(Initial version) |
m (Added a few more pro/con's and fixed some typos) |
||
Line 5: | Line 5: | ||
= Details = | = Details = | ||
During the Fedora9 campaign, the fedora wiki was used to planning and | During the Fedora9 campaign, the fedora wiki was used to planning, recording and presenting test execution status. Examples can be found: | ||
* [[QA/TestPlans/Fedora9Install| Fedora 9 Installation Test Plan]] | * [[QA/TestPlans/Fedora9Install| Fedora 9 Installation Test Plan]] | ||
Line 17: | Line 17: | ||
** [[QA/TestResults/Fedora9PackageKit/Rawhide| PackageKit (rawhide)]] | ** [[QA/TestResults/Fedora9PackageKit/Rawhide| PackageKit (rawhide)]] | ||
While the information gathered was very helpful, and in the case of swfdec, was used to support dropping the proposed feature from F9, manually editing the wiki was not optimal. Additionally, tracking which test cases to include in a particular test run was easy to mis-type. | While the information gathered was very helpful, and in the case of swfdec, was used to support dropping the proposed feature from F9, manually editing the wiki was not optimal. Additionally, tracking which test cases to include in a particular test run was easy to mis-type. Manually adding results to a wiki-style table entry was also easy to mis-type. | ||
Pro's: | Pro's: | ||
* Recorded test cases provided manual steps needed for execution and verification | * Recorded test cases provided manual steps needed for execution and verification | ||
* The test plan outlined a plan of attack and provided a prioritization and organization for the proposed test cases | * The test plan outlined a plan of attack and provided a prioritization and organization for the proposed test cases | ||
* Wiki worked well for test plan and case content | |||
* Low barrier to entry ... a wiki is straight forward and uncomplex | |||
Con's: | Con's: | ||
* Gather alternative views of test metrics from the wiki is not possible | * Gather alternative views of test metrics from the wiki is not possible | ||
* Wiki not so great for hard metrics (date, tester, pass/fail execution metrics) | |||
= Tasks = | = Tasks = | ||
Line 45: | Line 48: | ||
|- | |- | ||
|} | |} | ||
= Open issues = | |||
* Should the testopia instance just re-use [http://bugzilla.redhat.com bugzilla] product and components? | |||
= Lessons Learned = | |||
The following list outlines important issues uncovered during the testopia proof of concept: | |||
[[Category:QA]] | [[Category:QA]] |
Revision as of 15:29, 13 June 2008
Summary
The Fedora QA team is interested in evaluating Testopia as a test planning and results recording solution.
Details
During the Fedora9 campaign, the fedora wiki was used to planning, recording and presenting test execution status. Examples can be found:
- Additional test results recorded for:
While the information gathered was very helpful, and in the case of swfdec, was used to support dropping the proposed feature from F9, manually editing the wiki was not optimal. Additionally, tracking which test cases to include in a particular test run was easy to mis-type. Manually adding results to a wiki-style table entry was also easy to mis-type.
Pro's:
- Recorded test cases provided manual steps needed for execution and verification
- The test plan outlined a plan of attack and provided a prioritization and organization for the proposed test cases
- Wiki worked well for test plan and case content
- Low barrier to entry ... a wiki is straight forward and uncomplex
Con's:
- Gather alternative views of test metrics from the wiki is not possible
- Wiki not so great for hard metrics (date, tester, pass/fail execution metrics)
Tasks
STATUS | OWNER(s) | TASK |
---|---|---|
DEFERRED | Create testopia.fedorahosted.org site (trac, tickets, git, mailing list etc...) | |
COMPLETE | User:dmalcolm | Rough packaging of testopia so that we can install and test on publictest2 |
INPROGRESS | User:dmalcolm, User:jstanley | Package review for testopia RPM
See bug#450013 |
COMPLETE | User:jkeating | Apply for fedora infrastructure sponsorship. Jesse has volunteered to be our sponsor |
COMPLETE | User:dmalcolm | Apply for shared "publictest environment" to house proof of concept |
INPROGRESS | User:jlaska | Migrate existing F9 data into testopia for sample data set |
Open issues
- Should the testopia instance just re-use bugzilla product and components?
Lessons Learned
The following list outlines important issues uncovered during the testopia proof of concept: