XM: A Customised Software Tool to Facilitate Learning and Professional

Skills Development in Operating Systems

Alina-Delia C

alin

, Rares

Florin Boian

and Horea-Bogdan Mures

Department of Computer Science, Babes¸-Bolyai University, 1 M. Kog

alniceanu Street, 400084, Cluj-Napoca, Romania

Keywords:

Operating Systems, UNIX, Teaching Tools, Evaluation Tools, Class Management, Cheating Detection.

Abstract:

In this paper, we present the XM tool and a novel perspective on teaching Operating Systems as an under-

graduate course, focused towards the development of professional skills. We propose a modular content course

plan and a suite of software tools that are speciﬁcally designed to support academic activities that include

learning, assessment, cheating detection, and attendance management. With this approach, we are able to

provide students with access to a better educational environment that will support their learning experience

and prepare them for real-life industry scenarios. For teachers, it provides more efﬁcient class management

for grading and attendance tracking, allowing more time to spend on personalised feedback for students.

1 INTRODUCTION

From textbook content to online resources used in

teaching Operating Systems (OS), most course ma-

terials are highly theoretical and abstract (Kankuzi,

2019). While it is a fundamental component of all

undergraduate computer science degree programmes,

the allocated time and resources allow for a deep

and detailed study of only the most important con-

cepts. Expanding on this requires approaches and

tools that support the demonstration of OS concepts

in the context of a real-world, live scenario, such as

using command-line tools for ﬁle system navigation,

compilers, but also managing processes, resources

and users (Vavreckov

a, 2020).

In this paper, we propose a dynamic approach to

teaching OS by utilising a suite of software tools de-

signed to offer a great hands-on learning experience.

This is performed with the help of live programming

and interaction with the Unix operating system, all

the while improving the detection and prevention of

cheating, even in online environments. To evalu-

ate the impact of the customised software system on

learning and teaching, we formulate and answer three

speciﬁc research questions, presented in Section 3.

https://orcid.org/0000-0001-7363-4934

https://orcid.org/0000-0002-0993-8842

https://orcid.org/0000-0003-4777-7821

2 BACKGROUND

In terms of teaching operating systems, there are

many attempts to shift the highly theoretically ori-

ented courses towards a more practical approach, us-

ing tools and focusing on understanding the OS con-

cepts in a hands-on experience. A few decades back,

the various operating systems used were much sim-

pler, and the OS course was oriented towards the

senior-level students or graduates. However, in time

things became more complex and with multiple types

and generations of systems, the theoretical concepts

became more abstract and disjoint from the experien-

tial part, with heavy content (Weina et al., 2018).

In this context, studies (Weina et al., 2018) em-

phasise that students often have difﬁculties linking

this course to others, as well as replicating in experi-

ments some key concepts (such as memory manage-

ment, concurrent behaviour). To solve these issues,

they use a related thinking principle of teaching, pro-

viding a simple OS to demonstrate aspects, invite stu-

dents to learn through experiments, and make associa-

tions. Another paper (O’Brien, 2017) describes an ap-

proach using the Linux SystemTap monitoring tool to

capture and illustrate speciﬁc operating system con-

cepts such as scheduling, ﬁle system implementation,

and memory management. This tool offers an addi-

tional simple and valuable resource to solidify the un-

derstanding of theoretical concepts.

Virtual machines have also been widely used to

provide a safe learning environment to students in OS

330

alin, A., Boian, R. and Mure¸san, H.

XM: A Customised Software Tool to Facilitate Learning and Professional Skills Development in Operating Systems.

DOI: 10.5220/0012590300003693

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 16th International Conference on Computer Supported Education (CSEDU 2024) - Volume 1, pages 330-337

ISBN: 978-989-758-697-2; ISSN: 2184-5026

courses. Using Linux as the production operating sys-

tem and VMware as the platform, a virtual kernel de-

velopment lab was created, covering useful subjects

including POSIX Threads, ext3 ﬁle system, and ker-

nel patch application (Nieh and Vaill, 2006). The

EduFUSE tool was developed, extending an existing

framework to support both the C and Java languages

(Thompson et al., 2020), compatible with Linux and

MacOS, to study and learn about ﬁle systems.

