Develop Reference

Getting user agent value. Server side vs Client side?

I need to pass user agent value into front end.
I can get this value using $_SERVER['HTTP_USER_AGENT'] and write it into front end.
(Actually I will be using Mage::helper('core/http')->getHttpUserAgent(), but I think it's just a magento helper to call above mentioned function.)
Or I can use get navigator.userAgent with js on client side.
Which better and why? My primary concern is speed.
p.s. I understand that UA can be easily manipulated. We are not basing any serious functionality on the value, it's used as a secondary parameter.

I would personally use navigator.userAgent. Mainly, because passing values from PHP to JavaScript is pretty ugly in my opinion. Also, the value will be exactly the same for both. Even if someone decides to edit their useragent.
I think simplicity takes the cake here.

Performance will depend on the purpose. If you need this inside php, then use the server variable with helper getter you mentioned above. For js use navigator object.
In general both navigator.userAgent and HTTP_USER_AGENT are variables of Request Header and are both already present in memory (of server or users browser in case of js). So no measurable performance difference is possible.

How can I prevent someone from altering or avoiding my JavaScript logic when adding buttons to jQuery UI dialog?

I am using a jQuery UI dialog and I am adding buttons to the dialog in JavaScript based on some entitlements logic. (I pass in a boolean from my server-side AJAX call if I am entitled and then I show different buttons based on that flag.)
What concerned me is what is preventing someone from using developer tools like Firebug and putting a breakpoint on that line that does the check and either altering the flag or dragging to skip over that entitlements check.
So my question is specific to adding buttons onto a jQuery UI dialog (because its not like you can add the buttons from the server side since its a jQuery plugin), but I guess it highlights a more general point around any entitlements logic on the client side being "vulnerable". So if there are any general best practices around this point I would be interested (but still looking for an answer to my specific example).
NOTE: I am also doing a server-side entitlement check on POST as a backup, so I am still "protected" but I am still concerned about the point above.

Nothing prevents people from altering client-side code, it is inevitable.
You can, however, add buttons of the kind of "server-side", you just retrieve a string using the AJAX call, which happens to be a JavaScript function that adds buttons. And on the client side do eval() on that string which will execute the retrieved JS function and will add the buttons. Moreover, you can transmit your entire JavaScript code that way, so the client cannot skip anything since all is being executed in the eval().
A quick example:
Server-side function returns
string banana= "alert('test');";
return banana;
and client side does
eval(response.d);
Here is a theoretical example: FIDDLE

You cannot control what clients will do with your scripts, nor what requests they will make of your server. You must design your back-end API (not your JS client) to be the "gating mechanism" between the user and your system. It's best not to think of the JS as part of your system, but as a separate client that you ship as a reference implementation for your API.
But, if you wanted to at least make it difficult for users to mess with your code, you could minify and concatenate your JS scripts with something like Closure.

As the other person suggested, you cannot implement security on the client for exactly the reason you point out. You could use basic auth, or try setting up a token based approach.

What risk or liablities in using eval() in the following manner

I'm working on creating one of those robot games. The user creates a robot and then puts it in a battlefield with other robots.
I'd like to let the users use javascript to program their bots. I'll provide a number of functions for them to call, but they also can build thier own. (sorta)
To date, the only solution I have come up with is to use the javascript eval() function to execute the code the users have written.
I want to know two things:
Anyone have any alternative suggested implementations that still allow the users to write in javascript?
Can the users do anything with this flaw that they could not do using the firefox javascript debugging tools? (ie: on their own without my use of the eval() function)
Note: The javascript code is stored within mySQL. ajax is used to pull the jscript out and display to users. ajax is used to send javascript updates back into SQL. All code submitted by users and about to be inserted in the database is run through a "clean()" function.

So basically you will allow UserA to write javascript which will be evalled in UserB's browser?
If so, then that sounds like a fairly bad idea ;)
You could use a middle layer such as http://code.google.com/p/google-caja/wiki/CajaCajole to make it a bit safer.
An example of what they could do is: write javascript which will present what looks like your login page, then send the username and password to another server.
Another example would be to inject a script tag which then gets the 'full' payload which could get up to all kinds of mischief, like fx showing a friendly popup with the new exclusive downloadable Portal game that you got a special deal with Steam to make available etc etc. Just download and Run! Then it creates a hidden iframe to some trojan cdn. :)

I never thought I'll say this, but Project Narcissus might be of use to you. It's a JavaScript engine written in JavaScript.

