# Metrics adventure (aka Meeting-log) December 7, 2021

Update LNMetric adventure December 7, 2021.

Reference to last meeting log: [November 22, 2021](https://hackmd.io/@lnmetrics/metrics_adventure1/edit)

Partecipants: Vincent, Dott. Carlo Sartiani.

### Table of Content

- [x] Where we are
- [x] [Medium blog post](https://vincenzopalazzo.medium.com/introduction-to-ln-open-metrics-96a7c859f4e2)
- [x] [Mainlist announce](https://lists.ozlabs.org/pipermail/c-lightning/2021-November/000213.html)
- [x] Anonymus Person Metrics ([clboss](https://github.com/ZmnSCPxj/clboss))
- [x] System architecture
- [x] Possible direction of the thesis
- [x] :bulb: Estimation of the goodness and choose what is the best connection for one node?
- [x] Group Meeting before the end of the year
- [x] Appendix

### Meeting discussion

:pushpin: Where we are

#### :speech_balloon: Medium blog post
__Nothings to report here.__

#### :speech_balloon: Mainlist announce
__Nothings also here__

#### :speech_balloon: Anonymus Person Metrics

There are some metrics that clboss does, like the metrics discussed in the following post available at the following [link](https://github.com/ZmnSCPxj/clboss/issues/81#issuecomment-984179078).

However, when I start a clean node without any connection and any knowledge of the nodes there are on the network clboss need to rely on a third party that offers the best node for some definition of best, in the network. With LN Open Metrics is possible we can solve this problem and propose some suggestions to this new node. We can make metrics by filtering the options by the user input.

:pushpin: Possible direction of the thesis

#### :speech_balloon: Summary

Having a unique system to collect data like LN Open Metrics https://github.com/LNOpenMetrics give some flexibility to propose
a way for all the people that want to make research that involves data collection to add it in a well-tested system like the system proposed in the [c-lightning mainlist](https://lists.ozlabs.org/pipermail/c-lightning/2021-November/000213.html).

In addition, proposing some metrics to involve the user to share the data with a server sounds good, and maybe we need to give to the user the ability to express some preference or give some input to the server to get some result customizable from the user viewpoint.

If there is time, with the previous idea we can think also to implement some additional plugins that query the server and start to bootstrap the channels connected with the metrics and user input help.

There is also an additional problem that is not in the range of the Master's thesis but would be cool to have some proposals like:

- How we can prove that the data received from the node are the real data? or are not modified from the user to be the best node in the world.
- How we can manage the metrics if the node doesn't have enough data to calculate them? e.g: We want to know the forward rating but the node doesn't forward nothings because it was not active, in this case, what does this event mean? how do we need to classify this event?

:pushpin: Group Meeting before the end of the year

We try to set up something in January.

__The appendix is not logged__