproach is a bit away from what wicket was
> > > > made for...
> > > > > > >
> > > > > > > Rgards
> > > > > > >
> > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > Von: Korbinian Bachl [mailto:korbinian@whiskyworld.de]
> > > > > > > > Gesendet: Sonntag, 12. November 2006 10:32
> > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > Betreff: AW: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > >
> > > > > > > > Really fast :)
> > > > > > > >
> > > > > > > > i check it out soon and report,
> > > > > > > >
> > > > > > > > Regards
> > > > > > > >
> > > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > > > > Gesendet: Samstag, 11. November 2006 23:17
> > > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > > >
> > > > > > > > > should be fixed now.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > On 11/11/06, Korbinian Bachl
> > > > <ko...@whiskyworld.de> wrote:
> > > > > > > > > >
> > > > > > > > > > well, it works if it is:
> > > > > > > > > > a, the first access of the WebApp with no provious
> > > > > > > > ReDeploy b, the
> > > > > > > > > > root is accessed then
> > > > > > > > > >
> > > > > > > > > > if i redeploy just once, i cant access the
> > root anymore
> > > > > > > > > > -
> > > > > > > > > however the
> > > > > > > > > > direct path always works...
> > > > > > > > > > if i redeploy it and access the path before, it chains
> > > > > > > > > >
> > > > > > > > > > another danger thing is that following works:
> > > > > > > > > > assume you have IndexedPageParams and index.class
> > > > > > > > > > mounted
> > > > > > > > to /Index:
> > > > > > > > > >
> > > > > > > > > > if you then access it via /Index/foo/bar/Index/foo2
> > > > > > > > > >
> > > > > > > > > > and print out pageparams.toString you only
> > get 0=foo2 ,
> > > > > > > > meaning he
> > > > > > > > > > access it at last point, and not at first!
> > > > > > > > > > This violates JEE security !!!
> > > > > > > > > >
> > > > > > > > > > i also noticed that if you use a servlet and
> > > > > > > > > BookMarkablePageLinks the
> > > > > > > > > > URLs look like these
> > > > 127.0.0.1/<servlet>?param=value - and
> > > > > > > > > mounting
> > > > > > > > > > using a servlet was not possible at all
> > > > > > > > > >
> > > > > > > > > > Best Regards
> > > > > > > > > >
> > > > > > > > > > Korbinian
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > > > > > > Gesendet: Samstag, 11. November 2006 17:30
> > > > > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > > > > >
> > > > > > > > > > > does it always just work if always your
> > first access
> > > > > > > > > > > is
> > > > > > > > > the root of
> > > > > > > > > > > your webapp?
> > > > > > > > > > > So if you redeploy or restart then your
> > first request
> > > > > > > > > should be the
> > > > > > > > > > > root of your webapp.
> > > > > > > > > > > Then afterwards should everything just work.
> > > > > > > > > > >
> > > > > > > > > > > This is a bug yes i have to rollback code for
> > > > this because
> > > > > > > > > > > getServletPath() can't be depended on in a filter
> > > > > > > > configuration.
> > > > > > > > > > > Which we do now.
> > > > > > > > > > >
> > > > > > > > > > > johan
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > On 11/11/06, Korbinian Bachl
> > > > > > <ko...@whiskyworld.de> wrote:
> > > > > > > > > > > >
> > > > > > > > > > > > Hi,
> > > > > > > > > > > >
> > > > > > > > > > > > I issued a severe BUG
> > > > > > > > > > > >
> > > > > > > > > > > > http://issues.apache.org/jira/browse/WICKET-42
> > > > > > > > > > > >
> > > > > > > > > > > > Best Regards,
> > > > > > > > > > > >
> > > > > > > > > > > > Korbinian
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > >
> > >
> >
>
>
AW: Severe Bug in Wicket 2 / WICKET-42
Posted by Korbinian Bachl <ko...@whiskyworld.de>.
Yeah similar to that. I mean we currently even cant employ file-endings.
e.g: you can have /page/param/param
but not /page/param/param.html wich would be even more nice and logic (IMHO)
imagine a mount where you could specify "/" as begin and ".htm" as end. ->
you could have 1 handling-class for all data as well as natural URLs (if you
use IndexedURLParams).
if / is solely you could return the start page, if /.htm is called wicket
knows that it would be the handling-class with empty params...so we also
wouldnt crash with any other wicket functionality.
e.g: globalmount(".htm", new IndexedUrlParamStrategy("/",handler.class));
so you also would centralise URLs to just 1 class in app wich then can
delegate to the resulting pages.
was just a thought from what im in now as i need nice URLs and the problems
i stumbled in regarding static content, root etc.
what do you think about this? - i'm interested in any thoughts about it.
Best Regards,
Korbinian
> -----Ursprüngliche Nachricht-----
> Von: Johan Compagner [mailto:jcompagner@gmail.com]
> Gesendet: Montag, 13. November 2006 10:20
> An: wicket-dev@incubator.apache.org
> Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
>
> So this way?
>
> / == home page
> /?p=a == homepage with page params.
> /?interface:xxx == an callback to an existing page
> /?bookmarkablePage=xxx == an none mounted bookmarkable page request
> /input == mounted bookmarkable page
>
> and then
>
> /a == this is not mounted like input above.. Then it should
> go to a page that was mounted on /* ??
> So you want to have a fallback to a page?
>
> johan
>
>
>
> On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> >
> > Yeah, that is what i mean -
> >
> > e.g: mount to /*.html -> would mean it goes for /[everything].html
> > wich would introduce even more flexibility...
