forked from ocymum/trippi-sparql
-
Notifications
You must be signed in to change notification settings - Fork 3
/
RELEASE-NOTES.txt
66 lines (47 loc) · 2.82 KB
/
RELEASE-NOTES.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Trippi-SPARQL 1.0
Release Date: 9 January 2017
Additional Information may be found at:
- https://confluence.si.edu/display/RISSC/Trippi-SPARQL+1.0+Release+Notes
- https://github.com/Smithsonian/trippi-sparql
Team
Developers and Contributors
A. Soroka
Davis, Daniel
Jared Whiklo, University of Manitoba
Special Thanks to our Early Adopters
Nick Ruest, York University
Andy Wagner, University of Toronto
Jared Whiklo, University of Manitoba
John Yobb, University of Saskatchewan
Summary
Trippi-SPARQL 1.0 is the first release that incorporates all of the issues surfaced by the early adopters of this
utility in production environments. It achieves several major objectives:
- To be a drop-in replacement for the Mulgara Trippi implementation for Fedora 3
- Fully support SPARQL 1.1 tested with multiple triplestores
- Support RISearch
- Enable full use of a triplestore independently of the Fedora Repository API while preserving the legacy APIs
- Provide means to gradually update to Fedora 4
Trippi-SPARQL is formally tested only with Fuseki and Jena. However, our early adopters are using other SPARQL 1.1
compliant triplestores successfully, specifically BlazeGraph. Additionally, we test only with Fedora 3.8.1 using
Java 8.
Changes
Support Fedora's Default Aliases
A number of Fedora applications have been written to use a set of default, "convenience" namespace prefixes that are
provided by Fedora. Behind the scenes, Fedora rewrites those prefixes into canonical forms. For example, Islandora
assumes that those prefixes can be used in its code. The previous version of Trippi-SPARQL did not supply those
conveniences for queries not made through Fedora. This put the burden on client software to handle the default
namespace prefixes when not using Fedora's APIs and querying triplestores directly. In Islandora's case, a Tuque fork
was created to rewrite the default namespaces. By adding support in Trippi-SPARQL for Fedora's prefix rewriting,
client software using Fedora-specific namespaces and Fedora APIs should now "just work" without modification
permitting "drop-in" replacement of Fedora's APIs with queries made directly to the triplestore of your choice.
As we transition to Fedora 4, we feel it is best to rewrite client software to avoid assuming on such conveniences.
This version of Trippi-SPARQL makes this an incremental process. Clients can continue to use Fedora APIs including
RISearch, or use Fuseki or any other SPARQL 1.1 triplestore directly, (or both), and clients (including Islandora)
that work with Fedora's APIs should work immediately without change.
Parameterize URI base for graph name
Allows the base used to resolve any relative graph names to be configured.
Housekeeping and Bugs
- Improved README
- Improved POM
- Fixing bug in URI base configuration management
- Allow configuration of triplestore connection pool size