Notes from AWG group meeting on long term agenda


Mic Bowman
 

During the AWG call this morning we talked the "long term agenda" and our experiences. Some notes from that meeting (only partially pulled into cogent thoughts and I'm explicitly NOT drawing conclusions just providing content for the discussion tomorrow):

Observation: hyperledger is not a "wild garden", its more like a bunch separate greenhouses with very very thick glass walls.

Observation:  For open source developers, it is hard to justify work on common components when there is little or more realistically no hope or commitment that the component will be integrated by the platforms. To encourage broader participation by developers, we have to show that their work potentially matters.

Observation: Given the business motivations, a low cost solution that is specific to one platform will always win out over a more general solution that requires investment in customization/integration for the platform. That is, the cost bar is high for moving from "what I have right now that works OK" to "a general solution that might be useful far in the future".

Observation: If you want a component integrated into one of the platforms, you build the component for that specific platform and you engineer the integration yourself.

Some examples we talked about this morning (this is not intended to be a complete list, just ones that we had some experience with):

* Ursa: adopted by Indy, not be Fabric, if we had "infinite resources", what would the Ursa team need to do in order for the Fabric maintainers to adopt Ursa?

* Transact: Fabric appears to be sticking with its existing chaincode execution model rather than adopt Transact (at least for the moment), the cost of integration and the transition to the new model is not justified by the benefits of moving to the common module.

* DAML: appears to be available for several platforms, DAML developers (basically) integrated it into each of the platforms.

* PDO --> FPC --> Fabric, PDO --> Avalon: explicit goal was NOT to make a common technology module, but to "answer questions" that could inform other projects. Integration into Avalon and Fabric is almost a complete re-write from a code perspective (with the corresponding costs to each of the projects).

* Interoperability Lab: Highly modular plugin architecture (goal is to deliver value without requiring the underlying platforms to change). 

--mic


I’m assuming we will only have a small amount of time after the Solang discussion. Our goal is to *start* the framing so we can have some productive discussions in HGF and at the next TSC meeting. My assumption is that this is a long term item that will take us a few sessions to frame and several sessions longer to complete.

 

Join tsc@lists.hyperledger.org to automatically receive all group messages.