DataflowBroadcast around DataflowStream[?] error

[FelipeMSD]

Dear all, I don't know why this is happening now, because I didn't have this problem in the past, but when I use variable=Channel.of("string") and then variable.view() Nextflow returns me **DataflowBroadcast around DataflowStream[?]** . Has this ever happened to you? I'm even using nextflow.enable.dsl=2 in the begining of my main.nf file but it seems to not help...

[bentsherman]

Can you share the exact script you are running and your Nextflow version?

[FelipeMSD]

Hi @bentsherman . thanks for the prompt reply. For now I cannot share it but I made few tests and it seems the problem is with .map{[]}, as now I also had this message: WARN: Cannot serialize context map. Cause: java.lang.IllegalArgumentException: Unable to create serializer "com.esotericsoftware.kryo.serializers.FieldSerializer" for class: jdk.internal.reflect.DelegatingMethodAccessorImpl -- Resume will not work on this process. When I run variable.view() I can see the values but when I try x=variable1.map{[variable2]} and then x.view() what I have is: [variable1_string, DataflowBroadcast around DataflowStream[?]]. Any insight on this? ChatGPT told me to update Nextflow, which I did; check if my java version is higher than 8, which it is, and also try disabling the Kryo serialization in Nextflow by adding process.executor = 'local' to my script. Unfortunately, it doesn't seem to help me... It only happens for the Channel.of() channels. The file() channels appear on my .map{[]} list correctly.

[FelipeMSD]

I've been doing some tests here and I decided to call the parameter inside the .map{[]} array as x=variable1.map{[params.variable2]} instead. It's working this way now. 😉

[bentsherman]

Is variable2 another channel in your workflow? If so that is likely the problem, you cannot use channels inside the closure of a channel operator like that

[FelipeMSD]

I see... So I'll keep calling variable2 this way: x=variable1.map{[params.variable2]}. Thank you very much for the insights ;)