Recent approaches in computer science are based

on browser client-side web technologies, with almost

any programming language supported, even the ba-

sic Linux commands (Sharrock et al., 2022). While

they provide simple access to a practice terminal, they

skip on the very tedious process of setting up the re-

quired environment. Also, the problem of cheating

is not properly addressed in this case. Bailey et al.

present a tool named uAssign, aimed at teaching and

self-assessing Unix command line skills, based on a

web-embedded terminal (Bailey and Zilles, 2019).

Another aspect of teaching is related to class

management, student assessment, and grading, in

which cheating prevention is a major topic of inter-

est (Vykopal et al., 2022). This led to the devel-

opment of an Automatic Problem Generation soft-

ware (Fowler and Zilles, 2021) for creating person-

alised tasks for each student (to prevent cheating) for

a hands-on computer lab environment, for a cyberse-

curity course. It involves creating arbitrary networks

and virtual machines and then extending the tool to an

open-source platform of exercises and competitions.

3 PROPOSED COURSE

APPROACH. CASE STUDY

The Operating Systems course is part of the ﬁrst year,

second semester bachelor’s degree curricula for the

Computer Science study undergraduate programme

from a higher education institution. It is a mandatory

course in the 3-year computer science programme of-

fered by the university. The total number of students

enrolled each year is around 400, of which half study

in their native language programme and half in the

English teaching programme. There are no manda-

tory prerequisites, but basic knowledge of C program-

ming is recommended (and gaps are being addressed

in the course planning and laboratory practice).

The objectives of the discipline are focused on

learning the basics of Unix commands, shell pro-

gramming, and the standard C system API. Students

will learn to interact with an operating system be-

yond the standard user interface, using the com-

mand line. At the end, the students will have a

good understanding of processes (scheduling, concur-

rency, synchronisation, deadlocks), memory manage-

ment (allocation, loading, placement, replacement,

caching), ﬁle system organisation (including data cor-

ruption prevention, existing ﬁle systems and their

applicability), Linux command line (frequently used

commands, Sh/Bash programming), C ﬁle API (pro-

cess creation and inter-process communication API),

POSIX threads, and synchronisation mechanisms.

In this context, we introduce a set of software

tools, of which the XM tool is speciﬁcally built to ad-

dress the requirements and objectives of the Operating

Systems course, from the perspective of both students

and teachers. We assess the impact of our approach,

by answering the following research questions (RQ):

• RQ1: Does XM simplify grading and attendance

management for the teacher?

• RQ2: Does our approach facilitate the learning of

practical skills for industry by students?

• RQ3: What is the impact of our cheating detection

mechanism?

3.1 Modular Approach of the Course

Contents

The course structure and syllabus were developed in

line with the initial aims of putting practical skills and

competencies ﬁrst and theory second, and willing to

keep students engaged throughout the entire semester.

In terms of selecting and planning the course con-

tent for seminaries and laboratories, we used sprints

of 2-3 weeks for each chapter, followed by a forma-

tive evaluation, as described in Table 1. The lectures

follow a similar planning, providing the initial intro-

duction to these concepts (which are then detailed

and put into practice at the seminar and lab) and then

moving on to detailed theory about operating systems

starting from weeks 10-11 onwards.

In addition, many chapters are based on previ-

ously acquired knowledge and competencies, thus

this approach ensures greater participation of stu-

dents. Each chapter will end with a practical labo-

ratory evaluation in which students must demonstrate

their acquired skills and be individually assessed and

supported early on if they do not have the minimum

required skills to participate in the next chapter.

Continuous evaluation in the form of short tests

that measure basic learnt skills is more useful to both

students and teachers in adapting the rest of the con-

tent based on what was learnt and what has not yet

been handled well enough. In this sense, planning can

undergo slight changes, and the teacher can choose to

include more examples of gradual difﬁculty.