> >
> > currently, it goes to "/[neededCharacter]" and this catches
> > /[neededCharacter][anythingBehind]
> > e.g: /a ->
> > /a/b is valid
> > /aa/b is valid
> > /a is valid
> >
> > as / is mounted for the homepage and you cant mount a
> wildcard so far,
> > you still need a fixed point you can hook on.
> >
> > Regards
> >
> >
> > > -----Ursprüngliche Nachricht-----
> > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > Gesendet: Sonntag, 12. November 2006 20:15
> > > An: wicket-dev@incubator.apache.org
> > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > >
> > > what does a page mounted on /* means in your eyes?
> > > /[everything]??
> > >
> > > there is no such a thing like wild card mounting.
> > >
> > > johan
> > >
> > >
> > > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > >
> > > > Hi Igor,
> > > >
> > > > yes - but its not mounted to "/*" but to "/" - is it
> > > possible to have
> > > > the homepage mounted on "/*" ? means that this page gets
> > > all path as
> > > > indexedURLparams??
> > > >
> > > > if i mount it to /* he doesnt see it as wildcard but as
> fixed "*"
> > > >
> > > > Regards
> > > >
> > > >
> > > > > -----Ursprüngliche Nachricht-----
> > > > > Von: Igor Vaynberg [mailto:igor.vaynberg@gmail.com]
> > > > > Gesendet: Sonntag, 12. November 2006 18:39
> > > > > An: wicket-dev@incubator.apache.org
> > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > >
> > > > > the homepage is always mounted on /
> > > > >
> > > > > -igor
> > > > >
> > > > >
> > > > > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > > >
> > > > > > Johan,
> > > > > >
> > > > > > i checked all - everything is now behaving as expected.
> > > Great Work!
> > > > > >
> > > > > > One question however is still to me - why cant i mount a
> > > > > page to "/" ?
> > > > > >
> > > > > > so i could mount a PageDistribution class there which then
> > > > > handles the
> > > > > > PageParams logic... or would this get in conflict with
> > > the rest
> > > > > > of wicket ?
> > > > > > if so, would it be possible to have a page responsible for
> > > > > every path
> > > > > > behind / but for /wicket/ where the internal could
> be handled ??
> > > > > >
> > > > > > i know this aproach is a bit away from what wicket was
> > > made for...
> > > > > >
> > > > > > Rgards
> > > > > >
> > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > Von: Korbinian Bachl [mailto:korbinian@whiskyworld.de]
> > > > > > > Gesendet: Sonntag, 12. November 2006 10:32
> > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > Betreff: AW: Severe Bug in Wicket 2 / WICKET-42
> > > > > > >
> > > > > > > Really fast :)
> > > > > > >
> > > > > > > i check it out soon and report,
> > > > > > >
> > > > > > > Regards
> > > > > > >
> > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > > > Gesendet: Samstag, 11. November 2006 23:17
> > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > >
> > > > > > > > should be fixed now.
> > > > > > > >
> > > > > > > >
> > > > > > > > On 11/11/06, Korbinian Bachl
> > > <ko...@whiskyworld.de> wrote:
> > > > > > > > >
> > > > > > > > > well, it works if it is:
> > > > > > > > > a, the first access of the WebApp with no provious
> > > > > > > ReDeploy b, the
> > > > > > > > > root is accessed then
> > > > > > > > >
> > > > > > > > > if i redeploy just once, i cant access the
> root anymore
> > > > > > > > > -
> > > > > > > > however the
> > > > > > > > > direct path always works...
> > > > > > > > > if i redeploy it and access the path before, it chains
> > > > > > > > >
> > > > > > > > > another danger thing is that following works:
> > > > > > > > > assume you have IndexedPageParams and index.class
> > > > > > > > > mounted
> > > > > > > to /Index:
> > > > > > > > >
> > > > > > > > > if you then access it via /Index/foo/bar/Index/foo2
> > > > > > > > >
> > > > > > > > > and print out pageparams.toString you only
> get 0=foo2 ,
> > > > > > > meaning he
> > > > > > > > > access it at last point, and not at first!
> > > > > > > > > This violates JEE security !!!
> > > > > > > > >
> > > > > > > > > i also noticed that if you use a servlet and
> > > > > > > > BookMarkablePageLinks the
> > > > > > > > > URLs look like these
> > > 127.0.0.1/<servlet>?param=value - and
> > > > > > > > mounting
> > > > > > > > > using a servlet was not possible at all
> > > > > > > > >
> > > > > > > > > Best Regards
> > > > > > > > >
> > > > > > > > > Korbinian
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > > > > > Gesendet: Samstag, 11. November 2006 17:30
> > > > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > > > >
> > > > > > > > > > does it always just work if always your
> first access
> > > > > > > > > > is
> > > > > > > > the root of
> > > > > > > > > > your webapp?
> > > > > > > > > > So if you redeploy or restart then your
> first request
> > > > > > > > should be the
> > > > > > > > > > root of your webapp.
> > > > > > > > > > Then afterwards should everything just work.
> > > > > > > > > >
> > > > > > > > > > This is a bug yes i have to rollback code for
> > > this because
> > > > > > > > > > getServletPath() can't be depended on in a filter
> > > > > > > configuration.
> > > > > > > > > > Which we do now.
> > > > > > > > > >
> > > > > > > > > > johan
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > On 11/11/06, Korbinian Bachl
> > > > > <ko...@whiskyworld.de> wrote:
> > > > > > > > > > >
> > > > > > > > > > > Hi,
> > > > > > > > > > >
> > > > > > > > > > > I issued a severe BUG
> > > > > > > > > > >
> > > > > > > > > > > http://issues.apache.org/jira/browse/WICKET-42
> > > > > > > > > > >
> > > > > > > > > > > Best Regards,
> > > > > > > > > > >
> > > > > > > > > > > Korbinian
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > >
> > > >
> > >
> >
> >
>
Re: Severe Bug in Wicket 2 / WICKET-42
Posted by Johan Compagner <jc...@gmail.com>.
