feat(bundle-cost): reduce by 100x as in mainnet 61 #75
Conversation
| @@ -0,0 +1,13 @@ | |||
|
|
|||
| lower-bundle-cost.json: | |||
| jq '.tx.body.messages[0].content | .changes[0].value |= fromjson' tx-61.json >$@ | |||
There was a problem hiding this comment.
On first read I thought this was able to build dynamically from the ARCHIVE_NODE.
Please comment how this Makefile is intended to be used.
Incidentally, I suggest that lower-bundle-cost.json be produced during the build since it's a fast local operation to create from tx-61.json
There was a problem hiding this comment.
Please comment how this Makefile is intended to be used.
I added some comments. I don't think I quite covered this though. Feel free to add more or discuss it or something.
I suggest that lower-bundle-cost.json be produced during the build ...
I don't have a good sense of how the work should be partitioned in EVAL vs other stuff, so I'll leave this to you.
| clean: | ||
| rm -f lower-bundle-cost.json | ||
|
|
||
| realclean: clean |
There was a problem hiding this comment.
why two cleans? please comment
for naming, it's odd to imply that the other isn't real. consider clean-all
but I'd prefer to have lower-bundle-cost.json be produced in the build, in which case there'd be only one target and one clean here.
There was a problem hiding this comment.
realclean vs clean is a gnu convention
The difference in this case is: re-fetching from the network is expensive / wasteful while I was rapidly iterating on the jq step.
Will comment...
proposals/61:lo-bundle-cost/Makefile
Outdated
| lower-bundle-cost.json: | ||
| jq '.tx.body.messages[0].content | .changes[0].value |= fromjson' tx-61.json >$@ | ||
|
|
||
| hash=6EA7B46DFDD0973338BB50781B33D651FE0E8EE8ABF4D7BCB1BF8235334AAFE0 |
There was a problem hiding this comment.
very cool to pull from an archive node.
can we make this easier to work with? e.g. a default ARCHIVE_NODE value?
how does determine the correct hash value?
There was a problem hiding this comment.
I added the relevant query commands to go from the proposal number 61 to the tx hash. Overkill?
dckc
changed the title
turadg
force-pushed
the
dc-bundle-cost
branch
from
9499a3d
to
0ad4642
Compare
fixes #28