XM: A Customised Software Tool to Facilitate Learning and Professional Skills Development in Operating Systems

331

Table 1: Course contents by modules.

Chapter

sprint

Contents for seminar and laboratory

Weeks 1-3: - UNIX command line introduction. C pro-

gramming in the UNIX command line -

gcc, system functions, system manual. De-

tecting memory problems with valgrind.

Evaluation: Test on server (TC): debug a C program.

Weeks 3-5: - Regular expressions. Basics of grep, sed,

awk, ﬁnd, sort, uniq, cut.

Evaluation: Test on server (TGSA): solve a text pro-

cessing task using regex and several unix

commands.

Weeks 6-8: - UNIX Shell programming - basic sys-

tax, I/O, redirects, commands combina-

tion. Permissions, use of /etc/passwd and

other system ﬁles.

Evaluation: Test on server (ESh): write a Shell pro-

gram to solve system speciﬁc tasks (in-

volves ﬁles, directories, processes, users).

Weeks

9-11:

- UNIX Processes: fork, wait, exec, signal,

kill, exec. UNIX Signals. IPC: pipe, FIFO,

dup2, popen.

Evaluation: Test on server (EPr): write a C program to

solve a problem that use inter-process com-

munication between several processes.

Weeks 12-

14:

- POSIX threads, mutexes, RW Locks,

conditional variables, semaphores, barri-

ers. Programming examples with synchro-

nisation and deadlock prevention.

Evaluation: Test on server (ETh): write a C program to

model and solve a problem using threads

and efﬁcient synchronisation mechanisms.

3.2 Software Tools

Our platform of choice for teaching the OS course is

Linux, which is historically the platform of choice of

numerous schools teaching the same class. We also

chose Linux due to its pervasiveness in the production

systems used by the local industry, and its accessibil-

ity to students for home practice. Besides connect-

ing over SSH to a remote Linux server, students with

Windows computers can use Windows Subsystem for

Linux while students with Apple computers already

have in MacOS, a Unix based OS.

One of the course goals is to teach the students

to use an operating system beyond the usual visual

UI, in other words, we want them to get familiar

with the command line and learn to harness its power.

To achieve this goal, we chose to organise the entire

course around a Linux VM (2 vCPUs, 8GB RAM,

20GB SSD), to which the students would connect via

SSH and “spend the entire semester in the command

line”. As this makes for a great learning environ-

ment where students can have a hands-on experience

with a multi-user UNIX environment, it became ap-

parent that both teachers and students would beneﬁt

if it would also provide means for the other activities

involved in teaching: evaluation, grading, attendance

taking, proctoring, automatic grade calculation, stu-

dent status/grade reporting, etc. This led to the design

and development of a software tool named XM meant

for examination, but was later extended with all the

features listed above in year 2020.

3.2.1 The XM Tool

The guiding principles for the development of the XM

tool as a support for teachers and students were:

• Purely command line to promote learning of prac-

tical programming skills for students

• Assist teachers with class management to reduce

the time spent on non-learning tasks (e.g. atten-

dance taking, problem assignment, grading)

• Increase the difﬁculty of cheating and provide a

mechanism for detecting cheating attempts

The tool was developed on the Linux platform, to

fully support functionalities required for semester ac-

tivities such as attendance, evaluation and grading, in

a safe and cheating proof environment (see Figure 1).

The implementation language used is Go, together

with an SQLite database and is compatible with any

Linux distribution, as it makes use of automated user

account generation and management features. The

development principles are based on the Linux phi-

losophy ”write programmes that do one thing and do

it well” (Raymond, 2003) and rapid development, as

a user-friendly command line tool.

3.2.2 Managing Student Assessment Activities

Figure 1: Architecture of the XM Tool.

