| *** locallycompact has quit IRC | 06:10 | |
| *** locallycompact has joined #trustable | 06:11 | |
| reiterative | toscalix: I think that Resilience is the right term, but I still don't know where it should sit. My current thinking is that durability, reliability and resilience all belong under a broader top-level concept - perhaps robustness | 08:07 |
|---|---|---|
| reiterative | persia: I have explicitly avoided thinking about the mechanics of this, focussing only on what kinds of information might be relevant. But I thought that one of our goals was to establish how evidence (of whatever type) can be transmitted in coordination with the History. I don't expect *all* of this information to be made available or to be relevant in every case. I simply want to establish what *might* be considered relevant and/or what is | 08:17 |
| reiterative | already used in existing approaches to the trustability problem. | 08:17 |
| reiterative | I'm really happy that you are asking these questions :-) | 08:18 |
| reiterative | My goal here is to re-examine the elements that were identified in the original hypothesis - I don't yet have a coherent v2! | 08:19 |
| persia | I don't think it useful to think about how to transmit information in coordination with the History. That is diificult, and ends up proscribing mechanism. | 09:01 |
| persia | My preference is to think about what information is desired *within* the History, and accept that any Consumed History is only an assertion of any data within the History. | 09:02 |
| persia | That a Consuming Identity can frequently Consume History, and then compare current History to past History is one of the ways to validate the integrity of the assertion. | 09:03 |
| persia | Essentially, for the new list of words, I expect a consumer to either not have the information or have no way to validate the information, which makes metrics pointless (as trivially gamed). | 09:04 |
| persia | Whereas, if we instread construct metrics based on history (or comparison of histories: whether histories of the same or complementary artifacts is immaterial), then it becomes easier to say "this history is internally consistent, consistent with past snapshots, and consistent with complementary histories". If that is the case, it is easier to accept the assertion that it is a true history. | 09:05 |
| persia | If it is a true history, then we can use data stored in that history to assert qualities of the software. | 09:06 |
| *** toscalix has joined #trustable | 09:06 | |
| *** toscalix has quit IRC | 10:58 | |
| *** toscalix_ has joined #trustable | 11:02 | |
| *** traveltissues has joined #trustable | 11:40 | |
| *** traveltissues has quit IRC | 11:42 | |
| *** traveltissues has joined #trustable | 11:42 | |
| *** traveltissues has quit IRC | 11:47 | |
| *** traveltissues has joined #trustable | 11:47 | |
| *** traveltissues has quit IRC | 12:00 | |
| *** traveltissues has joined #trustable | 12:00 | |
| *** traveltissues has quit IRC | 12:40 | |
| *** traveltissues has joined #trustable | 12:47 | |
| *** milloni_ has joined #trustable | 14:03 | |
| *** milloni has quit IRC | 14:09 | |
| *** toscalix_ has quit IRC | 17:37 | |
| *** milloni_ is now known as milloni | 17:45 | |
| *** traveltissues has quit IRC | 20:44 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!