QA/TDAI/Automation Software: Difference between revisions

From MozillaWiki
< QA‎ | TDAI
Jump to navigation Jump to search
No edit summary
No edit summary
 
(6 intermediate revisions by 5 users not shown)
Line 1: Line 1:
<small>[[QA/TDAI|&laquo; QA/TDAI]]</small>
<table cellspacing="0" cellpadding="5" style="font-size: small;border: 1px solid black;margin: 15px 0 15px 15px;">
<table cellspacing="0" cellpadding="5" style="font-size: small;border: 1px solid black;margin: 15px 0 15px 15px;">
<tr>
<tr>
Line 39: Line 41:
<td><ul><li>Python-based: would be easy to extend</li></ul></td>
<td><ul><li>Python-based: would be easy to extend</li></ul></td>
<td><ul><li>Reliant on accessibility hooks.</li>
<td><ul><li>Reliant on accessibility hooks.</li>
<li>still very alpha</ul></td>
<li>still very alpha</li></ul></td>
<td>Under investigation (bc)</td>
<td>Under investigation (bc)</td>
</tr>
</tr>
Line 47: Line 49:
<td><ul><li>someone in China is already working to automate testing of Firefox (1.0.6)</li></ul></td>
<td><ul><li>someone in China is already working to automate testing of Firefox (1.0.6)</li></ul></td>
<td><ul><li>Requires Gnome</li>
<td><ul><li>Requires Gnome</li>
<li>Reliant on accessibility hooks.</li></ul>
<li>Reliant on accessibility hooks.</li></ul></td>
<td>Under investigation (coop)</td>
<td>Under investigation (coop)</td>
</tr>
</tr>
Line 62: Line 64:
<td><ul><li>promises automated and "hidden" testcase generation</li></ul></td>
<td><ul><li>promises automated and "hidden" testcase generation</li></ul></td>
<td><ul><li>Reliant on expert domain knowledge (hard to get) to seed testcase generation.</li>
<td><ul><li>Reliant on expert domain knowledge (hard to get) to seed testcase generation.</li>
<li>No existing examples (that are accessible at least)</li></ul>
<li>No existing examples (that are accessible at least)</li></ul></td>
<td>Waiting on feedback from Edjard and/or Monica. We tasked them to provide some proof-of-concept testcases for Software Update and Preferences.</td>
<td>Waiting on feedback from Edjard and/or Monica. We tasked them to provide some proof-of-concept testcases for Software Update and Preferences.</td>
</tr>
</tr>
Line 71: Line 73:
<td><ul><li>Captures a lot of information that we don't really care about.</li></ul></td>
<td><ul><li>Captures a lot of information that we don't really care about.</li></ul></td>
<td>Possibly reuse this design pattern in the future, but the tool itself is of limited use.</td>
<td>Possibly reuse this design pattern in the future, but the tool itself is of limited use.</td>
</tr>
<tr style="color: #303030; background-color: #FFFFFF;">
<td>[http://wtr.rubyforge.org/ WATIR]</td>
<td><ul><li>Used by many in the testing community</li>
<li>Designed for use in testing</li></ul></td>
<td><ul><li>Drives IE, not FF</li>
<li>Designed for testing web app, not browser</li></ul></td>
<td>davel to investigate</td>
</tr>
<tr style="color: #303030; background-color: #DFDFDF;">
<td>[http://samie.sourceforge.net/ SAMIE]</td>
<td><ul><li>Designed for use in testing</li></ul></td>
<td><ul><li>Drives IE, not FF</li>
<li>Very little recent activity</li></ul></td>
<td>davel to investigate</td>
</tr>
</tr>


</table>
</table>

Latest revision as of 06:40, 15 November 2007

« QA/TDAI

Software Advantages Disadvantages Status
Spider
  • acts as test harness for various suites of tests: JS, DOM, top site
  • Developer (Bob Clary) on staff
In use on QA test farm.
Selenium
  • Designed for web application testing, not end-to-end browser testing.
Under investigation
Eggplant
  • Cross-plaform (uses VNC)
  • Bitmap-based testing is inherently fragile.
Coop is working with Eggplant on the test farm to generate some proof-of-concept tests.
TestGen4Web
  • Designed for web application testing, not end-to-end browser testing.
Under investigation
dogtail
  • Python-based: would be easy to extend
  • Reliant on accessibility hooks.
  • still very alpha
Under investigation (bc)
LDTP
  • someone in China is already working to automate testing of Firefox (1.0.6)
  • Requires Gnome
  • Reliant on accessibility hooks.
Under investigation (coop)
Tomato Under investigation (coop)
Grubzilla
  • promises automated and "hidden" testcase generation
  • Reliant on expert domain knowledge (hard to get) to seed testcase generation.
  • No existing examples (that are accessible at least)
Waiting on feedback from Edjard and/or Monica. We tasked them to provide some proof-of-concept testcases for Software Update and Preferences.
Tracer
  • architecture is interesting
  • Captures a lot of information that we don't really care about.
Possibly reuse this design pattern in the future, but the tool itself is of limited use.
WATIR
  • Used by many in the testing community
  • Designed for use in testing
  • Drives IE, not FF
  • Designed for testing web app, not browser
davel to investigate
SAMIE
  • Designed for use in testing
  • Drives IE, not FF
  • Very little recent activity
davel to investigate