Saturday, September 29, 2018

Information Security

In my line of work, I have to know a lot about information security. My focus is more on software and websites, but some of my coworkers set up secure networks and workstations. Yet even when everything is done correctly, that security may still fail. In most cases, the weakest link turns out to be the users.

I've organized this post for easy skimming if you want to just read the bold text.  Or, you can dive down into the details and hopefully gain a little more understanding.

Everything can be hacked.


The question isn't whether something can be hacked—it's how hard it is to hack. A flower shop computer is definitely easier to hack into than a nuclear reactor, but both can still be hacked if you know how to do it.

Your goal should be to make it very difficult to gain access to your data. The inexperienced hackers should not be able to figure it out, and the experienced hackers should not consider it worth their time. If you're storing government secrets, you will want to go even further, but that is not within the scope of this blog post.

Find a trustworthy tech person.


Even if you know a thing or two about computers, find someone whom you can trust for tech support. And be sure to make it worth their while to help you when needed.

Be sure that you can trust them.


You may have to give a tech access to your data for them to fix something, so trust is very important here. I'd recommend going with someone experienced, because experience will tell them that they really don't want to know what you have on your computer.

Dont' jump between multiple techs.


Each additional person adds a potential point of failure. Keep the circle small.

Secure your own computer.


The news mostly reports about websites and databases being hacked, but most of the data breaches I've seen first hand have been on personal computers. Make sure that your computer is not the weakest link.

Keep your computer updated.


Never turn off automatic updates, and routinely update the software that you use. As security holes are found, they are fixed. You want the latest fixes.

This also includes your virus protection software. Make sure that it is updated regularly so that it knows about the latest viruses.

Never turn off your firewall.


If your "trusted tech" tells you to turn off your firewall for more than a few minutes, then call them stupid and find someone else. It's okay for them to turn it off for a minute to diagnose a problem, but it should never be turned off permanently.

Be careful opening files that you didn't create.


Programs are dangerous, Word documents are questionable, and pictures are only usually safe. I say "usually" because it is possible to put a virus into any type of file. Some are easier to do this with than others, so the easier ones such as programs and documents should be given an especially thorough looking-over.

If you do not trust the person who sent you the files, or if anything looks suspicious, then don't open it until you've verified that it is safe. This can be handled a variety of ways, including: talking to the person who sent it, conferring with your "trusted tech", or simply deleting the file.

Regularly back up your files.


Even if we are super careful, we may still get into trouble. As a backup plan, you should regularly backup of all of your important files. Not only is this a good idea from a security perspective, it is also a good idea from a data retention perspective.

Surfing the 'Net


You can go overboard with security by using VPNs, proxies, and other tools, or you can just accept the fact that some websites just aren't very trustworthy. I prefer to go with the later and follow a few simple security rules.

Be careful what websites you visit.


Websites that are taboo seem to be the worst about being malicious. I'm not sure why, but it's almost as if the people who make such websites don't have very high morals.

Even websites that aren't about taboo topics may be malicious. If you don't know the website, don't trust it. You may still want to visit it, but don't enter in any of your personal information or passwords.

Get an ad blocker.


Usually, the website has little control over the ads that it serves, and some ad platforms aren't very secure. Using an ad blocker will block a lot of potentially harmful content. As an added bonus, you get legitimate advertisements blocked too!

Do not trust tech support popups.


The #1 most common way that I see people infected is by calling fake "Microsoft" because of a popup when they visited a malicious website. Instead of calling the number on the screen, call your "trusted tech" that I mentioned earlier.

Pay attention to what your browser says.


Most popular browsers will tell you in the address bar whether a connection is secure. If the website is not secure, then do not enter any sensitive information on that website. Also, if your browser stops you with a big red warning screen, that means stop and close the tab.

Social Networking


While most popular platforms are secure, they usually do sell our data. Those third parties are the weak point. A chain is only as strong as its weakest link, so these super-secure systems may not really be so secure after all. That's okay so long as you keep it in mind.

Do not share any secrets.


Not only do your friends talk, the social network does too. If you don't want something to be public, then do not share it on a social network. It will be seen by more people than you intend to see it.

Your legal notices do not work.


I see a lot of posts that go something like, "I do not give Facebook permission to share my data." People expect for that to be legally binding, but it's not.

For such a text to be legally binding, all parties must intentionally agree to it, and there must be a record of such an agreement. I've never seen Zuckerberg comment "I agree" on such posts, but I guarantee you that all of us checked "I agree" while setting up an account.

