<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<!-- BEGINNING of header -->
<!-- $Id: test_policy.html,v 1.8 2001/05/05 10:23:30 amai Exp $ -->
<html>
<head>
<title>Example SUMMARY file</title>
</head>
<body bgcolor="#ffffff" text="#000000">
<!-- END of header -->
<p> At the top of the file, please add an entry for yourself and your
hardware configuration to the SUMMARY file when you create or modify it. For
example: <br>
<pre>
Tester: Jon A. Christopher (JAC) <jac8792@tamu.edu>
SGI-Motif 1.2.4
SGI-LessTif 0.85.2
Tester: Rick Scott (rws) <rwscott@omnisig.com>
HP Motif 1.2
Linux LessTif 0.85.2
</pre>
<p>The next section should be a keyword summary section.
For example from test/label, I've got:
<pre>
test1: POSSIBLE FAIL
test2: PASS
test3: GEOMETRY
test4: PASS
test5: ERROR
test6: POSSIBLE FAIL
</pre>
The keywords are, in order of desirability:
<ul>
<li> PASS - test is ok in all respects
<li> GEOMETRY - the test is ok, but has minor geometry differences/problems
<li> VISUAL - similar to above. The test seems to be ok, but there's a
visual defect somewhere.
<li> POSSIBLE FAIL - The test appears to function like the Motif version,
but the output diagnostics are different.
<li> FAIL - the LessTif version of this test didn't perform like its Motif
counterpart.
</ul>
Please be strict on the tests. There's no sense giving something a PASS
at this point, unless it really appears to be 100% correct. However, if
one test is a minor modification of another test, and the first test fails
for some reason, that doesn't necessarily mean the second test should be
given a FAIL. For example, if <tt>test1</tt> fails because of a missing
translation, and <tt>test2</tt> is the same test, but uses different
fonts, or something, the best way to report this would be:
<pre>
test1: FAIL
test2: PASS (except as noted in test1)
</pre>
Other possible keywords are:
<ul>
<li> UNKNOWN - The purpose if this test is a mystery to the tester.
Someone else will have to evaluate this test.
<li> ERROR - either the Motif or LessTif version of the test wouldn't
compile.
</ul>
<p> Following the keyword section is a detailed summary of each test. You
should date and initial each entry, describe what the purpose of the test
appears to be, and then describe any problems with the test, either
differences from some Motif behavior, or just plain misbehavior of the
test itself. For example:
<pre>
======================================================================
test9:
JAC Jun 23 1998
This test was written to demonstrate some problems with the redraw code
for message boxes. In particular, if you change the size of a label,
forcing the message box to increase in size, you sometimes get spurious
shadows in the larger dialog. The Motif version does not suffer from
these problems.
RWS Jul 4 1998
This has been fixed.
======================================================================
</pre>
<p> If there are extensive comments about a test in the summary file that
no longer apply (i.e. a bug has been fixed), it would be better to move
those comments to a separate entry at the end of the file than to delete
them entirely. That way we can see what the history of the tests has
been, but the top of the file will be fairly clean and easy to skim
through.
<p>Finally, and this is a general comment to LessTif developers, if you
make changes to a widget, you should re-run the tests for that widget and
see if any of the outstanding bugs are fixed. You should update the
SUMMARY file to reflect any changes. Also, if you notice anything not
already noted in the SUMMARY file, please add an annotation to the SUMMARY
file. Also, if you add any new tests, <b>please</b> describe it in the
SUMMARY file; sometimes it's impossible weeks later to figure out what a
given test program was supposed to be demonstrating.
<p>After you're done testing, please send in your changes to the SUMMARY
file to the mailing list, either as a patch or the whole file.
(Of course, if you're on the core team, just check it in!)
<!-- ---- Generic HTML footer -------------------------------------------- -->
<HR>
<address>
<a href="http://validator.w3.org/check/referer"><img
src="images/vh32.jpg" height=31 width=88
align=right border=0 alt="Valid HTML 3.2!"></a>
<a href="feedback.html">Feedback</a><br>
<SMALL>
Last modified on $Date: 2001/05/05 10:23:30 $
</SMALL>
</address>
</BODY>
</HTML>
