Types of Errors in Programming

Programming is a complex process requiring a great attention to details. Making mistakes here is, probably, inevitable. Some of the bugs, the most unpleasant ones, are logical by nature. It is extremely difficult to find them, and quite often such errors occur when a product is already in operation. Syntax errors are the most innocuous ones. In almost all languages, they are detected by interpreters and compilers. Some of the errors can be detected during automatic or beta testing.

Today we’re going to talk about the most common mistakes that almost every developer makes. Without a single doubt, being equipped with comprehensive knowledge on bugs that may arise, and, therefore, knowing what to expect, can help either avoid those mistakes, or better learn from them. Forewarned is forearmed, they say, huh?

What Are Software Development Problems?

Software development problems are attributed to errors in a program or in a system, due to which a program is characterized by unexpected behavior and, consequently, an unpredictable result. Most programming errors arise from developers’ mistakes related to writing a program’s source code or creating its design. Additionally, some errors occur due to incorrect operation of developer tools, for example, in case a compiler generates an incorrect code.

Software errors are localized and eliminated during testing and debugging of the program.

Bug Types in Software Testing

So, here are the types of bugs that you may encounter:

1. Syntax errors

Every programming language such as C, Java, Perl and Python has a specific syntax in which the code is written. When a programmer breaks the “grammar” of the programming language specifications, a syntax error is likely to occur. They result in the program not launching at all, or stopping to operate halfway. Usually, error notifications follow afterwards. These kinds of errors are easily eliminated at the compilation stage.

Have a look at the examples of syntax errors in software programming:

• missing punctuation mark;
• inconsistency of parentheses;
• incorrect forming of an operator;
• incorrect naming of variables;
• incorrect spelling of service words;
• absence of conditions for the end of a cycle;
• lack of array description, etc.

2. Logic errors

These are errors when the syntax itself is correct, although the code is not what you expected it to be, which means that the program operates successfully, but gives incorrect results. They are often more difficult to detect than syntax errors, as there is usually no error notification pointing out to the source of the error.

Logic errors are really tough ones. Perhaps, the toughest of all mistakes. When a program written in any language compiles and runs correctly but produces the wrong output, the flaw is in the underlying programming logic. Coping with logical errors requires a fundamental change in the algorithm.

Here are the examples of logic errors:

• incorrect indication of a branch of the algorithm after checking a certain condition;
• incomplete accounting of possible conditions;
• omission in one or more blocks of the algorithm in the program.

3. Compilation errors

In computer science, compilation refers to the “translation” of a program from a high-level language to a computer-understandable language of zeros and ones. Many types of errors can occur at this stage, including syntax errors. Sometimes, the syntax of the source code may be flawless, but a compilation error can still occur. This may be due to problems in the compiler itself. As a rule, these bugs are fixed at the development stage.

Compilation errors are the easiest ones to deal with, as the syntax of the language is usually strictly formalized, and the location of an error is communicated by a system in a detailed comment. The better the syntax rules of the language are formalized, the more errors can be detected by the compiler and, accordingly, the fewer errors will occur at the next stages.

4. Runtime errors

The program code was compiled successfully and a user is executing your product. Well, here is where runtime errors may come around. They can occur as a result of a crash or a lack of web server resources. Another cause may be the installation of new versions of programs over the already installed ones, which leads to errors in the system registry. In addition, runtime errors are often caused by various viruses deleting or modifying critical files of the operating system. This kind of error is usually fixed by returning to the coding stage.

5. Arithmetic errors

Many programs use numeric variables, and the algorithm may involve several mathematical calculations. Arithmetic errors are logical in their essence and can only be coped with by changing the algorithm. 

The examples of arithmetic errors include, but are limited to:

• incorrect indication of the variable type;
• incorrect determination of the order of actions;
• division by zero;
• extracting the square root of a negative number.

6. Resource errors

A resource error occurs when the program requires more resources than a computer or other media actually have. Resource bugs are difficult to track down because the machine you develop on is often of a higher quality than the servers that run your code on. It is also difficult to simulate real use from a local computer. 

Note these examples of resource errors:

buffer overloads;

use of an uninitialized variable;

access violations;

stack overloads.

7. Interface errors

They can come into being due to a software mismatch with the hardware interface or the application programming interface. In case of web applications, an interface error can be the result of a misuse of the web protocol. Basically, interface mistakes reflect an argument between how you intended the program should be used and how it is actually used. These flaws are often confusing, since may be considered as if they were on your side, when in reality they are on the caller’s side, and vice versa.