Cool idea.
eval does have a slight disadvantage against other methods of script injection.
You can create a function on the fly with Function. Try this:
var command = "alert(123)";
var doStuff = new Function(command);
doStuff();
eval runs in the private scope, Function runs in the global scope. That means if you have an internal value that bots aren't supposed to be able to modify, they might have access to it if you run their logic through eval, but they shouldn't if you use Function. More info here:
changing string to a function in javascript (not eval)

Many AJAX libraries can be set to execute the returned JS automatically. No need for eval().

The most important thing is to let pages containing user scripts run on a separate, "sandboxed" domain that has no session cookies from the main site that could be connected to user accounts and such.
That, together with some manual monitoring of the submissions, will already take away a lot of the script injection risks.
There will always be some risk of malicious code being run on the user's browser when allowing Javascript from your users, but it stands to reason that getting malicious JavaScript is a general risk on the Internet, and it's up to the client to protect against it.
What I wouldn't do is eval() user-entered JavaScript inside the main domain of the project. That opens too many real dangers of attack.

Javascript eval limits

Is there a limit to javascript's eval, like in lenght?
I'm trying to build an app where you can store JS code in the DB, which you can later load and eval in order to execute it, but i'm reaching a limit.
First of all, the code has to all be in one line. Any multiline statements are not executed.
Next, i'm reaching a limit in length (i guess). If i execute the code manually, it works, but put that same code in the db, load it via ajax, and try to execute it, and it fails.
Any ideas why?

You don't need to use eval and its not exactly a good thing to use. You could just have it print out to the page and it will run.
Here is the accepted answer on why you should not use eval:
Improper use of eval opens up your code for injection attacks
Debugging can be more challenging (no line numbers, etc.)
eval'd code executes more slowly (no opportunity to compile/cache eval'd code)

I have run into this also. As others have said here - eval comes in handy when you are generating the Javascript on the fly and then want to have it execute on the browser. My usages of this technique are to go small things like a simple function that will just make a call back to the server when a button is pressed. Depending upon the circumstances there might be two functions or just one. I've also used it to display information that changes from a database. The information is always just plain text. So no injection attack can be done.
Anyway, I too have run in to this limitation of the Javascript EVAL statement and it seems to me that there is a 1024 character limit. When I go over this I start getting weird things like eval just spitting out the original text. This is really evident because I hex everything before sending it to the browser so I can have things like single and double quotes in the text without it causing eval any problems. (And hexing everything helps prevent injection attacks.)
I also side with the person who said to use getscript in jQuery. It works just as well as the eval without the size limitations. The only extra step you have to take is to create the Javascript file first.
I hope this helps and answers the original poster's question. That being I believe the size limitation is 1024 bytes.

You could create a javascript function that creates a script-tag dynamically (createElement('script') and append it to the head- or bodytag) and point the source to your app. The src can contain parameters, used like a get request, like for example: src="jsapp.aspx?script=myscript&includefunction=loadfn" No eval needed. You can even define an onload handler for your new script tag. Plenty of documentation on the net for that.
You wouldn't even have to use XHR (AKA Ajax) for that.

Executing JavaScript from Flex: Is this javascript function dangerous?

I have a flex application that needs the ability to generate and execute JavaScript. When I say this, I mean I need to execute raw JavaScript that I create in my Flex application (not just an existing JavaScript method)
I am currently doing this by exposing the following JavaScript method:
function doScript(js){ eval(js);}
I can then do something like this in Flex (note: I am doing something more substantial then an alert box in the real Flex app):
ExternalInterface.call("doScript","alert('foo'));
My question is does this impose any security risk, I am assuming it's not since the Flex and JasvaScript all run client side...
Is there a better way to do this?

There's no need for the JavaScript function, the first argument to ExternalInterface can be any JavaScript code, it doesn't have to be a function name (the documentation says so, but it is wrong).
Try this:
ExternalInterface.call("alert('hello')");

This isn't inherently dangerous, but the moment you pass any user-provided data into the function, it's ripe for a code injection exploit. That's worrisome, and something I'd avoid. I think a better approach would be to only expose the functionality you need, and nothing more.

As far as I know, and I'm definately not a hacker, you are completely fine. Really, if someone wanted to, they could exploit your code anyway clientside, but i don't see how they could exploit your server side code using javascript (unless you use server side javascript)

I don't see where this lets them do anything that they couldn't do already by calling eval. If there's a security hole being introduced here, I don't see it.

Remember also that the script actions are controlled by the "AllowScriptAccess" tag in the statement. If the web page doesn't want these actions, they should not permit scripts to call out.
http://kb.adobe.com/selfservice/viewContent.do?externalId=tn_16494

ExternalInterface.call("eval", "alert('hello');");