3 ms·
It's a really cool feature of sqlite. I use it in a prover benchmark tool [1] where each run produces a new sqlite file that I can rsync elsewhere. The tool can
by c-cube 6y ago
It's a really cool feature of sqlite. I use it in a prover benchmark tool [1] where each run produces a new sqlite file that I can rsync elsewhere. The tool can then list these files and connect to multiple files to do joins across them (to compare results from different runs). It's very convenient.
[1] https://github.com/sneeuwballen/benchpress https://github.com/sneeuwballen/benchpress
- Cthulhu_ 6y agoIt's used in the application I'm trying to maintain / rewrite as a backup / revert mechanism; the database migration will create and update two identical databases, on save the revert database is attached and data is copied over. On revert the data is copied back. I mean the author didn't use transactions or anything so if there was any issue while reverting the user would end up losing data, but it's clever I guess. Not the best though. I'm rebuilding it and will likely be using versioning instead.
- thunderbong 6y agoThat's a very interesting approach! Is this available publicly?