Cosmetic Issues in Software Testing

A cosmetic issue is a visual flaw relating to the interface of a program that does not affect the functionality of the application. As a rule, those issues are characterized by low severity. However, the prioritization of cosmetic bugs depends on the situation: usually, such bugs found in the pilot are prioritized higher than those found in the test environment.

Cosmetic issues are more common for projects developed using Agile methodologies. Releases here are produced as sprints. They are usually called a sprint release or simply sr-xx, where “xx” refers to the release number. Therefore, it often happens that clients request many changes in the user interface.

Here are the examples of cosmetic issues / change requests that may arise:

• spelling mistakes;
• grammar mistakes;
• font size;
• colors;
• order of tabs;
• overlapping of texts, etc.

Key Challenges in Software Testing

To produce a high-quality product a developer’s team makes sure it undergoes several rounds of testing. Except for bugs that are detected and need to be fixed, there are other challenges associated with software testing. 

We picked up the most common types of issues in software testing:

• The company doesn’t have a unified center of testing competence, random tools are used;
• No documented testing standards are available; all product and testing departments use their own techniques and follow their own testing practices: developers may store everything on GitLab, while testers – in popular trackers, like Jira, YouTrack and TFS;
• A variety of testing frameworks for popular programming languages or testers’ personal techniques are used;
• Random types of reporting are practiced;
• Diversified ways of checking components and installers are used: testers can install everything manually, or use universal metaranners in TeamCity or in GitLab CI;
• Original types of tests are present: for one project one large mega test is used, followed by integration into the test framework, while for the other tests are extracted into a separate group of check-ups (with possible splitting into load and functional tests).

To avoid testing issues it’s vital to unify everything into a single complex, and generalize the knowledge structure applied by a company’s QA department. 

Debugging Process in Software Development

Debugging is a stage in software development, at which errors, bugs and flaws are detected, localized and eliminated. 

Fun fact, the word “debugging” dates back to 1947, when a bug got into a computer causing it to stop working. 

Debugging consists of three repetitive processes: testing, to find out there is a bug in the system, finding the location of the error in a software program and software documentation, and editing programs, as well as documentation to eliminate the detected error.

The sooner errors are detected or the wrong approach to the implementation of an action is revealed, the less the cost of these errors will be. That’s why the whole debugging process should be well-defined and diligently followed by the team.

In a more detailed paradigm debugging can be broken down into 6 consequent steps:

Problem Identification and Report

First of all, the problem needs to be detected. This is quite an obvious, but very important step, since the wrong interpretation of the bug detected may lead to the waste of time and increased costs.

Here are a few tricks to identify a bug correctly:

• Have a real look at the error. It’s good if you spot the error yourself, but if it comes from a user’s complaint, ask him to provide clear evidence of where the error appeared. For example, a screenshot or a screen recording featuring a functionality problem.
• Reproduce the error. To make sure you identified the existing bug you need to make sure you are able to reproduce it.
• Understand how it affects the system’s behavior. Make a comprehensive analysis of the error’s effects on the system. Talk to the developers, check the related documentation to identify the nature of a problem.
• Validate the identification. Confirm that the error is a thing and needs to be fixed.

Finding the Location of the Error

After the correct identification of the error, it’s time to find it in the code. At this stage we’re not focusing on the big picture, but instead narrowing down your search to small parts of the code to spot the bug.

Defect Analysis

At this step you need to analyze the error in its complexity. The 2 main aspects to focus on are:

• To check that there aren’t any other bugs stemming from or surrounding the initial bug.
• To make sure what are the risks of entering any collateral damage while fixing the bug.

Proving the Analysis

This is an important step when you possibly detect a few more errors in the detected parts of the code and write automated tests for these areas. Once written, you should run them and in case all of them fail, you can ascertain that your analysis is correct.

Defect Resolution by Making Required Changes to the System

At this stage the developers come with a solution to the problem and implement necessary changes to the system. We’re finally there!

Validation of Corrections

Now you can run all the test scripts and make sure they all pass. Hooray! Your program is spotless.

Debugging Techniques in Software Testing

Debugging can be as simple as fixing minor bugs or as tedious as collecting data, conducting comprehensive analysis, and scheduling upgrades.