CSEDU 2024 - 16th International Conference on Computer Supported Education

332

The XM design for evaluation and cheating preven-

tion is based on the use of one-time operating system

user accounts. Students do not have personal accounts

on the server; a new account is created for any activ-

ity requiring server access. The connection creden-

tials are sent by email to the students and the accounts

are locked after the activity is completed. The one-

time account acts as a frozen persistent reference of

the student’s work, allowing for grading and review-

ing/regrading of their solutions.

The accounts are conﬁgured at the time of creation

to record all user activity by starting upon every login

using the asciinema tool (Kulik, 2022). The console

activity recording can be replayed (Hassan, 2020b)

later on for ensuring authentic practical assessments

(Hassan, 2020a) as it allows easy detection of code

pasting or attempts to fetch material from outside the

examination account. During the examination, the in-

structor can monitor the student activity as a summary

of processes and connections, or can see details and

act using super-user privileges.

In order to assign evaluation sets for an activity

that will be graded, the instructor can attach the set(s)

when the room for that activity is created (Figure 2)

or at a later point (Figure 3).

Figure 2: Create room.

Figure 3: Add evaluation type to the room.

The evaluation sets are stored in a predeﬁned

folder structure where only instructors have access.

The XM tool, for each account linked with the activ-

ity, will randomly choose one problem from each of

the speciﬁed evaluation sets and will copy them into

the home folder of the account, one ﬁle per problem.

Upon login, the account owner will be able to access

the ﬁles containing the problem statements and solve

them. This way we ensure each student has a unique

time-limited task to prevent copying solutions.

3.2.3 Attendance Management

Some of the course activities such as seminars and

laboratories require attendance taking as faculty reg-

ulations impose a minimum attendance without which

the class is automatically failed. The students partici-

pating in a course activity are added in an XM room,

which acts as an attendance record. Adding the stu-

dents to a room has been designed to allow the pro-

fessor to quickly select the students from previous ac-

tivities, or search for new students by name or ID. In

case of online teaching, a simple copy-paste of the at-

tendance list from the conferencing application (e.g.

Microsoft Teams, Google Meet, Jitsi, etc) into XM

will result in XM tool matching the students by name

and add them automatically to the room and asking

for user input in cases that it cannot solve, as shown

in Figure 4. The instructor does not have to memo-

rise the command line options as XM provides Bash

auto-completion, and for every missing argument it

presents the user with a single/multi-choice interac-

tive option selection, as seen in the last three lines of

Figure 4. With this approach, attendance taking for a

seminar of 30 students is much faster.

Figure 4: Performing attendance.

3.2.4 Facilitate Grading Activity

The grading is performed by entering each one-time

student account, evaluating the solution, and then as-

signing a grade. To simplify the navigation between

accounts and grade assignment, XM provides room-

wise (often used in practical evaluation) and subject-

wise grading features. The subject-wise grading is

used for the written summative evaluations which

consist of 10 open-answer questions assigned to 150-

200 students at a time. Both approaches allow the

examiner to navigate among the student solutions in a

next/previous manner.

Grading consistency and efﬁciency are much eas-

ier achieved if the evaluator does not need to switch

from one problem to another, but rather focuses on

all solutions for the same problem. Reducing the ex-

aminer’s context switching when grading the practi-

cal evaluations does not have much impact, since the

number of distinct problems is small. For the summa-

tive evaluation, context switching minimisation is sig-

niﬁcantly more important as evaluators prefer to grade

consecutively 150 solutions to the same problem be-

fore changing context (problem statement), instead of

grading each student’s 10 solutions to the 10 different

problems before moving to another student. In Figure

5 is a sample grading session where the evaluator as-

XM: A Customised Software Tool to Facilitate Learning and Professional Skills Development in Operating Systems

333

signs a grade, navigates to the next student, and there

reconsiders the grade and changes it for a higher one.

Figure 5: Automating the grading activity.

The greatest impact is in the evaluation of the ﬁnal

written exam, allowing the grading of 400 students at

the same time, by subject, by 10 teachers, in less than

2 hours. In the traditional context of a paper exam,

with the same exam structure, it would have taken 6

hours for each of the 10 teachers, plus waiting times

for synchronisation among evaluators.

