This report summarizes the work on Hydra since October 2023. It serves as preparation for the monthly review meeting (see slides and recording), where the team updates project stakeholders on recent developments to gather their feedback on proposed plans.

Roadmap​

This month, several items were restructured on the project roadmap:

The latest roadmap with features and ideas

Notable updates​

• Release of version 0.14.0 is imminent as most of the stateless chain observation #1096 feature is done and only some internal refactoring missing.

• The user-submitted feature idea on supporting extended keys, #1141, has been refined. The purpose of this item is now clear, and although it's not currently a priority, work on it can be started.

• Two new items were created to improve the developer experience of the Hydra build toolchain #1146. Additionally, a follow-up on things learned by hydra-poll aims to create a full end-to-end example of a two-player game in a Hydra head #1098.

• There was already an item about reacting to main chain parameter changes on the roadmap, which included not only protocol parameter updates, but also hard-fork events. With the upcoming hard fork into Conway, the team split off the changes needed to navigate the hard fork and clarified them:

  • Conway support #1177 will make the hydra-node support Conway and allow users to keep heads open across the hard fork.
  • Drop Babbage support #1178 will eventually drop support for the Babbage era and retain maintainability.
  • The original item is only about reacting to protocol parameter changes #195 now.

• Substantial updates were made to the design of the next major feature Incremental decommit #1057, and work on it can be started.

• In general, the team prepared several of these 💭 idea items and many of them will turn into 💬 features soon.

Hydra development​

Issues and pull requests closed since the last report

This month, the team worked on the following:

MeshJS + Hydra​

At the beginning of month, the team attended the Cardano Summit in Dubai where they held the Hydra MasterClass workshop. There they also presented the Hydra-Poll dApp which is using MeshJS to build Hydra transactions on the frontend.

The Hydra team collaborated closely with the MeshJS team and as the outcome of this work The MeshJS team added some Hydra specific features to the MeshJS SDK. There is a nice low level API for building custom transactions which is pretty similar to what cardano-cli does.

Now you can use isHydra field which you can pass to this API which then uses another set of protocol paramters suitable for Hydra (eg. usually in Hydra we want to have zero transaction fees, reduce script execution units etc.).

Big thanks to the MeshJS team on giving much needed support. They were presenting these changes in the monthly review meeting so make sure to watch the recording.

Rendering the Hydra specification​

Besides working on the design of incremental commits itself, the team also tidied up the way the Hydra protocol specification is written.

So far, the specification was written in LaTex and rendered to a PDF. LaTeX was a good choice because of its expressiveness on math expressions, but not so great a choice for pull requests and the entry barrier is quite high. The majority of documentation around Hydra and Mithril is written using Markdown and rendered into a webpage using a static site generator, currently Docusaurus. Most modern Markdown renderers do support LaTeX-style math nowadays, even the normal GitHub file preview.

Hence, the team experimented with using the more standard Markdown format to write the specification, but still being able to use LaTeX math and LaTeX macros (which are still heavily used in the spec). The aim is to have both, a PDF and HTML rendering.

While not yet finished, the GIF above shows the progress being made using pandoc to resolve macros, producing a PDF via xelatex and Docusaurus compatible markdown. Using powerful extensions of pandoc one can even convert TikZ, GraphViz and Mermaid into both target formats.

Community update​

Beside the great work done by the MeshJS team towards supporting Hydra in the browser, here are some news and updates on community-led work streams related to Hydra:

• SundaeLabs has made good progress on their offline mode PR which should hit the main branch soon,

• TxPipe has been collaborating with IOG's creative engineering team to build Cardaminal, a command-line wallet, and demonstrated it during the monthly review meeting. Cardaminal is built in Rust, leveraging Pallas libraries, and is aimed at being a fully-fledged scriptable wallet providing users with powerful stateful transaction edition capabilities, native assets management, chain synchronisation, etc.

  The plan is to make Cardaminal compatible with Hydra - to be able to use it to interact with a head - and Mithril - using snapshots to synchronise the wallet.

• The team behind ZK Snarks in Hydra Catalyst proposal is making good progress and should be able to give a demo at the next monthly review meeting

Conclusion​

The monthly review meeting for November 2023 was held on 2023-11-27 via Google Meet with these slides and the recording.

The Hydra team is thankful to have again had so many great demos from the numerous projects in and around Hydra and Mithril projects - we should really schedule it to be one hour planned duration. Also, if you have not watched it already, make sure to check out the recording, the written report is merely expanding on a couple of topics but cannot render the cool demonstrations.

At first look, the demonstrated projects may appear not related - we even had a minimal command-line wallet on the show after all. But as we went on, these impulses spark discussions and it becomes clear that all of these projects are connected.

For example, the terminal-based wallet 'Cardaminal' as it was presented the first time in public this month, could itself become a Mithril client for faster synchronizing or even light-mode using the recently published Rust client library for Mithril. Furthermore, such a minimal wallet would be a great testing ground for a Hydra-aware wallet. The MeshJS integration also has further potential of Hydra specific fetcher and submitter providers.

On the other hand, we also stumbled over not-so-optimistic sentiment even in much more mature scaling solutions like the Bitcoin Lightning Network. This tweet summarizes that creating non-custodial [scalability/LN] solutions 'feel like the most thankless thing in the world'.

It is an exciting ride still, but events like this monthly review meeting with its colorful demonstrations of small steps to improve the usability and scalability of Cardano piece by piece with often unknown synergies is also empowering. This month again made us feel not 'only a tribe', but more a working group across the Cardano community .. the Cardano Scaling group!?

This report summarizes the work on Hydra since September 2023. It serves as preparation for the monthly review meeting (see slides and recording), where the team updates project stakeholders on recent developments to gather their feedback on proposed plans.

Roadmap​

This month, the team released version 0.13.0, and updated the project roadmap to include the next planned version 0.14.0:

The latest roadmap with features and ideas

Release 0.13.0​

• Implemented security fixes to Hydra on-chain scripts, all of which are only exploitable by other head participants (not any attacker)

• Added support for (externally) committing inline datums

• Improved stability by querying the cardano-node at the tip

• Improved state persistence

• See full release notes and a list of delivered features.

Notable updates​

• Moved the network resilience #188 issue out of version 0.13.0 to allow for a quick release of the security fixes. This feature has been split in two parts, both of which are completed and will soon be included in the 0.14.0 release.

• A new user submitted a feature idea about supporting extended keys #1141. Initially urgent, it's no longer a priority and requires further clarification.

• Although not part of Hydra version 0.14.0, the team completed the integration of Hydra support in Kupo #1078, which can already be used with Kupo version 2.7.0

• Implemented a new Hydra Poll DApp feature, which is an example to be demonstrated at the Cardano Summit 2023.

• Realized stateless chain observation #1096 as a dedicated feature, which will add a hydra-chain-observer tool and simplify upcoming feature implementation related to layer 1 transactions.

• Created a feature idea item to migrate all validators to Aiken #1092. This was moved to 1.0.0 though, as it would require a complete re-audit of the scripts and and there is no urgent requirement for performance improvements at this stage.

Hydra development​

Issues and pull requests closed since the last report

This month, the team worked on the following:

Network resilience to node crash​

Last month, the team worked on making the Hydra network resilient to transient network connectivity failure (see #188). However, this was just one aspect of the larger effort to improve the overall stability and reliability of the network. By completing #1079, the team also made the network resilient to arbitrary node failures, which is a much more likely event to happen should the network run for a significant amount of time.

Hydra nodes now persist the network messages they send to peers and together with the already present acknowledgment mechanism, this means a node recovering from a crash can resend messages which were not received by their peers.

Hydra Poll DApp​

As the Hydra team seeks to adhere more to the principles of dogfooding, we wanted to upgrade our basic Hydraw application to something more interactive. We, therefore, built a poll DApp with the goal of demonstrating how one can build an application living entirely in the frontend that can interact directly with a hydra-node, using its websockets-based API. This application was also showcased at the Cardano Summit 2023, where the Hydra team delivered a masterclass.

Upgrade to GHC 9.6 & Brick 1.10​

This month, we were able to update our compiler toolchain from GHC 9.2.8 to 9.6.3. This also required us to upgrade the hydra-tui from Brick 0.73 to 1.10 to stay compatible with the new version of the mtl.

The update to Brick necessitated a complete rewrite of the hydra-tui which greatly simplifies the application logic, using Brick's new elm architecture style combined with microlens-mtl.

Community update​

This month, the monthly review meeting was almost completely covered by demonstrations from the various teams working on and with Hydra. Don't miss out on the demonstrations of the various community contributions and make sure to also view the recording.

