stephen mallette commented on TINKERPOP-892:

Perhaps this was a bad oversight, but our IO formats are pretty well set at 
this point and I can't see us making a change to them for 3.x. I think we leave 
variables up to the graph provider to persist as they see fit. In TinkerPop 4.x 
we can re-think the role of variables and how they might be persisted. I'll 
make a note in the "future dev doc" about this for consideration.

> Saving Graph to file won't save graph variables
> -----------------------------------------------
>                 Key: TINKERPOP-892
>                 URL: https://issues.apache.org/jira/browse/TINKERPOP-892
>             Project: TinkerPop
>          Issue Type: Improvement
>          Components: structure
>    Affects Versions: 3.0.2-incubating
>            Reporter: Flavio Cordova
>            Priority: Trivial
> When you save a graph using writeGraph graph variables are not saved along 
> with all the vertices and edges, so you can't fully recover the graph when 
> you read the file.
> In the forum there's a thread called "Saving graph variables to File" that 
> will discuss the matter in details but basically "I understand the API should 
> handle variables and persist then properly, probably adapting it as a vertex 
> to be compliant to the format, but it should be transparent to the developer, 
> something like neo4J does for multi-properties. IF the way they are written 
> to the file were not good enough [for me] then I could be able to change it 
> writting a custom writer/reader, but not the way around."

This message was sent by Atlassian JIRA

Reply via email to