-
Notifications
You must be signed in to change notification settings - Fork 0
/
5discussion.tex
executable file
·39 lines (19 loc) · 4.53 KB
/
5discussion.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
%!TEX root = thesis.tex
\chapter{Discussion}
\label{chapter:discussion}
This Chapter will discuss the problem statements based on the proposal and implementation described in previous Chapters of this master's thesis. Possible future research topics around the subject matter of this thesis are presented related to the research questions.
\section{Benefits of Open Source}
Even though open-source license of a software does not guarantee the quality and excellence of the product compared to the closed source counterparts, the usage of open source solutions was beneficial for this project. As overall budget of the project was set rather low, open source products provided advantage over proprietary solutions.
Use of open source software provided possibility to evaluate the possible tools more throughly before actually taking them into the project. This was especially beneficial in terms of efficiently evaluating the potential tools and keeping the development time frame short.
Use of open source products also provided benefits to the customer organization whom the project was developed for. As the system under testing had strict security requirements, open source solutions provided visibility and transparency to the users and developers of the tested product.
Comparison between open source and proprietary products can be done in numerous different aspects and research done within this master's thesis was restricted due to time and cost limitations of the project. Future research could address this comparison more thoroughly by comparing AAT environments developed strictly with either open-source or proprietary components.
\section{Characteristics of Payment Terminals}
Different types of payment terminals were examined within this master's thesis and it was found out that due to the simple function of the payment terminal, the design usually involves few common parts: a display, a keypad and a card slot.
Scope of the master's thesis was limited to certain types of payment terminals and more exotic models were left out of consideration. Developed environment only supports payment terminals using chip card slot for inserting the payment card and use of other reading methods of the payment card, e.g. reading of magnetic stripe or NFC-chip, are not supported.
For the future work, possibility to support other reading methods of the payment card is suggested to being researched.
\section{Approaches for Test Automation}
Black box testing was used as a testing methodology within the work done in this master's thesis. Choice of the methodology was done entirely based on the definitions found from the literature around this topic. Use of black box testing methodology in the automated acceptance testing of embedded systems can be seen as most reasonable option as it imitates the final user most accurately. Other methodologies would have required more in-depth knowledge of the underlaying systems of the devices and this would not have emulated the final human user as accurately as black box testing methodology did in this case.
Methodology worked well in the AAT environment implemented in this master's thesis. AAT environment imitated final human user to the extent that it was possible to mostly automate the manual testing of the payment terminal, which was the goal of this project. As the AAT environment presented in this master's thesis concentrated on validating only the textual content of the payment terminal display, other visual validations were still left to be testes manually.
\section{Syntax for Test Suites}
Robot Framework was selected as a testing framework of the AAT environment. Choice of the framework and therefore also the test syntax was done based on literature review and examination of different tools. Robot Framework was selected for its modularity and versatile and human-readable test syntax. Use of RF proved to be robust and it was able to implement all the desired functionalities using the framework.
For future research, it is encouraged to arrange surveys and interviews related to the different acceptance testing frameworks. Research done within this master's thesis did not involve any investigation about current opinion atmosphere around the topic of acceptance testing tools used in testing of embedded software. This kind of research would be valuable to the future projects done in the field of automated acceptance testing. Multiple competitive testing tools exists and as the evaluation of the tools require extensive usage of different solutions, it would be beneficial if comparative and unbiased data would be widely available.