Update go-ipld-prime to the era of NodeAssembler.#31
Conversation
Codecov Report
@@ Coverage Diff @@
## master #31 +/- ##
==========================================
- Coverage 55.91% 55.83% -0.09%
==========================================
Files 3 3
Lines 279 283 +4
==========================================
+ Hits 156 158 +2
- Misses 88 89 +1
- Partials 35 36 +1
Continue to review full report at Codecov.
|
ribasushi
left a comment
There was a problem hiding this comment.
Functionality looks good to me. You may want to go mod tidy - there will be a notable shrinkage of go.sum as a result ( go.mod should remain intact )
|
Fascinatingly, that:
I've committed the results. However, these behaviors strike me as not so much minor bugs as they are boggling infirmities of core concepts -- seeing a system that appears to use hashes, but also does so in such inconsistent ways that one begins to wonder if those hashes are being checked or actually mean anything is deeply disquieting and something I have an especially strong allergy to. I formally give up trying to reason about go mod. It does not spark joy. |
Update go-ipld-prime to the era of NodeAssembler. This commit was moved from ipld/go-car@c222d79
Bumps dependencies and updates a few interfaces and package names that have shifted lately.
go moddid a lot of other things that make very little sense to me -- the two commands I ran werego get github.com/ipld/go-ipld-prime@masterandgo get github.com/ipld/go-ipld-prime-proto@master, and how that causedgo-ipld-formatto show up in the update effects (it's a test dependency only of a transitive lib, afaict?), or why lots of other things appear in the the 'go.sum' file that are even more distantly (un)related and also get many versions recorded which we haven't actually handled... I... I don't know. Someone may want to sanity check that; it's clear I simply don't understand whygo modis trying to do, and probably disagree with it in any case. But tests pass; so, I think the update overall is passable.