So this way?
/ == home page
/?p=a == homepage with page params.
/?interface:xxx == an callback to an existing page
/?bookmarkablePage=xxx == an none mounted bookmarkable page request
/input == mounted bookmarkable page
and then
/a == this is not mounted like input above.. Then it should go to a page
that was mounted on /* ??
So you want to have a fallback to a page?
johan
On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
>
> Yeah, that is what i mean -
>
> e.g: mount to /*.html -> would mean it goes for /[everything].html wich
> would introduce even more flexibility...
>
> currently, it goes to "/[neededCharacter]" and this catches
> /[neededCharacter][anythingBehind]
> e.g: /a ->
> /a/b is valid
> /aa/b is valid
> /a is valid
>
> as / is mounted for the homepage and you cant mount a wildcard so far, you
> still need a fixed point you can hook on.
>
> Regards
>
>
> > -----Ursprüngliche Nachricht-----
> > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > Gesendet: Sonntag, 12. November 2006 20:15
> > An: wicket-dev@incubator.apache.org
> > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> >
> > what does a page mounted on /* means in your eyes?
> > /[everything]??
> >
> > there is no such a thing like wild card mounting.
> >
> > johan
> >
> >
> > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > >
> > > Hi Igor,
> > >
> > > yes - but its not mounted to "/*" but to "/" - is it
> > possible to have
> > > the homepage mounted on "/*" ? means that this page gets
> > all path as
> > > indexedURLparams??
> > >
> > > if i mount it to /* he doesnt see it as wildcard but as fixed "*"
> > >
> > > Regards
> > >
> > >
> > > > -----Ursprüngliche Nachricht-----
> > > > Von: Igor Vaynberg [mailto:igor.vaynberg@gmail.com]
> > > > Gesendet: Sonntag, 12. November 2006 18:39
> > > > An: wicket-dev@incubator.apache.org
> > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > >
> > > > the homepage is always mounted on /
> > > >
> > > > -igor
> > > >
> > > >
> > > > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > >
> > > > > Johan,
> > > > >
> > > > > i checked all - everything is now behaving as expected.
> > Great Work!
> > > > >
> > > > > One question however is still to me - why cant i mount a
> > > > page to "/" ?
> > > > >
> > > > > so i could mount a PageDistribution class there which then
> > > > handles the
> > > > > PageParams logic... or would this get in conflict with
> > the rest
> > > > > of wicket ?
> > > > > if so, would it be possible to have a page responsible for
> > > > every path
> > > > > behind / but for /wicket/ where the internal could be handled ??
> > > > >
> > > > > i know this aproach is a bit away from what wicket was
> > made for...
> > > > >
> > > > > Rgards
> > > > >
> > > > > > -----Ursprüngliche Nachricht-----
> > > > > > Von: Korbinian Bachl [mailto:korbinian@whiskyworld.de]
> > > > > > Gesendet: Sonntag, 12. November 2006 10:32
> > > > > > An: wicket-dev@incubator.apache.org
> > > > > > Betreff: AW: Severe Bug in Wicket 2 / WICKET-42
> > > > > >
> > > > > > Really fast :)
> > > > > >
> > > > > > i check it out soon and report,
> > > > > >
> > > > > > Regards
> > > > > >
> > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > > Gesendet: Samstag, 11. November 2006 23:17
> > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > >
> > > > > > > should be fixed now.
> > > > > > >
> > > > > > >
> > > > > > > On 11/11/06, Korbinian Bachl
> > <ko...@whiskyworld.de> wrote:
> > > > > > > >
> > > > > > > > well, it works if it is:
> > > > > > > > a, the first access of the WebApp with no provious
> > > > > > ReDeploy b, the
> > > > > > > > root is accessed then
> > > > > > > >
> > > > > > > > if i redeploy just once, i cant access the root anymore -
> > > > > > > however the
> > > > > > > > direct path always works...
> > > > > > > > if i redeploy it and access the path before, it chains
> > > > > > > >
> > > > > > > > another danger thing is that following works:
> > > > > > > > assume you have IndexedPageParams and index.class mounted
> > > > > > to /Index:
> > > > > > > >
> > > > > > > > if you then access it via /Index/foo/bar/Index/foo2
> > > > > > > >
> > > > > > > > and print out pageparams.toString you only get 0=foo2 ,
> > > > > > meaning he
> > > > > > > > access it at last point, and not at first!
> > > > > > > > This violates JEE security !!!
> > > > > > > >
> > > > > > > > i also noticed that if you use a servlet and
> > > > > > > BookMarkablePageLinks the
> > > > > > > > URLs look like these
> > 127.0.0.1/<servlet>?param=value - and
> > > > > > > mounting
> > > > > > > > using a servlet was not possible at all
> > > > > > > >
> > > > > > > > Best Regards
> > > > > > > >
> > > > > > > > Korbinian
> > > > > > > >
> > > > > > > >
> > > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > > > > Gesendet: Samstag, 11. November 2006 17:30
> > > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > > >
> > > > > > > > > does it always just work if always your first access is
> > > > > > > the root of
> > > > > > > > > your webapp?
> > > > > > > > > So if you redeploy or restart then your first request
> > > > > > > should be the
> > > > > > > > > root of your webapp.
> > > > > > > > > Then afterwards should everything just work.
> > > > > > > > >
> > > > > > > > > This is a bug yes i have to rollback code for
> > this because
> > > > > > > > > getServletPath() can't be depended on in a filter
> > > > > > configuration.
> > > > > > > > > Which we do now.