Disclaimer: While I've done my research on this topic, I am not a lawyer and this is not legal advice. Consult a lawyer if you're serious about wanting to press this issue.


Be careful what third parties you allow access to.


Those "Sign in with [whatever]" buttons are very convenient! I no longer have to create an account for every service that I use. But they come with a downside – you are giving a third party access or control of some of your data.

On all popular social networks that have this feature, you can review which third parties have an access token and what data is shared with them. Review this and revoke access to the ones that you don't want on there.

Emails


Emails are not secure. We've been throwing technologies at it for years to improve its security, but it still sucks as a secure medium. Still, the convenience of it may be worth it if you're careful.

Verify unknown senders or emails.


If you receive an email from someone's "other" account that you haven't seen before, call and verify that it's them. If you receive an email that it doesn't make sense for you to have received, call and verify that they sent it. Or, just delete the emails if they don't look important.

Encrypt or don't send sensitive information.


If you want to leave your key under the doormat, that's fine, but don't be surprised when a robber finds it. It's the same concept here. If you do not encrypt sensitive information sent over email, then don't be surprised when a hacker gets it. I use this analogy because it's really up to you whether to go with convenience or security.

Be careful of attachments.


Every time that you open an attachment, you're at risk. Even if you trust the sender, who's to say that their computer hasn't been compromised. Weigh the risks, consider the convenience, and decide whether to gamble with your security.

Is it worth it?


As a software developer, it is my job to protect computer users. My aim is for my creations to never be the weakest link, but I can never dictate what else the user may have on their computer or how they may use it. Sometimes it feels like a losing battle, but that isn't how we should look at it.

With computers, we can do so many incredible tasks that we couldn't easily do before. The cost is a little bit of security. Is the benefit worth the cost? I think that it is. We just need to be aware of the risks and take appropriate precautions.

This is by no means an exhaustive list of precautions that can be taken, but I consider these to be some of the basics. I hope that this helps some of you to be safer on computers so that you can continue to enjoy that very useful technology.

Saturday, March 10, 2018

The Exception to the Rule

Exceptions in Programming Languages


I learned a neat little construct recently: exceptions!

Okay, so it's been 13 years since I first heard about them, but it wasn't until I read this StackOverflow question that I felt like I actually understood them. I finally felt like I knew when to use exceptions and when not to use exceptions.

I don't want to focus on the details of exceptions, but perhaps a little crash course is appropriate.

What are exceptions?


Exceptions can be thrown and caught. When you "throw" an exception, your program stops dead in its tracks. It picks back up wherever you have chosen to "catch" the exception. There's a little behind-the-scenes stuff too, but that's the gist of it.

    try {
        cout << "The program gets to here... ";
        throw exception();
        cout << "... and it never gets here." << endl;
    }
    catch (exception & e) {
        cout << "... and it picks up here." << endl;
    }

If you run this code, it will output the following:

The program gets to here... ... and it picks up here.

It works, but is this good code? Are there better ways to do the same thing?


Pros and Cons of Exceptions


Exceptions are slow. Yes, they can be terribly slow! It is much faster to use conditionals instead of exceptions. (Or, in the case of the program above, just drop the dumb lines that are only there for the example's sake.) Exceptions are just plain slow!

However, not having an exception is faster than conditional execution (like if, while, etc.). If you /can/ throw an exception but you don't, that's super fast! They are only slow when they are actually used.

That's the gist of the answers on the StackOverflow question I linked earlier.


When do you use exceptions?


After meditating on that StackOverflow question and its answers, I feel like I finally know when to use exceptions. Bare with me, and I'll explain.

A few givens:
  • Exceptions are slow, so you don't want to use them.
  • Having exceptions but not using them is fast.
  • Conditionals are almost as fast as unused exceptions. *
* at least in the languages I cared to research last weekend.

My conclusion is that exceptions should never happen. They should never be used. They should never be thrown. They should never be caught.

The exception to the rule "exceptions should never happen" is "an exception happens".


Thoroughly confused...


I get up and go to work every morning at 7:50am. I do this every weekday. This is normal to me.

Six months ago, I got in my car at 7:50am, and it wouldn't start. This is an exception! I had to change my plans because of this unexpected event.