3.2.5 Student Feedback and Status Reporting

The ﬁnal grade calculation and are done using a cus-

tom course-speciﬁc algorithm that yields a detailed

report of the student activity, which can be emailed

to students. In Figure 6 is a truncated sample report.

Sa and La are seminar and lab attendances, the grades

starting with T refer to quick tests, while those that

start with E are practical exam grades. The Ew01,

Ew02, ... are the grades for each of the 10 writ-

ten exam problems, and Ew is the cumulative written

exam grade. This detailed report resulted in a signiﬁ-

cant decrease in the number of contested ﬁnal grades.

3.3 Practical and Theoretical Student

Assessment

Formative evaluation allows the teacher to better ex-

plain the objectives and set the expectations, while

keeping students engaged and motivated by providing

feedback directly and more often. It is used in mon-

Figure 6: Computing the ﬁnal grade - detailed display.

itoring, sharing learning goals, and addressing mis-

conceptions (Grover, 2021).

As previously described, in our course planning

each chapter ends with a practical lab test. Each stu-

dent receives a unique account on the server and is

randomly assigned a task by the server XM tool to

be completed within a time-frame. Students will be

involved in doing live programming with command

line tools and text editors, and their server sessions

are recorded to prevent copy-paste or download activ-

ities. The teacher monitors the overall activity during

the exam: active sessions, running processes, and IP

connections, to prevent remote cheating connections

or server overload with processes.

Evaluation of these tasks is performed in the form

of a discussion with each student based on the solu-

tion they developed for the unique task assigned on

the server, randomised by the XM tool. This is per-

formed during the lab classes. Besides grading the ﬁ-

nal solution, teachers can identify common mistakes,

knowledge gaps, and acquired skills, and then dis-

cuss them with students, providing clariﬁcations. The

grading criteria are very strict, meaning that a code

with syntax errors is automatically graded with 1, a

program with a severe run-time crash is graded with

4, while 5 is the minimum passing requirement. The

maximum grade is 10, and it is obtained if the pro-

vided solution correctly addresses the given task in an

efﬁcient way (with respect to memory allocation, er-

ror checking, and data validation). We also discuss

code style, alternative solutions and limitations.

The summative theoretical evaluation consists of

a written examination and uses also the exam server,

but the tasks are more theoretical. Students are not

required a complete program, but a written response

related to a speciﬁc code or OS context, while still

using the skills developed (command line, text editor).

CSEDU 2024 - 16th International Conference on Computer Supported Education

334

4 SYSTEM EVALUATION AND

DISCUSSION

RQ1: Does XM Simplify Grading and Attendance

Management for the Teacher?

In order to assess the effectiveness of the XM tool

in providing a more efﬁcient management of teach-

ing related activities, the people involved in teach-

ing the OS lecture, seminar and laboratory activities

during the analysed period were required to answer

a questionnaire, as described in Table 2. The sur-

vey contained additional ﬁelds on the characteristics

of the respondents, such as gender and age. It is im-

portant to note that the XM tool is designed for OS

teachers speciﬁcally, who are proﬁcient in the use of

command-line tools, and the assessment of usability

is performed from this perspective.

Table 2: Teacher survey questions.

Item Question

1 In which year(s) did you teach OS?

2 To what extent did the XM tool reduce the time

you spent for attendance taking?

3 How did you perform attendance previously?

4 To what extent did the XM tool reduce the time

you spent for grading the written exam?

5 How did you perform grading previously?

6 How do you appreciate the system usability?

7 State your reason for the usability assessment.

Questionnaire items are mostly closed questions.

Item 1 is a multiple selection from a list of years (the

2016-2022 range). Items 2 and 4 require a single-

choice answer from a list of 4 options: None, 10%-

25%, 25%-50%, 50%-75%. Item 6 is a single choice

of a ﬁve-scale quality assessment: Excellent, Good,

Average, Poor, Very Poor. Items 3, 5 and 7 are open

questions to provide a reference point for the corre-

sponding previous closed questions.