Hypix​

Hypix is a continuation of the 'Hydraw' demonstration the team used throughout development. The project is spear-headed by Trym Bruset and will integrate CIP-68 NFTs with instant transaction processing in a Hydra head to realize collaboratively created art pieces. The project is progressing great with an open beta available soon.

The Hypix user interface

Kupo x Hydra​

Indexers of chain data, like Kupo, are useful to enable lightweight decentralized applications like Hypix. As the application state of Hypix, for example, is not only on the Cardano layer 1, but also in the Hydra head, Kupo was enhanced to run 'in front of' a hydra-node and provide the same API for UTXOs indexed from a Hydra head. This was a joint effort between @KtorZ, @v0d1ch and @ch1bo and has been released in version 2.7 of Kupo.

Kupo indexing a Hydra head

Offline-mode Hydra node​

Ledger-only mode for Hydra was one of the Catalyst Fund10 projects that got funded and work is underway to make it an integral part of Hydra software components. A new ADR has been proposed which details the rationale behind this new feature and an overview of what it brings and what impacts it will have on the hydra-node.

Moreover, this implementation is underway and should be part of the next release. This will enable running a single hydra-node as a kind of 'ledger-as-a-service', priming it with some UTXOs and allowing clients to submit and validate transactions without any relationship to layer 1.

Voting on Hydra​

The Hydra-for-voting project has been started earlier this year as a proof-of-concept to explore how a Hydra Head could help improve the voting or rather tallying process for Catalyst and other large-scale on-chain decision-making process. It's been on pause for a few months due to other priorities but it's been revived recently and put to use to tally votes for the Cardano Summit's awards recipients.

This development effort led to a couple of new features and fixes:

• The ability to submit a transaction for layer 1 using 'standard' CBOR-hex format
• Upgrade to the checked-in protocol parameters used by demo network which were out of sync with the internal ledger version
• Discussion on the usage of extended payment keys in in relation to the Hydra Head protocol

Hydra Tallying UI

ZK Snarks on Hydra​

The team has initiated discussions with another funded Catalyst Fund10 project, exploring the implementation of a zero-knowledge proof system for Cardano, utilizing Hydra and ZK Snarks. At this early stage, there is minimal progress to display, but the team is eager to start the development of this promising tool.

Conclusion​

The monthly review meeting for October 2023 was held on 2023-10-30 via Google Meet with these slides and the recording.

As mentioned above, the meeting was about 45 minutes of demonstrations - which clearly shows that we are in the wake of Catalyst Fund10 where several project proposals involving Hydra and Mithril got funded and at the same time everyone was preparing for the Cardano Summit 2023.

With this many community contributions and projects moving forward, there was not much time to discuss the Hydra roadmap or go into details about Mithril project status. This is great!

Besides, the project roadmaps seem to be well aligned where our next features in-line are most anticipated (incremental de-/commits) and the adoption of Mithril is showing a healthy level with clear next steps and features ahead.

It is getting increasingly hard to keep track of all the new developments around Hydra. However, that is also often not needed and the Cardano builders just set off and build things using Hydra these days.

November will start with the Cardano Summit, likely yielding even more feedback and feature ideas from users. Following this, there will be several productive weeks focused on feature development before the usual Christmas and New Year breaks across the teams.

This report summarizes the work on Hydra since August 2023. It serves as preparation for the monthly review meeting (see slides and recording), where the team updates project stakeholders on recent developments to gather their feedback on proposed plans.

Roadmap​

This month, the roadmap has not changed much as the team mostly clarified upcoming features.

The latest roadmap with features and ideas

Notable updates​

• Added and completed support for inline datums in commit API #1043, which was reported by a user.

• Current work on network resilience #188 yielded two new / follow-up items #1079 and #1080. To minimize the 'scope creep' from this topic, the team has narrowed down the direct goal to improve the situation incrementally.

• When detailing the design of incremental de-/commits, the feature was also split in two. Each of the features would have a different impact on the user experience and relates to other ideas:

  • Incremental commit #199 is a bit more complicated, but paves the way for directly open heads and could make Always abortable head #699 redundant.

  • Incremental decommit #1057 is fairly straightforward and can be evolved into 'partial fanouts', which solves similar problems as the split-fanout #190 and only signing closable snapshots #370 would address. Also, the optimistic head closure #198 feature is closely related to this item.

• Added a new feature to improve user journey of running the hydra-node by removing the command line defaults #1064. This came out of prior idea discussion #454 which highlights the life cycle of ideas and features on the Hydra project.

Cardano Scaling workshop in Nantes, France​

From 2023-09-12 to 2023-09-14, most of the Scaling Tribe members, that is the two Hydra and Mithril teams, met for a three-day workshop in the beautiful French city of Nantes. This workshop was run following the Open Space technique which makes room for every participant to experiment and focus on what's of interest for them at the time while being engaging for the whole team. Here is a selection of topics covered that were most relevant to Hydra.

Aiken validator experiment​

As part of thw workshop, we explored how hard it would be to implement one of the Hydra validators in aiken. After roughly two hours of work we had a hydra-node running, tests passing and benchmarks of verifying one step in the Hydra protocol using aiken.

The hydra-node just uses the UPLC binary and its hash to construct transactions, we could rely on the same interface, with the difference that the script binary gets produced by aiken. Aiken does produce a socalled blueprint plutus.json containing the contract. We staged that file into git and embedded it into a Haskell interface to be able to reference the compiled script as SerialisedScript from hydra-node.

Next, we ported the simplest contract we have, namely the Commit validator, to Aiken. This was fairly straight-forward and we test were passing after some fiddling with internal representations of data strctures (plutus-tx is encoding triples as Constr data on-chain).

The results? A significant reduction in script size, cpu and memory usage on collect and abort transactions - or in different words, doubling the supported number of parties in a Head. 🎉

Shallow cardano-node experiment​

Running a hydra-node requires running a cardano-node, and running a cardano-node requires synchronisation with the chain and about 150 gigabytes of data to be downloaded and verified, which is a significant barrier to quickly spin-up hydra-node and Hydra Heads. Given a Hydra node does not need to look at the history of the chain, a Shallow cardano-node that can run without the full database of immutable files would be fine, assuming we can retrieve a recent state of the chain with the same security guarantees than a fully synchronised chain. This is exactly what Mithril can provide!

We ran a small experiment, trying to start a cardano-node manually removing all but the last files from the immutable DB and it worked as long as one does not query the node for a block too old in the past. This lead us to envision an improved user experience for installing and running a hydra-node, by packaging it with a cardano-node and a mithril-client to retrieve only the minimum needed state and be able to start a fully functioning hydra-node in minutes.

Other topics​

• We continued working on Kupo x Hydra integration, benefiting from the presence of Kupo's creator and former Hydra team member, Matthias Benkort
• We brainstormed possible synergies between the Hydra and Mithril projects, something that solidified into the desire to explore this concept of "Shallow node" and its integration within Hydra
• We discussed and clarified our new Security Vulnerability Disclosure Policy
• Discussing the idea of a joint documentation effort, we realised the existing documentation of both teams could be improved, for example by following a Standard documentation framework.

Hydra development​

Issues and pull requests closed since the last report

This month, the team worked on the following:

Network resilience to disconnects #188​

If a Hydra node crashes, it can recover its state from disk thanks to the event-sourced persistency mechanism in place. But if a peer inside the head sends messages while the node is down, these messages are lost for this node and this can well lead to the head being stuck.

For instance, in the following picture, we see that if Bob was down when Alice acknowledged snapshot 10, then it will wait forever for this acknowledgement and will refuse to move forward with the head, making it stuck.

Head stuck because of lost message

Hydra is designed to deal with non-cooperative peers in a very safe way: close the head! And this situation could be generalized as a non cooperative peer situation. So the safe move is to just close the head.

But practically speaking, closing the head every time a node restarts or a network connection is interrupted can lead to an operations nightmare. So the question we're exploring is, can we distinguish between non cooperative peer or just transient communication issue?

• ADR 27 gives details about our current strategy
• #1074 is a first implementation of this strategy
• #1080 should specify the resulting network protocol so that others could implement compatible hydra nodes

Incremental commits and decommits #199​

In the beginning of this month, some progress on incremental commits and decommits was made. An early experiment about the off-chain / node logic showed that we can re-use the /commit validator and motivated design work on the feature items.

As mentioned above, the protocol extension was split into two items Incremental commits #199 and [Incremental decommits

1057. Both items contain

high-level description, interaction outline and a first design of the on-chain validators which is currently getting discussed by the Hydra contributors and researchers.

Especially the incremental commit is not trivial as the interaction between layer 1 and layer 2 is quite intricate. Observing a deposit on the Cardano mainchain may not be seen by all parties at the same time and this should not lock up the layer 2 process of transactions, while premature usage of committed funds is to be avoided. On the other hand, each step in the protocol needs to be secure in presence of an adversary and unlocking of committed funds must always be possible.

Once this validator design is clear, we can get started implementing one or both of the features to enable more flexible usage of the Hydra head protocol.

Refactored chain state #1049​

Following up the implementation of ADR#24, this marks the first enhancement in a series of refinements, and our goal was optimization.

We eliminated the need for recursive types in persisted chain states, transitioning to a non-empty history model, preserving full backward compatibility with existing systems.

Importantly, throughout this refactor, we've kept the interface between the chain and logic layers entirely unchanged. This ensures a smooth transition and seamless integration with existing components.

One of the notable advantages of this refactor is a reduced persistence footprint. We now store individual points of the chain state, enhancing storage efficiency.

Community​

Open source contributions​

This month some open source contributions were received, and the team was able to contribute back to one project:

• Hydra: Lightning network-style payments use case write-up by @k-solutions
• Aiken: Started an emacs aiken-mode by @ch1bo (triggered by our experiments)
• Hydra: Small fix in the docs by @caike (from RareEvo workshop)

Current scaling use cases being tracked​

Hypix​

• An update of hydraw with improved usability to stimulate user engagement
• A showcase for Hydra, with an option to productize

Book.io​

• A Web3 marketplace for tokenized digital eBooks and Audiobooks
• Requires NFT minting and distribution at scale (for each copy/instance of a book)
• Opened a discussion to investigate possible approaches to scaling their operations on Cardano

Midnight​

• DUST airdrop and side-chain operation

SingularityNet​

• A Web3 marketplace for AI services offered and consumed through a pay-per-use API
• Discussion ongoing about porting an existing Ethereum solution to Cardano
• A chance to build on previous Hydra Pay efforts and upgrade the design to support unidirectional micropayments channels

SundaeLabs​

• Sundae proposed a ledger-only mode of Hydra, modelled on Gummiworm (the protocol used by SundaeSwap)
• https://cardano.ideascale.com/c/idea/102138

Emurgo​

• Initial discussions to evaluate building a payments channel network using Hydra
• Could leverage work done on Hydra Pay by Obsidian

Conclusion​

The monthly review meeting for September 2023 was held on 2023-09-20 via Google Meet with these slides and the recording.

This concludes yet another fully-packed month, this time not only focusing on Hydra but also encompassing Mithril and the broader goal of supporting Cardano's scaling effort to the best of our ability. Meeting and working together in person demonstrated once more, if this were needed, how much a small group of highly motivated people can achieve. It also fueled our desire and consolidated our vision of being a small group of engineers with a clear purpose, scratching our "own itch", contributing to the whatever projects and tools is needed and best for the job at hand, relentlessly prototyping and experimenting collectively to find the best solutions solving actual people's problems.

It is also worth mentioning that Catalyst's Fund 10 outcome shoudl result in more contributions to Hydra (and hopefully Mithril!) in the near term.

This report summarizes the work on Hydra since July 2023. It serves as preparation for the monthly review meeting (see slides and recording), where the team updates project stakeholders on recent developments to gather their feedback on proposed plans.

Roadmap​

This month, the team released version 0.12.0, and the project roadmap has been slightly updated to focus 0.13.0 on network resiliency and bump incremental commit and decommit in priority:

The latest roadmap with features and ideas

Release 0.12.0​

• Support cardano-node 8.1.2

  • Updated client and Plutus versions

• Layer 2 protocol changes

  • Authenticated messages
  • Removed redundancy

• Event-sourced persistence

• New API endpoints

• Removal of internal commit endpoint

• Improved off-chain transaction processing performance

• Security fixes

• See full release notes and a list of delivered features

Development​

Issues and pull requests closed since the last report

This month, the team worked on the following:

Update the tutorial and include Mithril #997​

To prepare for the Hydra master class at RareEvo, the team has taken the originally written tutorial for Hydra version 0.8.0 and updated it to work with the latest versions.

The challenge was to write content that is both easily comprehensible and functional across a wide range of user platforms, including operating system and processor architectures.

Besides writing the tutorial, it is essential to ensure that it is kept up-to-date (eg, using continuous integration).

Support cardano-node 8.1.2 #1007​

To be able to use the latest Mithril snapshots to bootstrap a cardano-node for the hydra-node, we needed to make some updates.

The hydra-node uses the cardano-api to establish a connection with the node using the node-to-client protocols. The format there has slightly changed (although versioned), necessitating an update to the version of cardano-api used within the hydra-node.

The way Haskell dependencies are managed required an adjustment of the versions for both cardano-ledger and plutus. These versions are used to construct our off-chain ledger and on-chain Hydra Head protocol scripts, respectively.

As a result, this process has proven to be more intricate than it might initially sound. However, it has ultimately resulted in enhancements to the efficiency of our on-chain scripts.

TODO: how much?

Event-sourced persistence #913​

We want the hydra-node to be efficient in processing events to yield high throughput on processing transactions off-chain.

Work done as part of #186 has demonstrated that the primary bottleneck to faster transaction processing inside the node was the state persistence logic, which simply overwrites the full state with whatever new state has been produced.

For that reason, we changed the persistent state to a sequence of events according to ADR24. Persistence is now done incrementally by saving only the StateChanged deltas.

As a consequence, the first spike confirmed the following performance improvements: master ~300ms → spike ~6ms.

TODO: what are the numbers on master before/after?

Finally, this also opens up interesting possibilities for state observation in clients.

New API endpoints​

This release also includes several additions to the Hydra API. We added the /cardano-transaction endpoint to submit a transaction to the layer 1 network. This feature improves developer experience as Hydra clients do not need direct chain access (eg, connect to cardano-node) to be able to submit transactions.

The other new /protocol-parameters endpoint serves the currently configured protocol parameters used in hydra-node. This provides more flexibility when creating transactions for the head on the client side and avoids configuration or hard-coded values.

On top of this, we also included the hydra-node version inside of the Greetings message. This is very useful for debugging purposes and detecting possible version mismatches.

Removal of 'internal commit' endpoint #1018​

In the previous release, we made an announcement regarding the deprecation of committing to the head through the websocket command. Subsequently, we have taken steps to eliminate this client command, which in turn resulted in the removal of the fuel markers that were previously used to mark the UTXO of the internal Hydra wallet meant for use in the Head.

This simplifies the setup needed to run the Head protocol and improves security since users can directly commit funds from their wallets without sending them to the Head operator beforehand.

Community​

Hydra master class at RareEvo​

We were happy to run a Hydra master class session at RareEvo 2023. The session attracted 30+ attendees for the introductory parts including a presentation on Hydra and Mithril. About 10-15 participants remained for the practical part and following discussion.

We also streamed the event live on Discord, but were not able to interact much there. The responses were positive though and we should be doing more things in the public on this #hydra-live channel.

Several participants managed to use Mithril and synchronize a pre-production cardano-node; two teams effectively initiated Hydra heads, conducted fund transactions within them, and subsequently closed them. The major challenges, as expected, concerned installation and networking. In the future, we intend to ensure the availability of prebuilt binaries catering to a wider range of platforms.

Catalyst Fund10​

The team screened all the proposals mentioning Hydra and Mithril. We submitted 11 community reviews and noticed, in particular, the following proposals:

• Sundae Labs Hydra Ledger-only Mode
• Sundae Labs Hydra Transaction Stream Plugin
• Hydra as a B2B layer for DeFi - a white paper and an MVP
• Decentralized Demeter.run - Federated Frontend Hosting - New revenue stream for SPOs
• Mithril - Open-source contributor

Conclusion​

The monthly review meeting for August 2023 was held on 2023-08-23 via Google Meet with these slides and the recording, 'broadcasting live from warm and sunny Colorado'!

It has been an interesting and unusual month. Some of the team had been in Longmont, CO to prepare for the RareEvo event and we used the chance to have the meeting in a hybrid setting with some IO stakeholders attending live on-site and about 20 community members online.

This time, the demo was about the updated tutorial and demonstrating the full setup of the cardano-node, opening a Hydra head on the pre-production network, and submitting transactions off-chain in 15 minutes!

The feedback we received included inquiries about the timing, method, and extent of the audit for the Head protocol. While we will have an internal audit, which is already helping us improve the protocol, there are no plans for a significant external audit with funding. We also had the chance to look into and learn about some Catalyst Fund10 proposals involving Hydra. Hopefully, some or all of them get funded and we are looking forward to testing the Hydrazoa concept, implementing the ledger-mode operation, enabling federated Heads, and achieving other objectives.

