# Week 18

## Recap
Compared to the previous weeks, this week was definitely a lighter week in terms of actual implementation work. Most of this comes from the transition from working on purely Portal Wire communication systems to now implementing the History Network. I wanted to make sure that everything I was going to do was in line with the same ideas that Meldsun had, both to avoid overlap and to create a congruent system. We had a meeting at near the beginning of the week where we decided that I would work on implementing the FINDCONTENT, CONTENT, OFFER and ACCEPT packet logic for the History Network. I spent a good portion of my work this week just planning my approach on how I intended to tackle this. I ended up starting off my implementation with the outbound packet requests, meaning the sending of FINDCONTENT and OFFER and the recieving of CONTENT and ACCEPT responses. I was able to get the design down and implemented the outgoing portion in the HistoryNetwork class. One of the larger struggles had to do with appropriate planning on where I was going to handle the actual packet information processing for the History Network. Where I could either do so all in one place within the HistoryNetwork class, or outside in a new layer meant for such processing. I still haven't quite decided but I hope to sync ideas with Meldsun and maybe see what direction we would both like to go sometime this week.

## Learning
My learning this week was somewhat just general network designs, as well as continued reference to the Portal Wire Specifications for working on outbound communication.

### Concepts

#### Portal Wire Specifications
I mainly focused on the FINDCONTENT, CONTENT, OFFER and ACCEPT packets in the specifications. This was used to ensure that the clients outbound communication followed the appropriate steps. It was also useful in looking ahead to which systems are needed that aren't implemented yet. For the mean time they can then just be abstracted as an interface so that the logic associated with their use can be added.

##### Resources
1. https://github.com/ethereum/portal-network-specs/blob/master/portal-wire-protocol.md
2. https://github.com/ethereum/portal-network-specs/blob/master/portal-wire-test-vectors.md

## Tasks
This week had a bit less in regards to implementation work and more so focused on the design and planning of how to fit my FINDCONTENT and OFFER outbound systems into the HistoryNetwork class. In this way I could also potentially save some time down the road when needing to refactor or reorganize any systems I make. After solidifying my idea on how to go about implementation I went ahead and added the HistoryNetwork class implementations for findContent() and offer(). Both of which follow a similar design to the ping() and findNodes() methods that Meldsun had worked on. Additionally, like with the NODES response packet, the CONTENT response packet when containing the sub-type with selector 0x02 for an ENR list also needs to be filtered for the recieving nodes ENR and the responding nodes ENR before being processed. This part is still not totally complete as the HistoryNetwork class design is still partially preliminary, but one such way of doing it would be to keep a copy of the local nodes NodeRecord and the NodeRecord of the node recieving the FINDCONTENT packet, and then just using those values to look through the CONTENT response packet once recieved. Outside of that I also had to add basic database opperations as an interface for the FINDCONTENT packet once it is parsed. This way even though the database still needs to be developed, the general logic can still be implemented when working on content received in the CONTENT response packet. I will likely have to do something similar with a uTP interface for when the content size is too large in the CONTENT packet.

## Week 19 Plan
Heading into Week 19 my goals are much the same as this week. Except I am looking to get more done in terms of actual implementation work. I am hoping to continue in maintaining close communication with Meldsun regarding the outbound client packet design for the History Network. This should reduce the need for any refactoring or redesigns and should allow both of our contributions to outbound packet design to work together without any issue. My main goals are:

- Continue working on the FINDCONTENT and OFFER outbound packet communication for the History Network
- Potentially get into a proper database storage solution to build on the basic interface
- Add some additional logic to a layer previous to the HistoryNetwork class if needed for the management and use of the database interface
- Start designing the incoming communication method for FINDCONTENT and OFFER packets

We have decided on an interface design path where any additional systems that may be needed are to just be abstracted into an interface that can then be built on top of at a later time. This should speed up the general protocol implementation a bit and also prevent us from being restricted by any systems that just need implementation.