[03:41] *** melezhik joined [03:58] *** melezhik left [08:30] *** xinming left [08:31] *** xinming joined [13:03] *** Nemokosch joined [13:07] *** Nemokosch left [15:08] I'm not sure anymore it's a bug in Rakudo. The observation is that a request is emitted on the supplier of the cro request -> response pipeline. The supply in the first processor of the pipeline (RequestSerializerExtension) receives the 'emit' call, but is itself not tapped. Thus the request stalls (it's a preserving supplier, but the supply is never tapped). [15:10] According to the docs the code in supply blocks is called when the supply is tapped. The pipeline is basically built up using supply / whenever blocks. So as long as one of the supplies in the chain isn't tapped all upstream supplies aren' either. [15:12] *** rbt joined [15:20] *** rbt left [15:20] *** rbt joined [15:30] So I'll have to think up a way to find the one non-tapping processor... [16:20] *** rbt left [16:20] *** rbt joined [16:55] *** rbt left [16:56] *** rbt joined [17:53] *** Nemokosch joined [18:32] *** Nemokosch left [20:23] *** japhb left [20:30] *** japhb joined [20:55] *** n1to joined [22:45] *** n1to left