> > > > > > > > >
> > > > > > > > > johan
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > On 11/11/06, Korbinian Bachl
> > > > <ko...@whiskyworld.de> wrote:
> > > > > > > > > >
> > > > > > > > > > Hi,
> > > > > > > > > >
> > > > > > > > > > I issued a severe BUG
> > > > > > > > > >
> > > > > > > > > > http://issues.apache.org/jira/browse/WICKET-42
> > > > > > > > > >
> > > > > > > > > > Best Regards,
> > > > > > > > > >
> > > > > > > > > > Korbinian
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > >
> > >
> >
>
>
AW: Severe Bug in Wicket 2 / WICKET-42
Posted by Korbinian Bachl <ko...@whiskyworld.de>.
Yeah, that is what i mean -
e.g: mount to /*.html -> would mean it goes for /[everything].html wich
would introduce even more flexibility...
currently, it goes to "/[neededCharacter]" and this catches
/[neededCharacter][anythingBehind]
e.g: /a ->
/a/b is valid
/aa/b is valid
/a is valid
as / is mounted for the homepage and you cant mount a wildcard so far, you
still need a fixed point you can hook on.
Regards
> -----Ursprüngliche Nachricht-----
> Von: Johan Compagner [mailto:jcompagner@gmail.com]
> Gesendet: Sonntag, 12. November 2006 20:15
> An: wicket-dev@incubator.apache.org
> Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
>
> what does a page mounted on /* means in your eyes?
> /[everything]??
>
> there is no such a thing like wild card mounting.
>
> johan
>
>
> On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> >
> > Hi Igor,
> >
> > yes - but its not mounted to "/*" but to "/" - is it
> possible to have
> > the homepage mounted on "/*" ? means that this page gets
> all path as
> > indexedURLparams??
> >
> > if i mount it to /* he doesnt see it as wildcard but as fixed "*"
> >
> > Regards
> >
> >
> > > -----Ursprüngliche Nachricht-----
> > > Von: Igor Vaynberg [mailto:igor.vaynberg@gmail.com]
> > > Gesendet: Sonntag, 12. November 2006 18:39
> > > An: wicket-dev@incubator.apache.org
> > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > >
> > > the homepage is always mounted on /
> > >
> > > -igor
> > >
> > >
> > > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > >
> > > > Johan,
> > > >
> > > > i checked all - everything is now behaving as expected.
> Great Work!
> > > >
> > > > One question however is still to me - why cant i mount a
> > > page to "/" ?
> > > >
> > > > so i could mount a PageDistribution class there which then
> > > handles the
> > > > PageParams logic... or would this get in conflict with
> the rest
> > > > of wicket ?
> > > > if so, would it be possible to have a page responsible for
> > > every path
> > > > behind / but for /wicket/ where the internal could be handled ??
> > > >
> > > > i know this aproach is a bit away from what wicket was
> made for...
> > > >
> > > > Rgards
> > > >
> > > > > -----Ursprüngliche Nachricht-----
> > > > > Von: Korbinian Bachl [mailto:korbinian@whiskyworld.de]
> > > > > Gesendet: Sonntag, 12. November 2006 10:32
> > > > > An: wicket-dev@incubator.apache.org
> > > > > Betreff: AW: Severe Bug in Wicket 2 / WICKET-42
> > > > >
> > > > > Really fast :)
> > > > >
> > > > > i check it out soon and report,
> > > > >
> > > > > Regards
> > > > >
> > > > > > -----Ursprüngliche Nachricht-----
> > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > Gesendet: Samstag, 11. November 2006 23:17
> > > > > > An: wicket-dev@incubator.apache.org
> > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > >
> > > > > > should be fixed now.
> > > > > >
> > > > > >
> > > > > > On 11/11/06, Korbinian Bachl
> <ko...@whiskyworld.de> wrote:
> > > > > > >
> > > > > > > well, it works if it is:
> > > > > > > a, the first access of the WebApp with no provious
> > > > > ReDeploy b, the
> > > > > > > root is accessed then
> > > > > > >
> > > > > > > if i redeploy just once, i cant access the root anymore -
> > > > > > however the
> > > > > > > direct path always works...
> > > > > > > if i redeploy it and access the path before, it chains
> > > > > > >
> > > > > > > another danger thing is that following works:
> > > > > > > assume you have IndexedPageParams and index.class mounted
> > > > > to /Index:
> > > > > > >
> > > > > > > if you then access it via /Index/foo/bar/Index/foo2
> > > > > > >
> > > > > > > and print out pageparams.toString you only get 0=foo2 ,
> > > > > meaning he
> > > > > > > access it at last point, and not at first!
> > > > > > > This violates JEE security !!!
> > > > > > >
> > > > > > > i also noticed that if you use a servlet and
> > > > > > BookMarkablePageLinks the
> > > > > > > URLs look like these
> 127.0.0.1/<servlet>?param=value - and
> > > > > > mounting
> > > > > > > using a servlet was not possible at all
> > > > > > >
> > > > > > > Best Regards
> > > > > > >
> > > > > > > Korbinian
> > > > > > >
> > > > > > >
> > > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > > > Gesendet: Samstag, 11. November 2006 17:30
> > > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > > >
> > > > > > > > does it always just work if always your first access is
> > > > > > the root of
> > > > > > > > your webapp?
> > > > > > > > So if you redeploy or restart then your first request
> > > > > > should be the
> > > > > > > > root of your webapp.
> > > > > > > > Then afterwards should everything just work.
> > > > > > > >
> > > > > > > > This is a bug yes i have to rollback code for
> this because
> > > > > > > > getServletPath() can't be depended on in a filter
> > > > > configuration.
> > > > > > > > Which we do now.