I have come to the realization that exceptions in programming are exactly the same thing. In normal operation, how should the program work? Use exceptions for exceptions to this rule of how things should work.
  1. Use a conditional when: It's something that needs checked or just happens sometimes. Your program is coded to handle anything amiss.
  2. Use an exception when: Uh oh! Something happened, and the program wasn't built for this!
In short, exceptions never happen, and the exception to this rule is when they happen.


I need an example.


Perfect! Let's say you've got a text editor. Usually, you'll be trying to open valid files, but sometimes the user tells you to open a file that doesn't exist.

bool openFile(const char * filename)
{
    this->f = fopen(filename);
    if (!f)
        return false;
    _readContents()
    return true;
}

This is a member function of our TextEditor class. You try to open a file. If it doesn't exist, then no harm done. The program keeps on chugging away as usual, probably prompting them for a valid file name.

Now let's look at a different case. What if you are trying to open a configuration file for your program?

    FILE *f = fopen("myConfig");
    if (!f)
        throw exception("Config not found!");
    _parseConfig(f);

It's different. We expect the config file to exist. If it doesn't, we just can't continue. This is an exception.

We will, of course, catch the exception somewhere. If we're good, we might try to recreate the config and try again. But, for the moment, we've hit an unexpected obstacle that needs fixed before we can continue normal operation.


Thursday, February 23, 2017

SQL Injection: What it is and how to avoid it

A while back, we (where I work) took over a project from another company. I was going to make some small changes and have it up before my second cup of coffee! Then I opened the code and realized I would be working late that day instead.

Among other issues, the code was vulnerable to SQL Injection.

SQL Injection is one of the easiest hacks to do. It's also one of the easiest to protect against. Still, it's a pretty common vulnerability! As much as I would love to have a backdoor into tons of sites, I can't help but lose sleep over this problem being out there. I have to share a few solutions.

What is SQL Injection?


For an example, let's take a very simple login page. I won't post the code, but the basic idea is to take a username and password, compare it to what's in the database, and log them in if it's a match. Easy peasy!

Here's the SQL to match it up in the database:

SELECT id FROM users WHERE username='bob' AND password='1234'

If it doesn't return anything, the username and password is wrong. If it finds a match, it returns the user's ID.

What if we change the username to bob'--?

SELECT id FROM users WHERE username='bob'--' AND password='1234'

Whoah! We just made it so that we don't even need the password! We can log in as anybody so long as we know their username!

The problems don't stop there. We can change a page that displays a list of items to sort/filter however we wish. We can increase the number of results returned and use that to do a very good DOS attack.

If the webhost is particularly unlucky, we can even execute SQL for other websites that they host -- even if those sites aren't vulnerable! (Hint: "USE abcDatabase;")

How do we protect against it?


That's easy. Sanitize your inputs!

Way 1 - Escaping data


You can properly escape the data before putting it into the SQL query.

Most languages have a function to escape out harmful characters, and it's usually pretty easy to use. This post is language-agnostic, but the basic idea in most languages is to pass in a potentially harmful string and get back a properly escaped one. The SQL query becomes the following:

SELECT id FROM users WHERE username='bob''--' AND password='1234'

As you can see, the SQL injection attempt will now be unsuccessful!

This time.

Watch out!


What if we have the following query that deletes the message with a given ID?

DELETE FROM messages WHERE messageId=12

It looks pretty harmless. Of course, we'll be sanitizing any user input anyways, right?

What if we give our code "1 OR 1=1" as the ID to delete? There are no special characters to escape. The string will get substituted as-is. What does this do to our query?

DELETE FROM messages WHERE messageId=1 OR 1=1

Oh, it just deletes everything. Even though we escaped all of the harmful characters, we are still vulnerable to SQL injection!

The quick solution is to put single quotes around every "variable" you're substituting in. It will work, but then you might run into weird logic errors when doing comparisons. A better solution is to use parameterized queries. Read on.

Way 2 - Parameterized Queries


Parameterized queries will properly escape any data, will verify that the type of data matches up, and will (hopefully) make sure everything's dandy with the character encodings. They are the best way to do SQL queries.

SELECT id FROM users WHERE username=@username AND password=@password;
DELETE FROM messages WHERE messageId=@messageId;

It might look a little different in different languages, but that's the basic idea.

