in attendance: Oleksandr, Zooko (scribe), Mark_B, Amber First we looked at the last remaining test failure that happens on Mark's #1382 branch. The test failure is in test_each_byte ¹, when it is supposed to be corrupting the ciphertext hash tree. Mark_B's changes changed the corrupted share from share-num 0 to share-num 2, although this test really shouldn't be sensitive to *which* sharenum was being corrupted. We began to suspect that the offsets in "need_4th_victims" might be incorrect with respect to the layout specified in src/allmydata/immutable/layout.py ², but we stopped before confirming that. Mark is going to email Brian (the primary author of that test) and ask for help.
I drew Then we talked about what else Mark should work on, since he is mostly done with his original GSoC project (☺!). I asked what he was interested in. He suggested ¹ https://tahoe-lafs.org/trac/tahoe-lafs/browser/trunk/src/allmydata/test/test_download.py?annotate=blame&rev=3d771132a843a85578dc23a6cac55b4fae09fc64#L945 ² https://tahoe-lafs.org/trac/tahoe-lafs/browser/trunk/src/allmydata/immutable/layout.py?annotate=blame&rev=3668cb3d068b7f3a56cc88e7c45d4f81c0b4f499#L27 https://tahoe-lafs.org/trac/tahoe-lafs/ticket/1382# immutable peer selection refactoring and enhancements Regards, Zooko Wilcox-O'Hearn Founder, CEO, and Customer Support Rep https://LeastAuthority.com _______________________________________________ tahoe-dev mailing list [email protected] https://tahoe-lafs.org/cgi-bin/mailman/listinfo/tahoe-dev
