Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    jonathanindig
    @jonathanindig
    Glad to hear it! Please keep sending feedback our way, we really appreciate it.
    Scorch looks pretty neat as well, I’ll check it out :smile:
    raproth
    @raproth
    Is there something like opening notebooks in read-only mode?
    Frank Dekervel
    @kervel
    @koen-dejonghe coole notebook!
    (and i didn't know about numsca, another cool discovery)
    Koen Dejonghe
    @koen-dejonghe
    thank you
    Koen Dejonghe
    @koen-dejonghe
    @raproth If you just want to show off your notebook, then you could publish it on github (with the correct language tag as described above) .
    raproth
    @raproth
    @koen-dejonghe We often share URLs to notebooks to team-collegues to show us our work. In such cases we don't want the the recipient of the URL is able to accidentally modify (break) the notebook. spark-notebooks has such a feature ("read-only link")
    jeremyrsmith
    @jeremyrsmith
    You can get readonly, but you have to have some authentication enabled so it can tell what the permissions should be
    João Costa
    @JD557
    Is developing/testing the front-end easier with #839?
    Can you just compile the distributed frontend and everything works?
    jonathanindig
    @jonathanindig
    The way we develop / test the frontend is by running the server with the --watch flag
    and then npm run watch inside the frontend folder
    does that not work for you?
    João Costa
    @JD557
    I don't know, I was just curious :P
    jonathanindig
    @jonathanindig
    oh i see :) the changes in that PR mostly help reduce load times. now that everyone is working from home over here we’re more sensitive to that
    jeremyrsmith
    @jeremyrsmith
    @JD557 but yes it is easier :grinning:
    at least it's easier to make changes that you have to test on a remote machine because they're environment-specific features.
    It also gives a path for people to provide their own hacked frontend if they want, though I don't recommend it...
    but the main thing was that everyone is working from home now and the initial load of polynote is rough. This way we can have nginx serve all the static files directly, and it can serve them gzipped and do all its fast good stuff and it's much quicker to load
    I imagine most people are in the same boat now :grimacing:
    João Costa
    @JD557
    you guys have been shipping a ton of new features latelly.
    Is it related to everyone working from home, or is just a coincidence?
    (if that's OK to ask)
    jeremyrsmith
    @jeremyrsmith
    I think there was kind of a pause in releases for a while and then we did a fairly big release... the pace has been more steady than it seems, just didn't release regularly
    then did a release, noticed some problems, did another release, etc :grinning: that's how it goes sometimes
    Koen Dejonghe
    @koen-dejonghe
    hi, is there an example of a polynotebook that uses vega somewhere?
    I mean scala populating a vega spec
    jeremyrsmith
    @jeremyrsmith
    @koen-dejonghe you mean you want to write the vega spec but use data from Scala? We don't really have examples of published AFAIK, but if you use the plot UI it just generates a Vega Spec cell when you save the plot. It just has "Hide code" turned on; turn it off and you can see how it works
    Regrettably the error messages given by Vega Spec cells are pretty useless, so I imagine it would be challenging to write a Vega Spec cell from scratch. We should see if there's anything that can be done about that.
    but then with Vega
    jeremyrsmith
    @jeremyrsmith
    Yeah, we don't have one. But we really should, come to think of it.
    Will work on adding one.
    Unless you want to do that and make a PR? :grinning:
    Koen Dejonghe
    @koen-dejonghe
    yeah I could do that, but it's not clear to me how I can achieve that. Do I have to generate a json spec and then ...
    not sure how to proceed
    what is the plot UI?
    jeremyrsmith
    @jeremyrsmith
    Oh, if you have like a Seq of case classes, you can click on it in the symbol table and go to Plot. Or if there is a cell output that's a Seq of case classes, there is a "browse" and "plot" button in the output area where it shows the schema
    Koen Dejonghe
    @koen-dejonghe
    Nice. I didn't know about this.
    João Costa
    @JD557
    Is there anyway to easily plot two timeseries? Say, I have val ts1: List[Entry] and val ts2: List[Entry]
    I guess I could create a case class Foo(time: Long, entry1: Int, entry2: Int)...
    João Costa
    @JD557
    mmm, I might have found a bug in the bar chart... for some reason I can plot multiple lines, but multiple bars returns an "undefined" error
    João Costa
    @JD557
    finally sent a PR... building the dev docker image with a slow internet is a pain... :|
    Koen Dejonghe
    @koen-dejonghe
    @jeremyrsmith I have used matplotlib, which I think is also a very interesting feature of polynote. If you want, I can make a PR with the notebook.
    raproth
    @raproth
    is there a way to increase the log-level of the notebook? Especially during dependency resolution (jvm/scala)?
    raproth
    @raproth
    I'm especially interested to see some verbose output like from which repo an artifact was taken, and which timestamp was used (in case of SNAPSHOT artifacts)
    Anton Kulaga
    @antonkulaga
    Hey guys, I see that in 0.3.5 the host configuration does not work ( polynote/polynote#851 ), did you change anything in configuration format? I want to configure it for 0.0.0.0
    jeremyrsmith
    @jeremyrsmith
    @antonkulaga thanks for the heads up, config format for listen was certainly not changed intentionally. You should see log output when you start up that dumps the config, anything interesting there about Listen?
    Ahh, I think I see what happened. Will put up a PR.