Most respondents are male, aged 30-49, and have

been involved in teaching the OS course, seminar or

laboratory over several years, before and after the in-

troduction of the XM tool in 2020. In terms of as-

sessing the reduction in time spent with attendance

taking, 44% of the respondents estimate a reduction

with 25%-50%, while 56% of them estimate a reduc-

tion with 50%-70%. The alternative methods for at-

tendance taking used as comparison include: atten-

dance paper manually transcribed in a spread sheet,

attendance sheet/electronic spreadsheet, or manually

in predeﬁned lists (students signed their attendance on

paper then a spreadsheet is completed).

Regarding the impact of the XM tool in reduc-

ing time spent with grading, 89% of teachers chose

a time reduction of 50%-70% , whereas 11% of re-

spondents chose a reduction of 25%-50%. The pre-

vious grading methods for the written exam involve:

paper exam grading (sometimes sorting the papers by

subject), and the use of spreadsheets.

The usability of the XM tool is appreciated to be

Excellent by 89% of the respondents and Good by

11%. The respondents considering that the XM tool

interface is intuitive, easy to learn and use, similar to

other command line tools. It is also appreciated be-

cause it facilitates cheating detection, speeds up at-

tendance tracking and grading, and provides greater

ﬁdelity in the correctness of assessments.

In conclusion, XM is an easy to use command line

tool that greatly simpliﬁes the class management ac-

tivities involving attendance and grading by reducing

the time required, as opposed to classical approaches.

RQ2: Does Our Approach Facilitate the Learning

of Practical Skills for Industry by Students?

To evaluate the impact of this course as it is now struc-

tured and how students are receiving it, we have anal-

ysed the grades and collected student feedback.

Figure 7: Evolution of the average written exam grade com-

pared to the average of all practical exams grades.

We have analysed the grades between 2016 and

2022, to observe their evolution as teaching approach

and tools changed. The year 2020 was omitted from

these statistics, since due to the onset of Covid-19,

the examination structure was signiﬁcantly changed

from the norm. We compared the grades obtained at

the written examination and the grades obtained at the

practical examinations (i.e. a comparison of theoret-

ical knowledge vs. practical skills/knowledge). The

results are presented in Figure 7 and Table 3.

We also studied the correlation between the grades

of the written exam and the average of each practical

exam for each student per year (Figure 8 presents the

analysis for year 2022). In years 2016, 2018 and 2021

we have observed low values for R

: 0.232, 0.229

and 0.223, respectively. In 2021, due to COVID-19,

all exams were held remote and online, and there was

an increase in the number of grades between 8 and

10, potentially skewing the analysis. For years 2016

and 2018 the primary explanation is that there was a

discrepancy in the difﬁculty of the practical examina-

tions and the written examination. In the other years,

2017, 2019 and 2022, with R

values of 0.406, 0.325

XM: A Customised Software Tool to Facilitate Learning and Professional Skills Development in Operating Systems

335

and 0.3290 respectively, there is a better correlation

between the practical and theoretical examinations.

This indicates that the evaluation methods that were

employed gauge the level of understanding more ac-

curately than in other years.

Table 3: Average of Written Exam Grades (Avg.E) and

Practical Exam Grades (Avg.P).

Year Avg.E Abs(Avg.E-Avg.P) Avg. P

2016 4.87 2.78 7.40

2017 5.40 2.18 7.34

2018 5.78 2.12 7.62

2019 5.67 1.94 7.39

2021 7.65 1.27 7.56

2022 7.27 1.51 6.65

Figure 8: Correlation between theoretical exam grade and

average of practical exams grade in year 2022.

We have collected two types of feedback: (a) at

the lab and seminar in 2021 and 2022 using Microsoft

Forms, with a response of 20% of the students, and (b)

the overall course feedback collected in 2022 with 53

responses (meaning 13% of total enrolled students),

rating the overall experience of Very Good (38), Good

(14) and Bad (1).

