Added a bit of discussion

[tt2015.git] / a3 / 5discussion.tex

diff --git a/a3/5discussion.tex b/a3/5discussion.tex
index 87a78c1..7a4a96a 100644 (file)
--- a/a3/5discussion.tex
+++ b/a3/5discussion.tex
@@ -32,4 +32,13 @@ exhaustive search will also generate huge tests for big models.
 Both JTorX and \GAST use on-line checking. They test the SUT on the fly by
 syncing state transitions in the internal model of the SUT with observed output (after input) from the SUT.
 
-%iets over ease of use van test model generation
\ No newline at end of file
+For JTorX test generation was very simple since it accepts models in GraphViz
+DOT language, which is also the language used for describing the models for this
+report. For \GAST the model generation required a bit more work, as the model 
+is written as a function in Clean. This translation is however easily done by 
+hand. Because the model for \GAST is written as a function in Clean it is very
+expressive, the modeling method chosen for  JTorX however is a lot less 
+expressive, as it just allows for specification of states and transition labels
+between them. However JTorX allows for many different modeling languages, which
+can be more expressive than modeling using GraphViz models \cite{JTorXSite}.
+However, for our tests GraphViz models were expressive enough. 
\ No newline at end of file