> > > > > > > >
> > > > > > > > johan
> > > > > > > >
> > > > > > > >
> > > > > > > > On 11/11/06, Korbinian Bachl
> > > <ko...@whiskyworld.de> wrote:
> > > > > > > > >
> > > > > > > > > Hi,
> > > > > > > > >
> > > > > > > > > I issued a severe BUG
> > > > > > > > >
> > > > > > > > > http://issues.apache.org/jira/browse/WICKET-42
> > > > > > > > >
> > > > > > > > > Best Regards,
> > > > > > > > >
> > > > > > > > > Korbinian
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > >
> >
> >
>
Re: Severe Bug in Wicket 2 / WICKET-42
Posted by Johan Compagner <jc...@gmail.com>.
what does a page mounted on /* means in your eyes?
/[everything]??
there is no such a thing like wild card mounting.
johan
On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
>
> Hi Igor,
>
> yes - but its not mounted to "/*" but to "/" - is it possible to have the
> homepage mounted on "/*" ? means that this page gets all path as
> indexedURLparams??
>
> if i mount it to /* he doesnt see it as wildcard but as fixed "*"
>
> Regards
>
>
> > -----Ursprüngliche Nachricht-----
> > Von: Igor Vaynberg [mailto:igor.vaynberg@gmail.com]
> > Gesendet: Sonntag, 12. November 2006 18:39
> > An: wicket-dev@incubator.apache.org
> > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> >
> > the homepage is always mounted on /
> >
> > -igor
> >
> >
> > On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > >
> > > Johan,
> > >
> > > i checked all - everything is now behaving as expected. Great Work!
> > >
> > > One question however is still to me - why cant i mount a
> > page to "/" ?
> > >
> > > so i could mount a PageDistribution class there which then
> > handles the
> > > PageParams logic... or would this get in conflict with the rest of
> > > wicket ?
> > > if so, would it be possible to have a page responsible for
> > every path
> > > behind / but for /wicket/ where the internal could be handled ??
> > >
> > > i know this aproach is a bit away from what wicket was made for...
> > >
> > > Rgards
> > >
> > > > -----Ursprüngliche Nachricht-----
> > > > Von: Korbinian Bachl [mailto:korbinian@whiskyworld.de]
> > > > Gesendet: Sonntag, 12. November 2006 10:32
> > > > An: wicket-dev@incubator.apache.org
> > > > Betreff: AW: Severe Bug in Wicket 2 / WICKET-42
> > > >
> > > > Really fast :)
> > > >
> > > > i check it out soon and report,
> > > >
> > > > Regards
> > > >
> > > > > -----Ursprüngliche Nachricht-----
> > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > Gesendet: Samstag, 11. November 2006 23:17
> > > > > An: wicket-dev@incubator.apache.org
> > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > >
> > > > > should be fixed now.
> > > > >
> > > > >
> > > > > On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > > >
> > > > > > well, it works if it is:
> > > > > > a, the first access of the WebApp with no provious
> > > > ReDeploy b, the
> > > > > > root is accessed then
> > > > > >
> > > > > > if i redeploy just once, i cant access the root anymore -
> > > > > however the
> > > > > > direct path always works...
> > > > > > if i redeploy it and access the path before, it chains
> > > > > >
> > > > > > another danger thing is that following works:
> > > > > > assume you have IndexedPageParams and index.class mounted
> > > > to /Index:
> > > > > >
> > > > > > if you then access it via /Index/foo/bar/Index/foo2
> > > > > >
> > > > > > and print out pageparams.toString you only get 0=foo2 ,
> > > > meaning he
> > > > > > access it at last point, and not at first!
> > > > > > This violates JEE security !!!
> > > > > >
> > > > > > i also noticed that if you use a servlet and
> > > > > BookMarkablePageLinks the
> > > > > > URLs look like these 127.0.0.1/<servlet>?param=value - and
> > > > > mounting
> > > > > > using a servlet was not possible at all
> > > > > >
> > > > > > Best Regards
> > > > > >
> > > > > > Korbinian
> > > > > >
> > > > > >
> > > > > > > -----Ursprüngliche Nachricht-----
> > > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > > Gesendet: Samstag, 11. November 2006 17:30
> > > > > > > An: wicket-dev@incubator.apache.org
> > > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > > >
> > > > > > > does it always just work if always your first access is
> > > > > the root of
> > > > > > > your webapp?
> > > > > > > So if you redeploy or restart then your first request
> > > > > should be the
> > > > > > > root of your webapp.
> > > > > > > Then afterwards should everything just work.
> > > > > > >
> > > > > > > This is a bug yes i have to rollback code for this because
> > > > > > > getServletPath() can't be depended on in a filter
> > > > configuration.
> > > > > > > Which we do now.
> > > > > > >
> > > > > > > johan
> > > > > > >
> > > > > > >
> > > > > > > On 11/11/06, Korbinian Bachl
> > <ko...@whiskyworld.de> wrote:
> > > > > > > >
> > > > > > > > Hi,
> > > > > > > >
> > > > > > > > I issued a severe BUG
> > > > > > > >
> > > > > > > > http://issues.apache.org/jira/browse/WICKET-42
> > > > > > > >
> > > > > > > > Best Regards,
> > > > > > > >
> > > > > > > > Korbinian
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
>
>
AW: Severe Bug in Wicket 2 / WICKET-42
Posted by Korbinian Bachl <ko...@whiskyworld.de>.
