stronglink
stronglink copied to clipboard
btrask
A searchable, syncable, content-addressable notetaking system
Sort of related to #52. Caching is a little bit tricky on our search results because usually the page can't change, unless it is showing the newest results. We should...
Our main navigation options, "Log In | Compose | Upload" are kind of weird for a public blog where the user isn't logged in. When the public session is viewing...
sln-pipe is not particularly robust currently. It detects disconnections on the upstream (sending) side, but not on the downstream (receiving) side. I'm thinking we should add another API endpoint for...
Sort of a meta-issue: - [ ] Configuration interface #6 - [ ] Native GUIs for various platform (OS X, mobile?) - [ ] Building as a library with a...
We should keep statistics so we can estimate the number of results for any given query. This estimate must be much faster than generating a full list of results. Uses:...
When the user performs a query pertaining to a URI, for example "show me everything that links here," they are talking about file or a set of files sharing that...
Negation is hard because it can allow entries to be removed after the fact. Whether you want that or not depends on what you're doing with the results. I think...
Information: - https://lkml.org/lkml/2012/9/3/83 - https://github.com/LMDB/lmdb/blob/91fc43f17cec8ffa05a5c841831df1cfa0511469/libraries/liblmdb/mdb.c#L82 It's been fixed since Linux 3.6. For now, users should be careful running StrongLink on affected kernels.
There should be some way to manually pause syncs, including partially transferred files. We should also handle download resuming in an elegant way. Related to #1.
Would cut down copying and syscalls compared to the current system. The main bottleneck is `HTTPConnectionWriteChunkFile`, which is called once per result by the blog interface. For the best case...