by

Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Chris Travers
    @einhverfr
    working on the autodecode of json
    autoencode now works but there are some backwards incompatibilities
    Jean-Yves Sireau
    @jy-binary
    don’t worry about backwards incompatibilities - this module isn’t presently being used by anyone
    Chris Travers
    @einhverfr
    binary-com/perl-pubnub-pubsub#6
    one thing that will need to be documented is that we use Mojo::JSON for encoding and decoding and so booleans need to be set as that module expects.
    Chris Travers
    @einhverfr
    any further feedback on the pull request?
    Otherwise I am thinking I will go ahead and install it locally, then bump up the version number and merge once I verify that in practice it is solving the Couch->Pubnub problems
    Fayland Lam
    @fayland
    let me take a one more look
    wait a sec
    Chris Travers
    @einhverfr
    ok, I am trying to get my cookie refactor working on qa while waiting, so no worries.
    Fayland Lam
    @fayland
    could you help document the Message module also?
    Chris Travers
    @einhverfr
    yeah, sure. Thanks. Will do right away. (sorry intended to yesterday and forgot)
    Fayland Lam
    @fayland
    others seems all good. please go ahead to merge and release.
    Chris Travers
    @einhverfr
    I am having one final prolem which I should be able to fix quickly.
    It looks like Mojo takes the json string and adds extra escaping
    so I get json out on the other side like "{\"365\":0.285,\"240\":0.326,\"120\":0.177,\"90\":0.109,\"210\":0.286,\"330\":0.411,\"7\":0.058,\"60\":0.098,\"180\":0.151,\"30\":0.081,\"270\":0.354,\"150\":0.216,\"300\":0.384}"
    (I expect without the surrounding quotes or the backslashes)
    Fayland Lam
    @fayland
    you mean there is extra \" when encode?
    Chris Travers
    @einhverfr
    yeah
    I am working on it
    Chris Travers
    @einhverfr
    It may just be an issue with the PubNub console. I am going to consider it working for now.
    Fayland Lam
    @fayland
    good job. yo can go release I think. :)
    I see you have warn in the code, FYI
    Chris Travers
    @einhverfr
    fixing it
    also one issue I found on my tests, fixing that too
    if I dont see anything else in my extended manual testing will release tomorrow. I think those backslashes are an artifact of the console
    Fayland Lam
    @fayland
    yes, I believe so
    Chris Travers
    @einhverfr
    turns out it was not a console problem but me double encoding (sending in JSON so it would be double escaped)
    Chris Travers
    @einhverfr
    I am thinking of bumping the version number on the next release of this to 1.0.0 (and making backwards compatibility guarantees). Thoughts?
    Chris Travers
    @einhverfr
    ok, with no objections, merging the version number change and releasing 1.0.0
    Pavel Shaydo
    @trinitum
    Is there any reason to hurry with pushing 1.0.0? I usually wait a month - a couple of years before calling thins stable.
    Fayland Lam
    @fayland
    actually I guess nobody cares if it's 0.10 or 1.0.0. :)
    Pavel Shaydo
    @trinitum
    yeah, that's true ;)
    Chris Travers
    @einhverfr
    @trinitum, I like to move to 1.0.0 when I am ready to say "this API spec will not change without bumping up the first number." 0.x means no guarantees. 1.x means guarantees. That may just be me though.
    Chris Travers
    @einhverfr
    #7 is just a set of doc fixes, can be in 1.0.1 (fixes the synopsis so it works when copied and pasted)
    Chris Travers
    @einhverfr
    also one more thing I would like to change.
    Chris Travers
    @einhverfr
    if a connection fails on connect, we recurse immediately. I would like to eliminate this tail call for memory and stack management reasons.
    Chris Travers
    @einhverfr
    ok, I have a pull request for subscribe_multi functionality. It is tested using the examples/subscribe_multi.pl script. Are there any comments about the API? This allows us to publish one feed per type per instrument and then handle them on our side.
    Otherwise I plan to merge it and release 1.1.0 this pm.
    binary-com/perl-pubnub-pubsub#12
    (again goto coderefs are lateral call stack movement, not a traditional goto)
    Chris Travers
    @einhverfr
    ok 1.1.0 is out