Ok I see so in that case removing the authorname can be a good compromised

Stef
On Nov 22, 2008, at 4:02 PM, Simon Denier wrote:


On 22 nov. 08, at 13:08, Adrian Lienhard wrote:

The reason is that TestCase class >> #methodChanged: is executed each time a method is modified or compiled and it then creates for each test case class a new dictionary that stores the author name (and if not set, this will trigger the popup). This dictionary is then stored in another dictionary in a class variable...

Actually this is something related. Whenever you browse a TestCase class, OB scans the test history to look for previous results. If test results are not present, then an empty dictionary of results is sent back. But this dictionary is still initialized with a timestamp and the author name, so the name box can raise.

Do we need the author name in test results? It's good for traceability, but too much hassle. Easy to fix anyway.



Adrian

On Nov 22, 2008, at 12:43 , Stéphane Ducasse wrote:

simon, alex and others
can you check why the author name is asked while it should not?
This is annoying and I would like to see that fixed.

Stef

On Nov 19, 2008, at 10:02 PM, Simon Denier wrote:

Hi all

Just to notify a new enhancement to review in the PharoInbox.

SLICE-Issue338-SUnitHistoryAndIcons-simon_denier.2


Short doc:

Add:
- support history for test runs as well as progression filter to show test in progression, regression - support icons in OB to show the current state of a test: passed, failed, progressed, regressed...

History support:
- whenever some test cases run from the TestRunner, their results are stored in a class side variable TestResult class>>History
- contextual menu for history in TestRunner>>error pane:
1) History: access submenu to TestRunner instance history: can save and show progress between results ran in the same TestRunner; lost when this particular instance of TestRunner is closed 2) Store results as progress reference: create/update a lastStoredRun method on the class side of each TestCase which have just been run, containing the results 3) Show progress: make a diff between the last run and the lastStoredRun, showing progress or regress in tests (new passed, new failures, new errors)

The idea is that whenever you are satisfied with the current results of some of your tests, you can save this state as a reference point with the 'Store results...' action. This reference point can be compared with subsequent test results to show your progression or regression.

Test results are stored in a class side variable some are kept local to your image. Reference results are stored in a special method for each TestCase, so that they are saved with their package and can be used as a common reference for package developers. So a good practice is to store the latest test results as reference before committing your changes.


Icons support:
OB Package browser now has icons for test methods in TestCase subclasses:
- green circle: test has passed in the last run
- red circle: test has failed
- half green/half red circle: state unknown, test has no history or test has changed since last run
- (no icon): test has raised an error
- upside green arrow: test has progressed since last reference run
- downside blue triangle: test has regressed since last reference run

A test is in progression if it goes from error or failure to passed state. A test is in regression if it goes from passed to error or failure state.

The last two icons have other meanings with methods not related to test, so we should change them whenever someone has better icons.

Authors: Alexandre Bergel, Simon Denier, David Röthlisberger

--
Simon




_______________________________________________
Pharo-project mailing list
[email protected]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project



_______________________________________________
Pharo-project mailing list
[email protected]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project


_______________________________________________
Pharo-project mailing list
[email protected]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project


--
Simon




_______________________________________________
Pharo-project mailing list
[email protected]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project



_______________________________________________
Pharo-project mailing list
[email protected]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Reply via email to