There are several basic software debugging techniques:

Debugging by Outputting to the Screen

It is a process of considering results of the output commands previously placed in the code to track the stage at which the execution or the value of the variable of interest is stopped. This method is also known as printf-debugging.

Post-Mortem debugging

This kind of debugging occurs after the program execution has been interrupted by an error. It is used to examine the contents of the broken process memory, which is saved automatically by the operating system or by the developer himself using a previously defined instruction.

The “Wolf Fence” Algorithm

It was introduced by Edward Gauss in 1982 and is described in the following way: “Suppose there is a wolf in Alaska, how can we find it? Let’s build a fence in the center of the state, wait until the wolf howls and build another fence in the center of the half from which you can hear the wolf. Let’s repeat this process until the wolf is found”. Thus, when debugging programs, this algorithm comes down to minimizing the code until the error is found.

Delta debugging

This debugging technique is used in automatic testing and is applied by systematically decreasing the input data area until the values causing the software failure are found.

Consequences of Bugs in Software Testing

The consequences of bugs in software may vary from slight, to mild to catastrophic. History has seen dozens of cases when bugs caused enormous financial and even human losses.

Consequences of bugs may be the following:

• Image damages. Defective product offered to clients leaves them unsatisfied or questioned.
• Decrease in customers’ retention. For clients to become loyal to your product it has to be flawless and stand out from the competition. In a fast-moving dynamic environment, finding a fair alternative to your product in case it doesn’t meet users’ expectations is as easy as to post a photo in a social media network: a couple of clicks and you’re there.
• Reduced revenues and increased costs. It requires additional cash infusion to fix and maintain a defective product. Also, because users may switch to another product you’re losing additional income streams.

And now to the historical discourse:

• A small bug in General Electric Energy’s equipment monitoring software has left 55 million people without electricity. Residential buildings, schools, hospitals, airports were de-energized on the East Coast of the United States in 2003.
• The Therac-25 is one of the saddest cases in the history of medical device programming. Due to the race condition error, when the device quickly switched between magnetic and X-ray modes, the X-ray shutter did not have time to set up. Because of this, 10 patients were diagnosed with radiation sickness, which led to the death or amputation of the affected body parts.
• On October 7, 2008, a plane on a Singapore-Perth flight suddenly began to shake. After 3 hours of flight, the autopilot suddenly turned off, and the pilots were bombarded with errors. After 5 minutes, the aircraft’s captain turned off the autopilot and tried to figure out what had happened. Suddenly, the nose of the plane tilted 8.4 degrees, from which the passengers in the cabin, who were not wearing their seatbelt at that moment, crashed their heads against the ceiling. As a result 112 people were injured.
• In 2011, more than 400 prisoners were released too early from California prisons due to a software bug. The state was trying to free up some space in prisons. A script was launched to select the most “innocent” criminals, setting a “danger rating”. Since many of the prisoners were in various gangs or drug cartels, there was little information about them, therefore the script defined that the prisoner was not dangerous. Therefore, 450 dangerous criminals were released earlier than they were actually supposed to.
• Equifax Hack is the biggest loss of personal data. A bug has been discovered in the Apache Struts framework. The bug led to the fact that it was possible to carry out the so-called RCE attacks (remote code execution). Immediately after the fix, hackers began bombing sites on the Internet in search of admins who had forgotten to update the Struts version. The Equifax hack was done due to this bug.

Software Testing Problems and Solutions by JoinSoft

JoinSoft offers a full cycle of Quality Assessment and Software Testing services, from requirements testing and bugs discovery to pre-release testing and testing during deployment. 

We are a team of dedicated professionals with 5+ years of experience and passions for our work. We have a transparent workflow and involve you in each stage of the testing process. We always keep in touch with our clients to make sure they know how their project is doing.

We offer both manual and automated testing, as well as performance, functional, usability and security testing.

We apply an individual approach to each client’s project and use Agile methodologies and metrics for an efficient software testing process.

Final Thoughts About Open Issues in Software Testing

There are various types of software errors that can occur during the software development process and every programmer should be aware of them. Developing a system without a single mistake is, probably, impossible. However, the more complete the test coverage of the system is, the higher is the probability of detecting and fixing functional bugs. Besides, knowing the types of bugs that may come into being makes it easier to battle those issues. That’s what we are here for: 

1. To equip you with comprehensive information on various types of issues.
2. To provide high-quality services to help you create a product of your dreams!

Learn more about our Quality Assurance and Software Testing services and see you on the other side!