barak/cayley
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

Rename Provenance -> Label

Browse source 
's/Provenance/Label/g' 's/provenance/label/g' with human vetting.
This commit is contained in:
kortschak 2014-07-29 08:44:39 +09:30
parent 410202f3e0
commit 274d9ef57e
20 changed files with 416 additions and 416 deletions
Download patch file Download diff file Expand all files Collapse all files

16
TODO.md
View file

@ -26,11 +26,11 @@ Usually something that should be taken care of.
### Bootstraps
Start discussing bootstrap triples, things that make the database self-describing, if they exist (though they need not). Talk about sameAs and indexing and type systems and whatnot.
### Better surfacing of Provenance
### Better surfacing of Label
It exists, it's indexed, but it's basically useless right now
### Optimize HasA Iterator
There are some simple optimizations that can be done there. And was the first one to get right, this is the next one.
There are some simple optimizations that can be done there. And was the first one to get right, this is the next one.
A simple example is just to convert the HasA to a fixed (next them out) if the subiterator size is guessable and small.
### Gremlin features
@ -39,7 +39,7 @@ A simple example is just to convert the HasA to a fixed (next them out) if the s
A way to limit the number of subresults at a point, without even running the query. Essentially, much as GetLimit() does for the end, be able to do the same in between
#### "Up" and "Down" traversals
Getting to the predicates from a node, or the nodes from a predicate, or some odd combinations thereof. Ditto for provenance.
Getting to the predicates from a node, or the nodes from a predicate, or some odd combinations thereof. Ditto for label.
#### Value comparison
Expose the value-comparison iterator in the language
@ -66,7 +66,7 @@ The necessary component to make mid-query limit work. Acts as a limit on Next(),
Hopefully easy now that the AppEngine shim exists. Questionably fast.
### Postgres Backend
It'd be nice to run on SQL as well. It's a big why not?
It'd be nice to run on SQL as well. It's a big why not?
#### Generalist layout
Notionally, this is a simple triple table with a number of indicies. Iterators and iterator optimization (ie, rewriting SQL queries) is the 'fun' part
#### "Short Schema" Layout?
@ -75,7 +75,7 @@ The necessary component to make mid-query limit work. Acts as a limit on Next(),
### New Iterators
#### Predicate Iterator
Really, this is just the generalized value comparison iterator, across strings and dates and such.
Really, this is just the generalized value comparison iterator, across strings and dates and such.
## Longer Term (and fuzzy)
@ -83,7 +83,7 @@ The necessary component to make mid-query limit work. Acts as a limit on Next(),
There's a whole body of work there, and a lot of interested researchers. They're the choir who already know the sermon of graph stores. Once ease-of-use gets people in the door, supporting extensions that make everyone happy seems like a win. And because we're query-language agnostic, it's a cleaner win. See also bootstrapping, which is the first goal toward this (eg, let's talk about sameAs, and index it appropriately.)
### Replication
Technically it works now if you piggyback on someone else's replication, but that's cheating. We speak HTTP, we can send triple sets over the wire to some other instance. Bonus points for a way to apply morphisms first -- massive graph on the backend, important graph on the frontend.
Technically it works now if you piggyback on someone else's replication, but that's cheating. We speak HTTP, we can send triple sets over the wire to some other instance. Bonus points for a way to apply morphisms first -- massive graph on the backend, important graph on the frontend.
### Related services
Eg, topic service, recon service -- whether in Cayley itself or as part of the greater project.
@ -102,6 +102,6 @@ The necessary component to make mid-query limit work. Acts as a limit on Next(),
### All sorts of backends:
#### Git?
Can we access git in a meaningful fashion, giving a history and rollbacks to memory/flat files?
#### ElasticSearch
#### Cassandra
#### ElasticSearch
#### Cassandra
#### Redis