Click webbrowser button

  • Thread starter Thread starter the_only_steven
  • Start date Start date
T

the_only_steven

Hello all,

I'm trying to make my program autologin in a website but it isn't
really working.

The website has 2 textfields one for user name and other for pw.
When you filled in those you click the button (like on all websites).

I found how to fill in the 2 textfields but with the button i have a
problem.

This is the sourcecode from the website
<b>Username:</b></font><br>
<input type="text" name="login_username" size="20"><br>
<font color="#FFFFFF" face="Verdana" size="2">
<b>Password:</b></font><br>
<input type="password" name="login_password" maxlength="16"
size="20"><br>
<input type="submit" value="Login"></p>

the 2textfields are done like this
DirectCast(GetCurrentWebForm.item("login_username"),
mshtml.HTMLInputElement).value = Text
DirectCast(GetCurrentWebForm.item("login_password"),
mshtml.HTMLTextAreaElement).value = Text

The problem with the button is that it doesn't have a "name".

Anyone knows how i can click this button then?

ty,
Steven
 
I can but then i cant really protect my program from being used by
people that i dont want.

If i can find out how to click the button then i dont have to show the
webpage and i can refer to an other site (db) saying which user has
acces and which hasn't.
 
I can but then i cant really protect my program from being used by
people that i dont want.

If i can find out how to click the button then i dont have to show the
webpage and i can refer to an other site (db) saying which user has
acces and which hasn't.
Click to expand...

Why not build the security into the program instead of trying to interact
with site security?

The idea is you compress/encrypt the pages you upload that are private, then
you have your program access and decrypt/convert these pages (with or
without password checking by the program as per your overall specs) prior to
display.

This way the access security for your users is adequately separated from the
access security for you. So if someone does steal and reverse engineer your
program, all they get is access to the page compression/encryption and not
control of your site.

A system like this can allow you to distribute keys so that when the
security system reaches a certain risk of compromise, you re-encrypt under a
new key and distribute the new keys to the people still on your members
list. This can also be done in a public/private key system which although
more convenient, has mathematically fewer possible key combinations than
symmetric encryption thus asymmetric encryption, though more secure from key
theft, is less secure from both hacking and cracking.

Encryption only imposes a delay on acquiring your data, and the delay needs
to be engineered with respect to the pace of technological advancement and
the use-by data on the data. If this is indefinite, then you need to
introduce masking in addition to encryption. A mask is where a particular
cryptotext returns one of a dozen or more linguistically correct (expected
terminology and grammar in context of the known field) possible plaintexts
depending on which of a dozen or more keys are used. Of course other keys
return gibberish, but the amateur will stop hacking with the first plaintext
return - which is most likely a decoy. The professional will invariably get
all of the plaintext possibilities in time, but without obvious contextual
hindsight, direct evidence, or your surrendering of the correct key cannot
legally prove which of the plaintext possibilities is the real thing. The
ultimate mask returns a randomly generated and yet linguistically correct
plaintext for every incorrect key and is only possible if the encryption is
applied to a code (word-based) instead of a cipher (letter, character, or
syllable based). A good encrypted code mask has no legal remedy without the
key.

Intelligence analysis is another thing altogether and intelligence
misdirection an art unto itself - but may prove handy if you've heavy
commercial competition trying to second guess your direction. Masking is not
just a neat way to package selective misdirection. Of course, if you
distribute more than one key for a single cryptotext delivering multiple
plaintexts depending on key (IE member class - Figures for accountant and
shareholders, proposals for marketing to investigate, specifications for the
engineering team, and a dozen other useless plaintexts that support the
overall context) then you have the beginnings of a forest in which to hide a
tree.

Good luck...
 
Back
Top