Regarding the open questions from both (a) and

(b), we have analysed the feedback and grouped sim-

ilar responses. Although some experiences differ in

terms of how labs have been performed (”Teaching

method was very interactive”, ”I did not like there was

difference in difﬁculty of lab tests depending on each

teacher”, ”So i liked the modern overall approach:

everything was done in class no homework and in-

teractive seminars and labs”, ”Teachers happy to ex-

plain and interact during the lab”, ”tests graded on

the spot”). It is almost unanimous that the students

found tests hard and, for that reason, an important

motivator to learn and stay engaged throughout the

semester: ”The grading system ensured you could

pass as long as you had some practice”, ”The tests

really stimulated me to learn”, ”Having multiple ex-

ams/tests meant there was less material to study for

the written exam”, ”The topics at the end of the course

are too theoretical”, ”The evaluations were useful be-

cause it pressured me to not leave all the material for

the exam session”. Most students have stated that they

ﬁnd the subject interesting, that they have a deeper un-

derstanding of operating systems, and that they feel

that they will deﬁnitely use the concepts they have

learned later in their career: ”Linux is cool, and I feel

like I’ll always use the things I learned during this

course in some form or another!”, ”I liked a lot of the

things we studied, like how servers work, how mem-

ory is managed, how to program in Linux”.

In conclusion, the course method gave students a

challenge in a setting they thought would be helpful

for their future careers while also involving them in

the acquisition of practical skills.

RQ3: What Is the Impact of Our Cheating Detec-

tion Mechanism?

One crucial component of the toolkit set-up was

cheating detection, which is based on session record-

ings that can be checked for copy-and-paste-style ac-

tions. The total number of such cases identiﬁed

(graded with 0) for the ﬁve practical tests conducted

throughout the year are: 119 in 2021 and 153 in 2022.

In Figure 9, we have analysed the percentage of de-

tected cheating instances each year, before and after

the introduction of our monitoring tool in 2020.

Figure 9: Percentage of cheating for each academic year,

calculated as no. of grades 0 from the total no. of grades.

We can observe in Figure 9 that the percentage of

cheating that has been identiﬁed has decreased start-

ing 2018. We note that there is no visible difference

after adding our new approach to detect unauthorised

copy-paste actions, whereas previously this was done

solely through observations and students’ ability to

explain their answer.

In conclusion, our approach can detect cheat-

ing that involves copy-paste of code from external

sources. The percentage of cheating students is not

inﬂuenced by the introduction of such a system.

5 CONCLUSIONS

In this article, we present a novel approach and sev-

eral software tools designed to create an interactive

and practical laboratory experience for students tak-

CSEDU 2024 - 16th International Conference on Computer Supported Education

336

ing the undergraduate Operating Systems course. The

XM tool is a command line programme designed to

facilitate teaching activities for both teachers and stu-

dents. The methodology has been put into practice

and reﬁned during the last ﬁve academic years, with a

focus on developing professional and soft skills. We

have presented the tools and their role in keeping stu-

dents engaged throughout the semester by providing

ongoing feedback for each practical test.

The outcomes of this experience from both the

teacher’s and students’ perspectives show increased

satisfaction and conﬁdence in students’ ability to use

the concepts learnt forward, in industry. From the

teacher’s perspective, the system provides more efﬁ-

cient class management (in terms of attendance and

grading). The objective evaluation also shows the

high-quality practical skills achieved by the students.

Moreover, with time and adjustments each year, there

is a better correlation between practical and theoreti-

cal knowledge and skills, as shown by the higher cor-

relation of grades obtained during the semester and at

the end-of-term written exam. However, this needs to

be closely monitored and ﬁne-tuned in the future. We

note also the system’s capability to detect cheating,

and this knowledge discourages students, to some ex-

tent, from attempting copy-paste of external code.

As future work we aim improvements to the XM

software related to plagiarism detection, and to inte-

grate activity patterns detection using artiﬁcial intelli-

gence on usage data, to identify drop-out risk.

