- Home
- /
- CS Thinking
- /
- Mistakes
- /
- Systems, Networks & Impact
Common Mistakes in Systems, Networks & Impact
64 common pitfalls across 16 concepts, including 16 recurring stuck points
These are the pitfalls students hit across all 16 systems, networks & impact concepts, from Hardware & Software, Computing System and Operating System onward. Each entry names the concept it comes from, so you can tell a slip from a genuine gap and go straight to the explanation.
Firmware blurs the line — it's software permanently stored on hardware (like a router's built-in program).
Thinking hardware and software can function independently—they are useless without each other
Confusing firmware with hardware because it is embedded in a physical device
Assuming hardware problems can be fixed with software updates or vice versa
A computing system isn't just a desktop computer — phones, cars, smart thermostats, and game consoles are all computing systems.
Thinking computing systems are only desktop computers or laptops—embedded systems in cars, appliances, and medical devices are also computing systems
Ignoring the data component—hardware and software alone are not a complete computing system without data to process
Assuming all computing systems have screens and keyboards—many have specialized sensors and actuators instead
Apps run on top of the OS, not instead of it. Closing all apps doesn't shut down the OS.
Confusing the operating system with applications—the OS runs underneath all apps and manages them
Thinking closing all apps shuts down the operating system, which continues running in the background
Not updating the OS regularly, which leaves known security vulnerabilities unpatched
Storage is not the same as memory. Memory is temporary working space; storage is long-term data keeping.
Confusing storage capacity with RAM
Ignoring file size growth for images, audio, or video
Assuming deleting an app always removes all of its saved data
The internet is a network of networks, not a single thing. No one entity owns or controls it all.
Confusing the internet (global infrastructure) with the World Wide Web (websites accessed via browsers)
Assuming Wi-Fi and internet are the same—Wi-Fi is a local wireless connection that may or may not have internet access
Thinking a network requires physical cables—wireless, satellite, and cellular connections also form networks
The internet is not the same as the World Wide Web. The web is one service that runs on the internet (alongside email, streaming, etc.).
Confusing the internet (the network infrastructure) with the World Wide Web (websites and browsers that use the internet)
Assuming the internet has a central server or single point of control—it is fundamentally decentralized
Believing that data travels directly from sender to receiver—it actually passes through many intermediate routers
Packets from the same message can take different routes and arrive out of order — the receiving device reassembles them.
Assuming packets from the same message always travel the same route—they can take different paths and arrive out of order
Forgetting that packets have a maximum size (MTU), so large messages must be split into multiple packets
Confusing packet loss with a broken connection—protocols like TCP automatically detect and retransmit lost packets
A protocol is not software — it's a specification (set of rules). Software implements protocols.
Confusing a protocol (the specification or rules) with the software that implements it
Not understanding that protocols are layered—HTTP usually depends on TCP (HTTP/3 uses QUIC over UDP instead), which depends on IP, which depends on physical links
Assuming all network communication uses the same protocol—different applications use different protocols (HTTP for web, SMTP for email, FTP for file transfer)
No system is 100% secure. Cybersecurity is about managing risk, not eliminating it completely.
Assuming a single strong password is sufficient—multi-factor authentication and regular updates are equally important
Believing only large companies are targets—individuals and small organizations are frequently attacked
Thinking cybersecurity is purely a technology problem—human error (clicking phishing links, weak passwords) is the leading cause of breaches
'I have nothing to hide' misses the point — privacy is about control over your information, not about having secrets.
Accepting all permissions and cookie consents without reading what data is being collected
Assuming 'I have nothing to hide' means privacy does not matter—privacy is about control, not secrecy
Sharing personal information on social media without considering that it may be permanent and publicly accessible
Encryption protects data from reading, but it does not automatically prove the sender is trustworthy or the device is safe.
Thinking encryption only matters for governments or banks instead of everyday users
Assuming encrypted data is still readable without the correct key
Believing encryption solves every security problem on its own
Open source doesn't mean 'no rules.' Open source licenses specify exactly how you can use and share the code.
Assuming anything found online is free to use—most content is copyrighted by default
Thinking open source means 'no restrictions'—each open source license has specific terms for attribution, modification, and redistribution
Confusing fair use with unlimited use—fair use is a narrow legal exception with specific criteria
Accessibility isn't an afterthought or a bonus feature — it should be designed in from the start.
Treating accessibility as an afterthought to add at the end instead of designing for it from the start
Only considering visual disabilities—accessibility also covers hearing, motor, cognitive, and neurological disabilities
Assuming accessibility only benefits disabled users—features like captions, voice control, and responsive design improve the experience for everyone
Not every problem parallelizes well. Some steps still have to happen in sequence.
Assuming more processors always produce proportional speedup
Ignoring the overhead of coordination and communication
Trying to parallelize steps that depend heavily on each other
AI is not magic or general human intelligence. Most school-level AI systems are specialized tools trained for a narrow task.
Assuming AI outputs are always correct or unbiased
Ignoring the quality and bias of the training data
Treating AI as one single technology instead of many different methods and systems
Ethics isn't just about laws — something can be legal but still unethical. Ethics requires judgment, not just compliance.
Assuming that if something is legal it is automatically ethical—legality and ethics are different standards
Ignoring algorithmic bias by assuming that automated systems are inherently objective and fair
Believing ethics is only the developer's concern—users, companies, and policymakers all share ethical responsibility