| Commit message (Collapse) | Author | Age | |
|---|---|---|---|
| * | More debug output when failing to load resource plugins. | Christian Mollekopf | 2018-05-15 |
| | | |||
| * | Fixed more warnings and then disabled them | Christian Mollekopf | 2018-03-25 |
| | | | | | ..because there is a boatload more to fix. | ||
| * | The path is not called "plugins" on macosx. | Christian Mollekopf | 2018-03-16 |
| | | |||
| * | Support for storage upgrades | Christian Mollekopf | 2018-01-30 |
| | | |||
| * | Avoid storing the password in the configuration | Christian Mollekopf | 2017-09-18 |
| | | | | | | | | The password (or any other secret), is now cached in the client process (in-memory only), and delivered to the resource via command. The resource avoids doing any operations against the source until the secret is available. | ||
| * | static members can apparently still lead to crashes... | Christian Mollekopf | 2017-08-13 |
| | | |||
| * | Don't set capabilities as property on creation. | Christian Mollekopf | 2016-12-15 |
| | | | | | | | Instead we make it part of the plugin. This ensure we also have access to the proper capabilities when creating a resource via sinksh. | ||
| * | Used the CommandProcessor as central place for all command processing. | Christian Mollekopf | 2016-11-28 |
| | | |||
| * | Implement debug stream operators for query. | Christian Mollekopf | 2016-11-07 |
| | | |||
| * | Specify the resource name in one place | Christian Mollekopf | 2016-10-21 |
| | | |||
| * | Avoid using the resource object to remove the data from disk. | Christian Mollekopf | 2016-10-07 |
| | | | | | | ...because creating it will potentially start transactions on the database we're about to remove. | ||
| * | Refactored the generic resource to use separate classes for | Christian Mollekopf | 2016-05-28 |
| | | | | | | | | | changereplay and synchronization. This cleans up the API and avoids the excessive passing around of transactions. It also provides more flexibility in eventually using different synchronization strategies for different resources. | ||
| * | Fromatted the whole codebase with clang-format. | Christian Mollekopf | 2016-03-03 |
| | | | | | clang-format -i */**{.cpp,.h} | ||
| * | Executed database removal in the resource instead of the client. | Christian Mollekopf | 2016-02-08 |
| | | | | | | The resource doesn't really notify all clients properly about the removal, but the tests all still pass. | ||
| * | Build without warnings | Christian Mollekopf | 2016-01-31 |
| | | |||
| * | Renamed Akonadi2 to Sink | Christian Mollekopf | 2016-01-20 |
| | | | | | (except for documentation). | ||
| * | Let clients tell the resource when they no longer require a revision. | Christian Mollekopf | 2015-10-19 |
| | | |||
| * | Untangled the include dependencies a bit. | Christian Mollekopf | 2015-08-13 |
| | | | | | We no longer depend on clientapi.h from everywhere. | ||
| * | Made the pipeline an implementation detail of the resource. | Christian Mollekopf | 2015-07-30 |
| | | | | | | This removes one dependency from the Listener and will allow us to test the Listener better. | ||
| * | Forward revision updates through resource | Christian Mollekopf | 2015-07-28 |
| | | |||
| * | Pass command around as QByteArray | Christian Mollekopf | 2015-07-23 |
| | | | | | | | | Simpler api, GenericResource didn't honor size anyways, and we copy the command for now to avoid sideeffects of data coming in in the meantime (although that should generally work since data is always appended). | ||
| * | Adapt to KAsync namespace change | Dan Vrátil | 2015-05-15 |
| | | |||
| * | A way to ensure all messages have been processed. | Christian Mollekopf | 2015-01-25 |
| | | | | | | As queries become reactive this should become less important. We can then just wait until all results become available. For tests it is in either case useful though. | ||
| * | DomainTypeAdaptor factory, per type preprocessor pipeline configuration. | Christian Mollekopf | 2015-01-02 |
| | | |||
| * | Buffers wrapped into entity buffer, async command progress tracking. | Christian Mollekopf | 2014-12-28 |
| | | |||
| * | a basically-working Pipeline implementation | Aaron Seigo | 2014-12-17 |
| | | | | | | still a skeleton rather than a full body with flesh and blood, but it is getting there! | ||
| * | add pipelines (as a sketch only), message ids and message responses | Aaron Seigo | 2014-12-16 |
| | | |||
| * | add a synchronize method | Aaron Seigo | 2014-12-16 |
| | | |||
| * | go into the correct subdir | Aaron Seigo | 2014-12-16 |
| | | | | | | | this will allow us to have different plugin types in future with nicer separation of the libraries from each other (faster scan times, less chance of name collision, etc) | ||
| * | first run at a plugin loader | Aaron Seigo | 2014-12-16 |
| | | |||
| * | split out the resource bits into their own file | Aaron Seigo | 2014-12-16 |
 |
index : sink | |
| A data access layer handling synchronization, caching and indexing |
| summaryrefslogtreecommitdiffstats |