Hi Igor,
yes - but its not mounted to "/*" but to "/" - is it possible to have the
homepage mounted on "/*" ? means that this page gets all path as
indexedURLparams??
if i mount it to /* he doesnt see it as wildcard but as fixed "*"
Regards
> -----Ursprüngliche Nachricht-----
> Von: Igor Vaynberg [mailto:igor.vaynberg@gmail.com]
> Gesendet: Sonntag, 12. November 2006 18:39
> An: wicket-dev@incubator.apache.org
> Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
>
> the homepage is always mounted on /
>
> -igor
>
>
> On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> >
> > Johan,
> >
> > i checked all - everything is now behaving as expected. Great Work!
> >
> > One question however is still to me - why cant i mount a
> page to "/" ?
> >
> > so i could mount a PageDistribution class there which then
> handles the
> > PageParams logic... or would this get in conflict with the rest of
> > wicket ?
> > if so, would it be possible to have a page responsible for
> every path
> > behind / but for /wicket/ where the internal could be handled ??
> >
> > i know this aproach is a bit away from what wicket was made for...
> >
> > Rgards
> >
> > > -----Ursprüngliche Nachricht-----
> > > Von: Korbinian Bachl [mailto:korbinian@whiskyworld.de]
> > > Gesendet: Sonntag, 12. November 2006 10:32
> > > An: wicket-dev@incubator.apache.org
> > > Betreff: AW: Severe Bug in Wicket 2 / WICKET-42
> > >
> > > Really fast :)
> > >
> > > i check it out soon and report,
> > >
> > > Regards
> > >
> > > > -----Ursprüngliche Nachricht-----
> > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > Gesendet: Samstag, 11. November 2006 23:17
> > > > An: wicket-dev@incubator.apache.org
> > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > >
> > > > should be fixed now.
> > > >
> > > >
> > > > On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > >
> > > > > well, it works if it is:
> > > > > a, the first access of the WebApp with no provious
> > > ReDeploy b, the
> > > > > root is accessed then
> > > > >
> > > > > if i redeploy just once, i cant access the root anymore -
> > > > however the
> > > > > direct path always works...
> > > > > if i redeploy it and access the path before, it chains
> > > > >
> > > > > another danger thing is that following works:
> > > > > assume you have IndexedPageParams and index.class mounted
> > > to /Index:
> > > > >
> > > > > if you then access it via /Index/foo/bar/Index/foo2
> > > > >
> > > > > and print out pageparams.toString you only get 0=foo2 ,
> > > meaning he
> > > > > access it at last point, and not at first!
> > > > > This violates JEE security !!!
> > > > >
> > > > > i also noticed that if you use a servlet and
> > > > BookMarkablePageLinks the
> > > > > URLs look like these 127.0.0.1/<servlet>?param=value - and
> > > > mounting
> > > > > using a servlet was not possible at all
> > > > >
> > > > > Best Regards
> > > > >
> > > > > Korbinian
> > > > >
> > > > >
> > > > > > -----Ursprüngliche Nachricht-----
> > > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > > Gesendet: Samstag, 11. November 2006 17:30
> > > > > > An: wicket-dev@incubator.apache.org
> > > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > > >
> > > > > > does it always just work if always your first access is
> > > > the root of
> > > > > > your webapp?
> > > > > > So if you redeploy or restart then your first request
> > > > should be the
> > > > > > root of your webapp.
> > > > > > Then afterwards should everything just work.
> > > > > >
> > > > > > This is a bug yes i have to rollback code for this because
> > > > > > getServletPath() can't be depended on in a filter
> > > configuration.
> > > > > > Which we do now.
> > > > > >
> > > > > > johan
> > > > > >
> > > > > >
> > > > > > On 11/11/06, Korbinian Bachl
> <ko...@whiskyworld.de> wrote:
> > > > > > >
> > > > > > > Hi,
> > > > > > >
> > > > > > > I issued a severe BUG
> > > > > > >
> > > > > > > http://issues.apache.org/jira/browse/WICKET-42
> > > > > > >
> > > > > > > Best Regards,
> > > > > > >
> > > > > > > Korbinian
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > >
> > >
> >
> >
>
Re: Severe Bug in Wicket 2 / WICKET-42
Posted by Igor Vaynberg <ig...@gmail.com>.
the homepage is always mounted on /
-igor
On 11/12/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
>
> Johan,
>
> i checked all - everything is now behaving as expected. Great Work!
>
> One question however is still to me - why cant i mount a page to "/" ?
>
> so i could mount a PageDistribution class there which then handles the
> PageParams logic... or would this get in conflict with the rest of wicket
> ?
> if so, would it be possible to have a page responsible for every path
> behind
> / but for /wicket/ where the internal could be handled ??
>
> i know this aproach is a bit away from what wicket was made for...
>
> Rgards
>
> > -----Ursprüngliche Nachricht-----
> > Von: Korbinian Bachl [mailto:korbinian@whiskyworld.de]
> > Gesendet: Sonntag, 12. November 2006 10:32
> > An: wicket-dev@incubator.apache.org
> > Betreff: AW: Severe Bug in Wicket 2 / WICKET-42
> >
> > Really fast :)
> >
> > i check it out soon and report,
> >
> > Regards
> >
> > > -----Ursprüngliche Nachricht-----
> > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > Gesendet: Samstag, 11. November 2006 23:17
> > > An: wicket-dev@incubator.apache.org
> > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > >
> > > should be fixed now.
> > >
> > >
> > > On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > >
> > > > well, it works if it is:
> > > > a, the first access of the WebApp with no provious
> > ReDeploy b, the
> > > > root is accessed then
> > > >
> > > > if i redeploy just once, i cant access the root anymore -
> > > however the
> > > > direct path always works...
> > > > if i redeploy it and access the path before, it chains
> > > >
> > > > another danger thing is that following works:
> > > > assume you have IndexedPageParams and index.class mounted
> > to /Index:
> > > >
> > > > if you then access it via /Index/foo/bar/Index/foo2
> > > >
> > > > and print out pageparams.toString you only get 0=foo2 ,
> > meaning he
> > > > access it at last point, and not at first!