ACKNOWLEDGEMENTS

The publication of this article was supported by the

2023 Development Fund of the UBB.

REFERENCES

Bailey, J. and Zilles, C. (2019). Uassign: Scalable inter-

active activities for teaching the unix terminal. In

Proceedings of the 50th ACM Technical Symposium

on Computer Science Education, SIGCSE ’19, page

70–76, New York, NY, USA. Association for Com-

puting Machinery.

Fowler, M. and Zilles, C. (2021). Superﬁcial code-guise:

Investigating the impact of surface feature changes on

students’ programming question scores. In Proceed-

ings of the 52nd ACM Technical Symposium on Com-

puter Science Education, pages 3–9, New York, NY,

USA. Association for Computing Machinery.

Grover, S. (2021). Toward a framework for formative as-

sessment of conceptual learning in k-12 computer sci-

ence classrooms. In Proceedings of the 52nd ACM

Technical Symposium on Computer Science Educa-

tion, pages 31–37, New York, NY, USA. Association

for Computing Machinery.

Hassan, I. (2020a). Automated authentic assessment: Ap-

plied to an undergraduate course in network and sys-

tem administration. In 2020 IEEE Global Engineering

Education Conference (EDUCON), pages 183–191,

Porto, Portugal. IEEE.

Hassan, I. (2020b). Teaching cybersecurity to computer

science students utilizing terminal sessions recording

software as a pedagogical tool. In 2020 IEEE Fron-

tiers in Education Conference (FIE), pages 1–8, Upp-

sala, Sweden. IEEE.

Kankuzi, B. (2019). Balancing theory and practice in an

introductory operating systems course. In ICT Educa-

tion: 47th Annual Conference of the Southern African

Computer Lecturers’ Association, SACLA 2018, Gor-

don’s Bay, South Africa, June 18–20, 2018, Revised

Selected Papers 47, pages 362–375. Springer.

Kulik, M. (2022). Asciinema. last accessed on August 8,

2022.

Nieh, J. and Vaill, C. (2006). Experiences teaching op-

erating systems using virtual platforms and linux.

SIGOPS Oper. Syst. Rev., 40(2):100–104.

O’Brien, D. (2017). Teaching operating systems concepts

with systemtap. In Proceedings of the 2017 ACM Con-

ference on Innovation and Technology in Computer

Science Education, pages 335–340, New York, NY,

USA. Association for Computing Machinery.

Raymond, E. S. (2003). The art of Unix programming.

Addison-Wesley Professional, Boston.

Sharrock, R., Collin, E., Labat, T., Hamonic, E., Bonfert-

Taylor, P., and Goudzwaard, M. (2022). Teaching and

learning programming with linux using in-browser

client-side web technologies: Exploring the key fea-

tures for achieving systems and tools scalability. In

Proceedings of the Ninth ACM Conference on Learn-

ing @ Scale, L@S ’22, page 427–430, New York, NY,

USA. Association for Computing Machinery.

Thompson, L., Clarke, J., and Sheehan, R. (2020). Edufuse

a visualizer for user-space ﬁle systems. In Proceed-

ings of the 2020 ACM Conference on Innovation and

Technology in Computer Science Education, ITiCSE

’20, page 549–550, New York, NY, USA. Association

for Computing Machinery.

Vavreckov

a, S. (2020). The combination of skills train-

ing for it administrators and programmers. In ISSEP

(CEURWS Volume), pages 152–159.

Vykopal, J.,

abensk

y, V., Seda, P., and

Celeda, P. (2022).

Preventing cheating in hands-on lab assignments. In

Proceedings of the 53rd ACM Technical Symposium

on Computer Science Education V. 1, pages 78–84,

New York, NY, USA. Association for Computing Ma-

chinery.

Weina, F., Ping, Y., and Shuai, L. (2018). Research on the

related teaching method of computer operating system

course. In International Conference on E-Learning,

E-Education, and Online Training, pages 292–297,

Cham. Springer, Springer International Publishing.

XM: A Customised Software Tool to Facilitate Learning and Professional Skills Development in Operating Systems

337