[Edbrowse-dev] Login failure with 3.5.1

[Edbrowse-dev] Login failure with 3.5.1

[Charles Hallenbeck]

Hi people,

I still find submit buttons that don't submit, and have a brief log of
one such, with db7 selected in edbrowse. Any ideas about this?

db7
.
{Login}
g
findField returns 270, javascript:__doPostBack('ctl00$ctl13$LoginButton','')
go 1 nogo 1 jsh 1 dead 0
click 0 dclick 0 over 0
undoCompare no undo map
jSyncup starts
jSyncup ends
javascript:
javascript:__doPostBack('ctl00$ctl13$LoginButton','')
TypeError: __doPostBack is not a function
function script$$pending



edbrowse returns to the line on which I typed g.

Chuck


-- 

 Chuck in Ghent, northeast of Hudson on the Hudson.

"An excess of reason is itself a form of madness."
 -- Kim Stanley Robinson

Re: [Edbrowse-dev] Login failure with 3.5.1

[Kevin Carhart]

hi Chuck,

I think this syntax:
> javascript:__doPostBack('ctl00$ctl13$LoginButton','')

is a microsoft aspx trying to do ajax.  I don't know what to do, but I'm 
interested in figuring it out and am (still) working on it.

I had an idea for how to support pages with ajax.  I wonder what Karl, 
Chris & Adam think of this?   I'm interested in bringing in 
information from the javascript debugger that is available as part of 
spidermonkey.  jsdbgapi.c is like the JSAPI only with ways of 
addressing stack frames - I think?

So I think if you are running Chuck's site and stepping through moments in 
time, you would eventually hit a stack frame that contained all the 
parameters to doPostback, all assembled and ready to go (hopefully).  A 
'step' command would work neatly with the CLI. If a full URL is sitting 
there in memory just for one step's worth of time before being sent to an 
ajax routine as a parameter, we could grab it, browse it and work with 
the output.  So we don't have to "support ajax" if we can deconstruct 
ajax and do what the ajax would do, but turn-based and CLI rather than 
constant, chatty network activity for little bits, and with a 
graphical bias.  What do you think?

thanks
Kevin




On Fri, 20 Jun 2014, Charles Hallenbeck wrote:

> Hi people,
>
> I still find submit buttons that don't submit, and have a brief log of
> one such, with db7 selected in edbrowse. Any ideas about this?
>
> db7
> .
> {Login}
> g
> findField returns 270, javascript:__doPostBack('ctl00$ctl13$LoginButton','')
> go 1 nogo 1 jsh 1 dead 0
> click 0 dclick 0 over 0
> undoCompare no undo map
> jSyncup starts
> jSyncup ends
> javascript:
> javascript:__doPostBack('ctl00$ctl13$LoginButton','')
> TypeError: __doPostBack is not a function
> function script$$pending
>
>
>
> edbrowse returns to the line on which I typed g.
>
> Chuck
>
>
> -- 
>
> Chuck in Ghent, northeast of Hudson on the Hudson.
>
> "An excess of reason is itself a form of madness."
> -- Kim Stanley Robinson
>
> _______________________________________________
> Edbrowse-dev mailing list
> Edbrowse-dev@lists.the-brannons.com
> http://lists.the-brannons.com/mailman/listinfo/edbrowse-dev
>

--------
Kevin Carhart * 415 225 5306 * The Ten Ninety Nihilists

Re: [Edbrowse-dev] Login failure with 3.5.1

[Adam Thompson]

[-- Attachment #1: Type: text/plain, Size: 2121 bytes --]

On Fri, Jun 20, 2014 at 12:49:24PM -0700, Kevin Carhart wrote:
> 
> hi Chuck,
> 
> I think this syntax:
> >javascript:__doPostBack('ctl00$ctl13$LoginButton','')
> 
> is a microsoft aspx trying to do ajax.  I don't know what to do, but I'm
> interested in figuring it out and am (still) working on it.

Yeah, I've run into it a few times.

> I had an idea for how to support pages with ajax.  I wonder what Karl, Chris
> & Adam think of this?   I'm interested in bringing in information from the
> javascript debugger that is available as part of spidermonkey.  jsdbgapi.c
> is like the JSAPI only with ways of addressing stack frames - I think?
> 
> So I think if you are running Chuck's site and stepping through moments in
> time, you would eventually hit a stack frame that contained all the
> parameters to doPostback, all assembled and ready to go (hopefully).  A
> 'step' command would work neatly with the CLI. If a full URL is sitting
> there in memory just for one step's worth of time before being sent to an
> ajax routine as a parameter, we could grab it, browse it and work with the
> output.  So we don't have to "support ajax" if we can deconstruct ajax and
> do what the ajax would do, but turn-based and CLI rather than constant,
> chatty network activity for little bits, and with a graphical bias.  What do
> you think?

I suspect if you're a js expert it'd be possible,
however it'd be difficult at best due to the way ajax js is usually constructed.
In addition, I've never seen the debugger in the new js version (that's not to
say it's not there), and my general instinct is not to rely on things like this.
At the end of the day if someone told me I had to debug all the js to get the
ajax parameters, then enter them manually and then hope it works,
I'd probably tell them no thanks.

However, I like the idea of finding a CLI-friendly way to do ajax.
I think something like the edbrowse timer behavior may be a possible approach.
First of all we need code to allow js to do http, and then we can construct the ajax machinary around this.

Cheers,
Adam.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 473 bytes --]

Re: [Edbrowse-dev] Login failure with 3.5.1

[Kevin Carhart]

Question for Chuck, could you tell me what particular website this was? 
Unless you'd rather not.

thanks
Kevin

On Fri, 20 Jun 2014, Charles Hallenbeck wrote:

> Hi people,
>
> I still find submit buttons that don't submit, and have a brief log of
> one such, with db7 selected in edbrowse. Any ideas about this?
>
> db7
> .
> {Login}
> g
> findField returns 270, javascript:__doPostBack('ctl00$ctl13$LoginButton','')
> go 1 nogo 1 jsh 1 dead 0
> click 0 dclick 0 over 0
> undoCompare no undo map
> jSyncup starts
> jSyncup ends
> javascript:
> javascript:__doPostBack('ctl00$ctl13$LoginButton','')
> TypeError: __doPostBack is not a function
> function script$$pending
>
>
>
> edbrowse returns to the line on which I typed g.
>
> Chuck
>
>
> -- 
>
> Chuck in Ghent, northeast of Hudson on the Hudson.
>
> "An excess of reason is itself a form of madness."
> -- Kim Stanley Robinson
>
> _______________________________________________
> Edbrowse-dev mailing list
> Edbrowse-dev@lists.the-brannons.com
> http://lists.the-brannons.com/mailman/listinfo/edbrowse-dev
>

--------
Kevin Carhart * 415 225 5306 * The Ten Ninety Nihilists