We had to make changes to some of the virtual operations that were added to allow voting and payout information to be moved to hivemind. These problems were all identified by hivemind’s automated testing system:
https://gitlab.syncad.com/hive/hive/-/merge_requests/85
https://gitlab.syncad.com/hive/hive/-/merge_requests/86
https://gitlab.syncad.com/hive/hive/-/merge_requests/88
https://gitlab.syncad.com/hive/hive/-/merge_requests/90
We also added some new tests for hived and made some improvements to the automated testing system itself. Notable additions to the testing system include replay tests and tests for saving/restoring state information (new feature added in Eclipse release).
https://gitlab.syncad.com/hive/hive/-/merge_requests/87
https://gitlab.syncad.com/hive/hive/-/merge_requests/89
We made a few more small improvements to the hivemind’s automated testing system, but most of the time was spent fixing hivemind bugs that were identified by the testing system. Most of these bugs were due to the movement of data from hived to hivemind (to reduce memory usage), and several fixes also required changes to hived as well, as previously mentioned.
Improve automated testing for hivemind:
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/38
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/46
Hivemind bug fixes for bugs found by API tests
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/39
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/40
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/41
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/42
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/43
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/44
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/45
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/48
General code improvement
https://gitlab.syncad.com/hive/hivemind/-/merge_requests/47
We added support for Dapplr videos. As part of this work we reviewed security associated with iframe usage in condenser. Iframes allow content from other sites such as youtube, threespeak, dapplr, etc to be displayed inside of posts.
https://gitlab.syncad.com/hive/condenser/-/merge_requests/81
https://gitlab.syncad.com/hive/condenser/-/merge_requests/79
https://gitlab.syncad.com/hive/condenser/-/merge_requests/77
We’re still doing user-interface work for decentralizing lists; the latest change is to allow a user to add a “description” to his lists, so that other users can see the criteria used to make each list. These descriptions will be shown to new users when they first pick lists to follow and they will also be shown on the pages where the user has followed other people’s lists. This change also requires an update to hivemind to store these descriptions.
In my last post, someone asked how to follow the work of individual programmers. Most of our devs don’t post on-chain, but if you’re interested in getting a snapshot view of what individual programmers are working on, you can use the links below:
https://gitlab.syncad.com/jsalyers
https://gitlab.syncad.com/ABW
https://gitlab.syncad.com/bwrona
https://gitlab.syncad.com/Trela
https://gitlab.syncad.com/kmochocki
https://gitlab.syncad.com/dkedzierski
https://gitlab.syncad.com/dan
https://gitlab.syncad.com/Ickiewicz (new dev)
https://gitlab.syncad.com/klesniak (new dev)
In addition to the devs above, we have several other devs contributing on a part-time basis to various projects such as condenser and hived, plus some full-time and part-time devops guys working on various Hive-related tasks (server setup and maintenance, manual testing, performance testing, etc).
We’re planning to replace the pyresttest testing framework used by hivemind’s automated testing system with the https://github.com/taverntesting/tavern framework. Pyresttest is no longer actively developed, and tavern also provides better reporting and better integration with gitlab’s automated testing environment. Tavern will also allow us to mark individual tests within a testing job as “allowed to fail” so that the overall test job doesn’t report as a failure, which can be very useful when making changes that are expected to temporarily break some tests.
Unfortunately, we weren’t able to complete enough testing this week that we were comfortable to make a release candidate for hived, mainly because we keep finding hivemind-related bugs that required us to make changes in the virtual operations needed by hivemind.
But at this point we have enough tests passing that I think we’ll be able to provide a hived release candidate early next week. That means we are a little over 1 week behind compared to our original, optimistic schedule for HF24.
