forked from gousiosg/github-mirror
-
Notifications
You must be signed in to change notification settings - Fork 0
/
CHANGELOG
110 lines (96 loc) · 4.83 KB
/
CHANGELOG
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
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
= Version 0.11
* Retrieve members by processing MemberEvents to counter API change
* Removed the request caching layer. Requests are cached in the persister
* Change default DB isolation to REPEATABLE READ for stronger isolation
* Finer-grained (commit level) transactions when processing forks
* More accurate and uniform logging
* Tool to push logs to InfluxDB for monitoring
* Drop ext_ref_id from all tables
* More efficient retrieval of events, 100 in one go
* Tool to retrieve all user details and support for marking users deleted
* Support for retrieving repo events when using ght-retrieve-repo
* Non-recursive retrieval of pull requests leads to 1/3 API calls
* Custom rate limits for Github API tokens
* Tooling for MySQL dumps in CSV files
* General bug fixes and cleanups
= Version 0.10
* Base class for multiprocess queue clients
* Make retrieval of pull request commits faster
* Drop table forks
* Remove merged and user_id fields from pull_request
* Ruby 2.0.0 support
* Fast(er) path to project forking. Copy commits from base repo instead of getting them from Github.
* Github API tokens for authentication
* Support for disabling the cache
* A new real-time log analyzer
* Mark fake users as such
* Organization members are now retrieved
* Authenticated MongoDB connections
* General bug fixes and cleanups
= Version 0.9
* Remove dependency to EventMachine-based AMQP client. We now use bunny.
= Version 0.8.1
* New tool to retrieve specific entities and their dependencies
* New tool to retrieve repositories en masse
* Support for resuming when exception occurs while processing items in loops
* Support for finer grained transactions when processing large entities
* Commit comments are now indexed per owner/repo (was just by comment id)
* Remove the unused daemon mode
* Various exception fixes and more detailed logging
= Version 0.8
* Retrieve and process issue labels
* Retrive and process actors for pull request events
* Retrieve pullreq comments for projects with no issuetracker
* Better heuristics for avoiding dulpicate entries in pull request histories
* The event loader now loads events by IDs to reduce pressure to the queue
* Compound indexes in MongoDB by default
* X-RateLimit-Reset header support
* Remove lots of dead code, general cleanups
= Version 0.7.3
* Support for running in standalone mode
= Version 0.7.2
* Support for running in standalone mode
= Version 0.7
* Full support for issues (comments, labels etc) and pull requests
* Cleaned up retrieval of pull request commits
* Cleaned up association of commits with repositories.
* Removed the forks table. Forks are now tracked by the forked_from field in projects
* Use Github's HTTP headers for request throttling
* Support for setting user agent header as per Github API requirements
* Support for marking projects as deleted (run fixes/update_deleted.rb)
* New tool (ght-retrieve-user) to retrieve all data for a single user
* Support for running without a config.yaml file
* Technical report in doc/
= Version 0.6
* Support retrieval of issues, issue events and issue history
* Support for setting username/password for performing requests
* Respect by default Github's x-ratelimit-remaining header
* Selective processing of events for user-specified repos
* New tool (ght-get-more-commits) to retrieve all commits for a repository
* New tool (ght-process-events) to process just one event by id
* Retrieve 100 items at once by default on multipage requests
* Rename watchers -> stargazers, as per Github API change
* Fixes to bugs that permitted efficient processing of multipage requests
* Several fixes on how pull requests are being processed
* Users with invalid git setups are now allowed
* Compatibility with Ruby 1.8 restored
= Version 0.5
* Generic methods for retrieving items that are bound to repositories
* Processing of pull requests with commits, comments and history
* Processing of project forks
* New tool (ght-load) to filter and load events to the queue
* New tool (ght-rm-dupl) to delete duplicate entries from collections (events & commits supported)
* Project wide requesting result caching for multi-page requests
* Better logging in various places
* Better defaults for MySQL (UTF8 + InnoDB tables)
* Commits are now seperated from projects. Project forks can share commits.
* Support for setting the IP address to use for retrieval on multi-homed
* Compatibility with Ruby 1.9 (now default) and JRuby
* Proper modularization, following the cake design pattern
* Never retrieve arrays of results from MongoDB
= Version 0.4
* Implement support for retrieving watches and project members
* Support for processing FollowEvents, WatchEvents, CommitCommentEvents, MemberEvents
* Projects are exclusively associated to users
* Remove dependence on Github API v2
* Remove license headers