> > > > This violates JEE security !!!
> > > >
> > > > i also noticed that if you use a servlet and
> > > BookMarkablePageLinks the
> > > > URLs look like these 127.0.0.1/<servlet>?param=value - and
> > > mounting
> > > > using a servlet was not possible at all
> > > >
> > > > Best Regards
> > > >
> > > > Korbinian
> > > >
> > > >
> > > > > -----Ursprüngliche Nachricht-----
> > > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > > Gesendet: Samstag, 11. November 2006 17:30
> > > > > An: wicket-dev@incubator.apache.org
> > > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > > >
> > > > > does it always just work if always your first access is
> > > the root of
> > > > > your webapp?
> > > > > So if you redeploy or restart then your first request
> > > should be the
> > > > > root of your webapp.
> > > > > Then afterwards should everything just work.
> > > > >
> > > > > This is a bug yes i have to rollback code for this because
> > > > > getServletPath() can't be depended on in a filter
> > configuration.
> > > > > Which we do now.
> > > > >
> > > > > johan
> > > > >
> > > > >
> > > > > On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > > >
> > > > > > Hi,
> > > > > >
> > > > > > I issued a severe BUG
> > > > > >
> > > > > > http://issues.apache.org/jira/browse/WICKET-42
> > > > > >
> > > > > > Best Regards,
> > > > > >
> > > > > > Korbinian
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > >
> > > >
> > >
> >
> >
>
>
AW: Severe Bug in Wicket 2 / WICKET-42
Posted by Korbinian Bachl <ko...@whiskyworld.de>.
Johan,
i checked all - everything is now behaving as expected. Great Work!
One question however is still to me - why cant i mount a page to "/" ?
so i could mount a PageDistribution class there which then handles the
PageParams logic... or would this get in conflict with the rest of wicket ?
if so, would it be possible to have a page responsible for every path behind
/ but for /wicket/ where the internal could be handled ??
i know this aproach is a bit away from what wicket was made for...
Rgards
> -----Ursprüngliche Nachricht-----
> Von: Korbinian Bachl [mailto:korbinian@whiskyworld.de]
> Gesendet: Sonntag, 12. November 2006 10:32
> An: wicket-dev@incubator.apache.org
> Betreff: AW: Severe Bug in Wicket 2 / WICKET-42
>
> Really fast :)
>
> i check it out soon and report,
>
> Regards
>
> > -----Ursprüngliche Nachricht-----
> > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > Gesendet: Samstag, 11. November 2006 23:17
> > An: wicket-dev@incubator.apache.org
> > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> >
> > should be fixed now.
> >
> >
> > On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > >
> > > well, it works if it is:
> > > a, the first access of the WebApp with no provious
> ReDeploy b, the
> > > root is accessed then
> > >
> > > if i redeploy just once, i cant access the root anymore -
> > however the
> > > direct path always works...
> > > if i redeploy it and access the path before, it chains
> > >
> > > another danger thing is that following works:
> > > assume you have IndexedPageParams and index.class mounted
> to /Index:
> > >
> > > if you then access it via /Index/foo/bar/Index/foo2
> > >
> > > and print out pageparams.toString you only get 0=foo2 ,
> meaning he
> > > access it at last point, and not at first!
> > > This violates JEE security !!!
> > >
> > > i also noticed that if you use a servlet and
> > BookMarkablePageLinks the
> > > URLs look like these 127.0.0.1/<servlet>?param=value - and
> > mounting
> > > using a servlet was not possible at all
> > >
> > > Best Regards
> > >
> > > Korbinian
> > >
> > >
> > > > -----Ursprüngliche Nachricht-----
> > > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > > Gesendet: Samstag, 11. November 2006 17:30
> > > > An: wicket-dev@incubator.apache.org
> > > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > > >
> > > > does it always just work if always your first access is
> > the root of
> > > > your webapp?
> > > > So if you redeploy or restart then your first request
> > should be the
> > > > root of your webapp.
> > > > Then afterwards should everything just work.
> > > >
> > > > This is a bug yes i have to rollback code for this because
> > > > getServletPath() can't be depended on in a filter
> configuration.
> > > > Which we do now.
> > > >
> > > > johan
> > > >
> > > >
> > > > On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > > >
> > > > > Hi,
> > > > >
> > > > > I issued a severe BUG
> > > > >
> > > > > http://issues.apache.org/jira/browse/WICKET-42
> > > > >
> > > > > Best Regards,
> > > > >
> > > > > Korbinian
> > > > >
> > > > >
> > > > >
> > > >
> > >
> > >
> >
>
>
AW: Severe Bug in Wicket 2 / WICKET-42
Posted by Korbinian Bachl <ko...@whiskyworld.de>.
Really fast :)
i check it out soon and report,
Regards
> -----Ursprüngliche Nachricht-----
> Von: Johan Compagner [mailto:jcompagner@gmail.com]
> Gesendet: Samstag, 11. November 2006 23:17
> An: wicket-dev@incubator.apache.org
> Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
>
> should be fixed now.
>
>
> On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> >
> > well, it works if it is:
> > a, the first access of the WebApp with no provious ReDeploy b, the
> > root is accessed then
> >
> > if i redeploy just once, i cant access the root anymore -
> however the
> > direct path always works...
> > if i redeploy it and access the path before, it chains
> >
> > another danger thing is that following works:
> > assume you have IndexedPageParams and index.class mounted to /Index:
> >
> > if you then access it via /Index/foo/bar/Index/foo2
> >
> > and print out pageparams.toString you only get 0=foo2 , meaning he
> > access it at last point, and not at first!
