Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Noel Maersk
    @veox
    So, basically, wait 'till the tx got included for the first time, then check if the block became an uncle.
    Mandar Vaze
    @mandarvaze
    I'm getting "Transaction Failed" error when testing my code.
    I have a basic "Token" contract, very similar to one listed at https://www.ethereum.org/token.
    The tests work for balanceOf() mintToken() as well as burnFrom()
    Only transferFrom is failing.
    How can I troubleshoot ?
    @veox I know there is truffle.js that contains the configuration.
    Paul Tiplady
    @paultiplady
    Thanks Noel -- RE: uncles I'm just trying to figure out how to integration-test my code, since Rinkeby doesn't seem to get uncles (I could be wrong there). Can I manually mine blocks on a local tester chain to produce an uncle'd transaction?
    I'll continue to dig and report back, it seems like there should be a way.
    Piper Merriam
    @pipermerriam
    @mandarvaze you should check out https://github.com/carver/ethtoken.py
    doesn't implement the non ERC20 methods you listed but might be a good reference
    Paul Tiplady
    @paultiplady
    What does a tx receipt look like if the block was uncled? Presumably some of the fields get nulled out.
    Piper Merriam
    @pipermerriam
    @paultiplady I think those two concepts are disparate. Receipts only come from transactions that are mined in blocks on the main chain of blocks. The transactions in uncles are not actually part of the state, they don't have receipts, and they may have either already been included in a previous block, are included in the current block, or have not yet been included in any block.
    Noel Maersk
    @veox
    Yup! Uncles ("ommers" in yellowpaper) can be described as references, or valid headers; their bodies are not present. If you got a receipt (for a txhash) once (indicating inclusion in block X), waited, fetched it again (for the same txhash) - and either got a different receipt, or no receipt at all - then you could reason that block X has become an ommer.
    Mandar Vaze
    @mandarvaze
    @pipermerriam OK. There are not enough test though :(
    Anyway, I was looking for troubleshooting tips.
    avatar-lavventura
    @avatar-lavventura
    Hello how could I solve this error on web3 side: web3: TypeError: 'toBytes' is not a function
    Noel Maersk
    @veox
    @avatar-lavventura: see if you didn't mean web3.Web3.toBytes().
    Depends on what web3 is.
    avatar-lavventura
    @avatar-lavventura
    I was using on the node side, i think web3 on nodejs does not have that function. When I tried on web3.py it worked.
    Jason Carver
    @carver
    I have not seen an equivalent function in web3.js. There's not an exact equivalent to python's bytes in javascript, that I know of, so they tend to use hex-encoded strings everywhere.
    Mandar Vaze
    @mandarvaze
    @pipermerriam It turns out there was a "bug" in my code. Once I fixed it, the tests started passing.
    Thanks for the awesome populus
    :thumbsup: :clap:
    Michael Egorov
    @michwill
    Hi @pipermerriam , somehow getting populus.contracts.exceptions.BytecodeMismatch: None of the known addresses matched the expected bytecode when using populus 2.2.0 and solc 0.4.20 (0.4.19 is fine). Is it something to expect?
    (when trying to get_or_deploy_contract)
    Mandar Vaze
    @mandarvaze
    @pipermerriam Is there a plan to "bring back" migrations ? I saw it was removed some time ago.
    What is the substitute, if there is no plan.
    Thanks.
    Mikko Ohtamaa
    @miohtama
    @mandarvaze What kind of goals you are trying to accomplish?
    Mandar Vaze
    @mandarvaze
    @miohtama "Migrate from truffle to populus for the smart contract development" - hence looking at parallels. That is all.
    I just got started using populus.
    Piper Merriam
    @pipermerriam
    @mandarvaze yes no official timeline. There's some hiring I'm working on getting done to be able to support/develop populus more actively
    Mandar Vaze
    @mandarvaze
    @pipermerriam Why was this "feature" removed ?
    Piper Merriam
    @pipermerriam
    It was fundamentally broken. Migrations cannot be global. They are chain specific.
    And even then, they aren't exactly chain specific because some may/may-not need to be appied to multiple chains.
    My plan for re-implementing it is going to look more like a framework for writing migration scripts. Each script will be executable on it's own, or chained with other scripts, executed multiple times on the same chain (for some kind of utility script), etc.
    Bryant Eisenbach
    @fubuloubu
    off-chain migrations? or on-chain like truffle? with off-chain you can write your own migration tracking contract like truffle, I personally don't like the additional overhead they add there.
    Mandar Vaze
    @mandarvaze

    I get the following error when I run populus chain reset test1 where test1 is a chain I created using populus chain new test1

    AttributeError: 'Project' object has no attribute 'get_blockchain_data_dir'

    Is this known issue, or Do I have a bad configuration ?

    my project.json is a copy of ...lib/python3.6/site-packages/populus/assets/defaults.v8.config.json (without this my py.test does not work)
    Noel Maersk
    @veox
    @mandarvaze Sounds like you're using v2.2.0. Perhaps report this issue on the tracker, and try v2.1.0?.. (You'll also have to change version of the config from 8 to 7 in project.json.)
    I haven't used populus chain myself, so don't know if that issue is present in Populus v2.1.0.
    Mandar Vaze
    @mandarvaze

    @veox I checked "git blame" on that file, looks like it hasn't changed in 2 yrs. So I have a feeling the problem would be there even in 2.1.0

    I'll open an issue.

    Mandar Vaze
    @mandarvaze
    Can someone help me understand why ethereum/populus#444 pull request failed ? It is a REALLY simple change.
    I clicked on the travis-ci link, but I still don;t understand what is wrong (or how to fix it)
    I think it is unrelated to the change. There seems to be some version conflict related to eth-utils in the build logs (on travis ci)
    Noel Maersk
    @veox
    @mandarvaze Yes, populus can't use eth-utils 0.8.x, but that's not specified in its requirements. Some dependency pulls in eth-utils 0.8.x, so populus breaks.
    eth-utils 0.7.4 is one that works for me.
    Noel Maersk
    @veox
    Populus currently lacks a project maintainer.
    Bryant Eisenbach
    @fubuloubu
    @veox I think for right now @pipermerriam and @carver are focusing on updates to web3.py (releasing a major version soon), but I believe they will be supporting this again when that seems stable. Correct me if I'm wrong!
    Piper Merriam
    @pipermerriam
    Yes, I've been avoiding making any announcements on the topic because I don't have a timeline I can commit to. My team is growing and we should have the resources necessary to properly support this tool in the near/mid-term future
    Bryant Eisenbach
    @fubuloubu
    awesome, congrats on more resources!
    Mandar Vaze
    @mandarvaze
    @veox I didn't find eth-utils mentioned in the requirements-dev.txt ? Where should the specific version be mentioned ? Seems like minor change, I can fix it, if OK.
    Found eth-utils>=0.7.1 in setup.py ? Should we change it to eth-utils==0.7.4 ?
    Mikko Ohtamaa
    @miohtama
    @mandarvaze yes
    or eth-utils<0.8
    you can test it by doing a pull request and seeing that Mr. Travis happily churns through it
    or run tox locally
    Mandar Vaze
    @mandarvaze
    Done. See ethereum/populus#453
    travis-ci is in progress - at least two tasks (?) passed, which were failing earlier.
    Noel Maersk
    @veox
    :sparkles:
    Mandar Vaze
    @mandarvaze

    Travis CI successfully completed for #453

    Is there a way to rerun just the travis-ci for #444 (and other PRs like #440 and #445 that had failed since Feb 10th - I think)