Now you need the code that puts in the data. This depends a lot on the language, but the basic idea is to say what type of data each parameter is and then to set it to a value. In VB.Net, here is the code for the second query:

Dim query = "DELETE FROM messages WHERE messageId=@messageId"
Dim cmd As New SqlCommand(query, connection)
cmd.Parameters.Add("@messageId", SqlDbType.Int)
cmd.Parameters("@messageId").Value = 12
cmd.ExecuteNonQuery()

Parameterized queries are more verbose and somewhat confusing at first, but do use them. Use parameterized queries. Use parameterized queries.

Conclusion and Final Thoughts


SQL Injection is very easy to do, very common, and yet very easy to protect against. The simple solution is to just use parameterized queries everywhere. This OWASP page has some handy code snippets for your reference.

Here's a rule: Always use parameterized queries. Never insert your data directly into a query string (even if properly validated and escaped first). I don't care what your teacher or textbook says. Use parameterized queries. The only exception to this rule is when you know the risks and make an informed decision to go another route.

Saturday, January 12, 2013

Playing a Pitch in an Android App with AudioTrack

 I've been working on an android app lately that can be used to tune instruments. I've run into a road block with the pitch detection, so I'm starting over and making the code nicer. While I'm at it, I figure I may as well share some of what I've learned!

Sunday, January 6, 2013

AVL vs. Red-Black: the conclusion

    I accidentally deleted my original blog post about this, and I can't recover it despite trying all sorts of different tricks. So, here it is reposted. Sorry for any broken links.

Thursday, January 3, 2013

Year of Change


December 21, 2012 came and went, and the world is still here.  I didn't look a lot into the Mayan Calendar, but I think I remember reading that every time one of their circle calendar thingeys rolls over, it's a new era that brings change.

Every year, on December 31, people decide to make new years resolutions.  This year, why not assume the Mayans had something going right and make a resolution to see change?

People have been chanting about "change" a lot lately, but what has really changed?  And why are we still living relatively ordinary lives?  Myself, I am terrified of change!  I like my comfort zone, and I'm sure most of you prefer to not take risks, too.

We all have opportunities.  Why not take them?  You have a lot to lose, but you have even more to gain!

I stopped spending so much time programming a while back.  I got a guitar and played it until my fingers blistered and cracked.  On the 31st, I bought a new car.  Yesterday, someone called me and asked me to come work at some company.

Change.  I have an option to refuse it and continue to live in comfort.  I think I'll make a change in my habits and seek to change.  I'm taking the risk, making a gamble, and I think the odds are in my favor.  Will you join me?

I got my first blow yesterday.  I'm apparently not as great at singing and playing the guitar as I thought.  I knew I wasn't great, but am I not at least a little good?  Nope!  But I also found some encouragement in an unlikely place.  I found a youtube video of one of my favorite singers when she was younger.  I love her to death, but she just could not sing!  Now, she's going on world tours!  If she can do it, why can't I?  And if I can do it, why the hell are you still reading this?  Get out there and live a little!

Tuesday, December 11, 2012

Resistance Accomplished With Penguin Jokes!


My heart is saddened and worried as I resist the urge to dive into my old computer habits to avenge my newest love.  "That would be so cruel!"  Or, "I should get caught!"

I'm reminded of the last guy who thought he was clever with his 1337 5k*11z...  I sometimes miss his crazy ideas of how to annoy me.  And the guy before him who avoids me like the plague?  So sad...  so sad.

My evil side is hurt that I abandoned my computer world even though it was out of evil anger.  Yes, folks, with the exception of little scripts here and there to help me accomplish a task, I quit programming.  It's caused more harm than good, so I am looking for a new talent.

My newest talent is joke telling!  No, not really, but I am looking for good penguin jokes to make me feel better.  They must be penguin jokes!  Here are a few I've found so far (and 2 I knew):

FAT PENGUIN!!!!!!!!!!!!!!!!!!!!!!!!!!!
"What?"
First joke, to break the ice!!!

Q: What do you call a penguin in the desert?
A: Lost!

Q: Why don't you ever see penguins in Britain?
A: Because they're afraid of Wales!

Two penguins are standing on an ice floe. The first penguin says, "you look like you're wearing a tuxedo." The second penguin says, "what makes you think I'm not?"

Penguins are black and white.  Old TV shows are black and white.  Therefore, penguins are old TV shows.

Q: And, lastly, how do you express your delight at a penguin joke?
A: Cool!

Any more?  Please share them below!


P.S. Don't take anything you read above seriously except for the request for more penguin jokes.  Most of it is silliness.  It's sad that I feel like I have to say this, but that's the world for you!  So, no hate mail from penguins!  Thanks!