Graeme Geldenhuys wrote 10 lines
Op 2010-05-30 21:33, Joost van der Sluis het geskryf:
>
> Sigh:
http://www.freepascal.org/~joost/wwwroot/testsuiteweb.cgi
The new one looks very nice.... Some pointers on things I noted (hope you
don't mind):
* once you view the details of a test run, there is not Back button or Main
button to get back to the main screen.
I have a really big one for that in my browser. But I'll see if that can be added.
* On the main screen I have no idea what date format is used? Maybe use
the ISO 8601 (international date/time) format to stop any possible
confusion.
Yeah, and the time-part is also included, so filtering on it will be difficult. Have to fix that.
Show quoted text 6 lines
* In your initial post you mentioned that it is tests run each night. Yet
all the tests listed on the first page you see are (I think - that's if
I am interpreting the date format correctly) 2003-11 and 2003-10. That's
rather long ago, and I don't really see the point in keeping such a
archive for daily tests runs. Surely the daily results is all you want
to see, or a maximum of say last 7 days.
Click on the date-column header, it will sort on the date in reversed order. But that should indeed be the default.
Show quoted text 5 lines
* Failure details. On the main screen it says the one test run caused 33
failures. I click on that row which opens the details. Page 1 or 66!
1) Maybe that could be reduced to some summary/collapsed view first.
2) How do I get a listing of only the 33 tests that failed? I don't
want to scroll 66 pages looking for them.
You can sort on the 'OK' column, but in that case you should filter out the skipped-tests. (Which is currently not possible)
I'll add a 'failed' column. And a way to filter tests.
* In the test listed below, which was a test run for 2003-10-31, there is
a test result dated 2007-01-28?? See attached image. So somewhere your
details query is pulling wrong data, or the data is corrupt.
Good question. That date is indeed wrong. (should be 8/10/03 according to the 'old' website.)
Joost.