Beta-testing SuperMemo

From SuperMemopedia
Jump to navigation Jump to search

Beta-testing strategy

SuperMemo beta-testing strategy has been refined in the 1990s. With the arrival of internet, the procedure resulted in SuperMemo having a reputation of a stable application. However, it is also known to be beginner-unfriendly. It is relatively easy to mess up SuperMemo collection by just deleting its files or making simple errors (e.g. in backup).

Procedure

  • Alpha-testing in house. Most dedicated users of SuperMemo at SuperMemo World use the new program for their own learning. The threats are well-identified in advance. This is why there were no major mishaps at this early stage. Beta-testing usually begins with the program well-tested and relatively safe.
  • Beta-testing over the internet. Users are invited incrementally depending on the testing needs. The circle is expanded at no more than 2-3 users per day. The list of testers does not exceed a hundred. Large numbers cause chaos. At some point, there are more suggestions, support questions, or insignificant bugs than actually bug-documenting material. See: How to join beta-testing of the new version?
  • Suggestions are classified on a tasklist of the next future release. Bugs are kept on a separate tasklist (see: How many bugs are there in SuperMemo?).
  • Bugs are fixed from the top of the tasklist where priority is set by Value/Cost
  • Gamma testing lasts post-release. Beta-tester use a version of the program equipped with enhanced error reporting

Design contest

Wrath0flog1c proposed:

Redesign Contest: Can boost innovation, no extra cost. We propose launching a contest inviting users to propose new designs for SuperMemo. This will not only stimulate creative solutions but also foster a sense of collective ownership among the users. The designs could then be voted on and potentially incorporated into the future updates

Beta-testing

Forming a Beta Testing Team: Once we have potential new features or designs from the contest, a team of dedicated users could serve as beta testers. Their feedback would help refine these features before a wider release, improving the quality of updates

Scepticism

As SuperMemo interface is huge, redesigning it from scratch, implementing and testing seems like a monumental job. Isn't it better to write a perfect app from scratch? Why struggle with the limits imposed by trade secrets or code disclosure formalities? Instead of listening to his own imagination, the designer would need to follow tracks that have been beaten for 35 years. Many of those are blind alleys.

See also