> > This violates JEE security !!!
> >
> > i also noticed that if you use a servlet and
> BookMarkablePageLinks the
> > URLs look like these 127.0.0.1/<servlet>?param=value - and
> mounting
> > using a servlet was not possible at all
> >
> > Best Regards
> >
> > Korbinian
> >
> >
> > > -----Ursprüngliche Nachricht-----
> > > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > > Gesendet: Samstag, 11. November 2006 17:30
> > > An: wicket-dev@incubator.apache.org
> > > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> > >
> > > does it always just work if always your first access is
> the root of
> > > your webapp?
> > > So if you redeploy or restart then your first request
> should be the
> > > root of your webapp.
> > > Then afterwards should everything just work.
> > >
> > > This is a bug yes i have to rollback code for this because
> > > getServletPath() can't be depended on in a filter configuration.
> > > Which we do now.
> > >
> > > johan
> > >
> > >
> > > On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > > >
> > > > Hi,
> > > >
> > > > I issued a severe BUG
> > > >
> > > > http://issues.apache.org/jira/browse/WICKET-42
> > > >
> > > > Best Regards,
> > > >
> > > > Korbinian
> > > >
> > > >
> > > >
> > >
> >
> >
>
Re: Severe Bug in Wicket 2 / WICKET-42
Posted by Johan Compagner <jc...@gmail.com>.
should be fixed now.
On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
>
> well, it works if it is:
> a, the first access of the WebApp with no provious ReDeploy
> b, the root is accessed then
>
> if i redeploy just once, i cant access the root anymore - however the
> direct
> path always works...
> if i redeploy it and access the path before, it chains
>
> another danger thing is that following works:
> assume you have IndexedPageParams and index.class mounted to /Index:
>
> if you then access it via /Index/foo/bar/Index/foo2
>
> and print out pageparams.toString you only get 0=foo2 , meaning he access
> it
> at last point, and not at first!
> This violates JEE security !!!
>
> i also noticed that if you use a servlet and BookMarkablePageLinks the
> URLs
> look like these 127.0.0.1/<servlet>?param=value - and mounting using a
> servlet was not possible at all
>
> Best Regards
>
> Korbinian
>
>
> > -----Ursprüngliche Nachricht-----
> > Von: Johan Compagner [mailto:jcompagner@gmail.com]
> > Gesendet: Samstag, 11. November 2006 17:30
> > An: wicket-dev@incubator.apache.org
> > Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
> >
> > does it always just work if always your first access is the
> > root of your webapp?
> > So if you redeploy or restart then your first request should
> > be the root of your webapp.
> > Then afterwards should everything just work.
> >
> > This is a bug yes i have to rollback code for this because
> > getServletPath() can't be depended on in a filter
> > configuration. Which we do now.
> >
> > johan
> >
> >
> > On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> > >
> > > Hi,
> > >
> > > I issued a severe BUG
> > >
> > > http://issues.apache.org/jira/browse/WICKET-42
> > >
> > > Best Regards,
> > >
> > > Korbinian
> > >
> > >
> > >
> >
>
>
Re: Severe Bug in Wicket 2 / WICKET-42
Posted by Eelco Hillenius <ee...@gmail.com>.
> This is a bug yes i have to rollback code for this because getServletPath()
> can't be depended on in a filter configuration. Which we do now.
I introduced that bug two weeks ago. Sorry. Johan and I already discussed it.
Eelco
AW: Severe Bug in Wicket 2 / WICKET-42
Posted by Korbinian Bachl <ko...@whiskyworld.de>.
well, it works if it is:
a, the first access of the WebApp with no provious ReDeploy
b, the root is accessed then
if i redeploy just once, i cant access the root anymore - however the direct
path always works...
if i redeploy it and access the path before, it chains
another danger thing is that following works:
assume you have IndexedPageParams and index.class mounted to /Index:
if you then access it via /Index/foo/bar/Index/foo2
and print out pageparams.toString you only get 0=foo2 , meaning he access it
at last point, and not at first!
This violates JEE security !!!
i also noticed that if you use a servlet and BookMarkablePageLinks the URLs
look like these 127.0.0.1/<servlet>?param=value - and mounting using a
servlet was not possible at all
Best Regards
Korbinian
> -----Ursprüngliche Nachricht-----
> Von: Johan Compagner [mailto:jcompagner@gmail.com]
> Gesendet: Samstag, 11. November 2006 17:30
> An: wicket-dev@incubator.apache.org
> Betreff: Re: Severe Bug in Wicket 2 / WICKET-42
>
> does it always just work if always your first access is the
> root of your webapp?
> So if you redeploy or restart then your first request should
> be the root of your webapp.
> Then afterwards should everything just work.
>
> This is a bug yes i have to rollback code for this because
> getServletPath() can't be depended on in a filter
> configuration. Which we do now.
>
> johan
>
>
> On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
> >
> > Hi,
> >
> > I issued a severe BUG
> >
> > http://issues.apache.org/jira/browse/WICKET-42
> >
> > Best Regards,
> >
> > Korbinian
> >
> >
> >
>
Re: Severe Bug in Wicket 2 / WICKET-42
Posted by Johan Compagner <jc...@gmail.com>.
does it always just work if always your first access is the root of your
webapp?
So if you redeploy or restart then your first request should be the root of
your webapp.
Then afterwards should everything just work.
This is a bug yes i have to rollback code for this because getServletPath()
can't be depended on in a filter configuration. Which we do now.
johan
On 11/11/06, Korbinian Bachl <ko...@whiskyworld.de> wrote:
>
> Hi,
>
> I issued a severe BUG
>
> http://issues.apache.org/jira/browse/WICKET-42
>
> Best Regards,
>
> Korbinian
>
>
>