| Title | No automated performance test |
| Status | open |
| Priority | nice |
| Assigned user | Gareth Rees |
| Organization | Ravenbrook |
| Description | There is a risk that changes will affect performance but we won't discover this because there is no automated test case that checks performance. |
| Analysis | The problem is, what can we use as the baseline? We can't use absolute numbers because we need tests to be portable across different platforms. But even with relative numbers, what do we use as the baseline? Could compare against malloc performance? RB suggested [1] that it might be possible to get reliable performance figures out of Travis. Alternatively, we could set up our own continuous integration server (see job003834) and get performance figures out of that. |
| How found | inspection |
| Evidence | [1] <https://info.ravenbrook.com/mail/2016/03/03/21-27-46/0/> |
| Created by | Gareth Rees |
| Created on | 2014-05-01 16:02:45 |
| Last modified by | Gareth Rees |
| Last modified on | 2016-03-06 11:20:59 |
| History | 2014-05-01 GDR Created. 2016-03-06 GDR Updated based on [1]. |