At the RareEvo event, we had the chance to meet and communicate with various people from the community. This ranges from known Hydra collaborators to tech-savvy stake pool operators, to representatives of successful applications running on Cardano for scaling purposes like book.io.

Also with a new full-time contributor on board, we are keen to add more functionality while the first applications prepare to utilize Hydra in production on mainnet.

This report summarizes the work on Hydra since June 2023. It serves as preparation for the monthly review meeting (see slides and recording), where the team updates project stakeholders on recent developments to gather their feedback on proposed plans.

Starmap update​

Every couple of months, the team reflects on the themes identified for 2023 (as presented in January), and creates a high level list of possible deliverables, eg using impact maps. This time, the team has come up with the following items:

• Sustainable open-source 💚

  • Hydra master-class workshop: A concrete session we aim to conduct in an upcoming event or online as a fallback. This will also be a motivation to update our existing tutorials and improve the first-start experience.

  • Engage and support developer community on Catalyst proposals: There are around 30 proposals mentioning Hydra in the current Catalyst funding round and we want to review, comment, try to align, and maybe even promote some proposals. The goal is to provide a platform for many new ideas while avoiding fragmentation and maximizing the reuse of work.

  • Further open up GitHub repositories: We want to provide a welcoming experience to non-IOG contributors and provide an easy-to-navigate group of repositories relating to Hydra.

• Mainnet mature application 💜

  • Make Hydra trivial to use: To enhance user and developer experience, we aim to enable a quick and frictionless setup of Hydra, possibly involving Mithril to get started. Our goal is to facilitate the opening of a distributed Hydra head with two nodes on a testnet within just a few minutes.

  • Build a basic head explorer: We have previously experimented with the explorer, and now we are looking to set it up for the long term. The goal is not only to assess Hydra adoption but also to provide a valuable tool for DApp developers. Additionally, we aim to create a visualization that end users can use to answer important questions, such as 'who is running this head?'

• Drive adoption 💙

  • Deliver the first version of HydraNow to mainnet: It's important that the very promising use case of payments using Hydra is made available to first users on mainnet. As a concrete example, this would be a mobile application 'HydraNow', which uses a hydra-pay service to enable peer-to-peer payment channels.

  • Enable uni-directional payment channels: This will expand the capabilities of any payment channel created using Hydra, allowing deposits and withdrawals while keeping the head open.

Work packages related to these items appear on the technical roadmap (especially in this view) and in our community activities.

Roadmap​

This month, the team released version 0.11.0, which shifts the way Hydra clients need to interact with the hydra-node when committing funds. The project roadmap was also updated to reflect the latest developments and ideas from the starmap:

The roadmap without idea items

Release 0.11.0​

• Completes the layer 2 ledger isomorphism with timed transaction support

• Adds commit from external wallets API, which allows to:

  • Commit multiple UTXO
  • Commit from scripts
  • Commit directly from any key

• Deprecated: Internal commits via the WebSocket API

• See full release notes and a list of delivered features

Notable updates​

• Two new items _elevated from user ideas and improved the API capabilities:

  • The ability to read protocol parameters via API #735
  • The ability tosubmit layer 1 transactions through hydra-node #966

• [Removing commit from internal wallet #954](Submit layer 1 transactions through hydra-node#954) is a consequence of deprecating this functionality in the previous 0.11.0 release. This will make the code base and documentation cleaner as there will be only one way to commit funds to a head.

• The first new item derived from the starmap is to Update tutorial and include Mithril #997. The tutorial is already a bit outdated, this will slim it down and incorporate Mithril to get the fastest (secure) bootstrap possible of a hydra-node.

• As a result of focusing on the very concrete use case of payments for mainnet deployment (and hence 1.0.0 release), multiple items need to (and can) get reprioritized to match the requirements of this use case.

The latest roadmap with features and ideas

Development​

Issues and pull requests closed since last report

This month, the team worked on the following:

Authenticate network messages #727​

The Hydra Head protocol paper states:

Each party then establishes pairwise authenticated channels to all other parties.

Hydra: Fast Isomorphic State Channels, p.14

Guaranteeing the authenticity and integrity of the messages received from one's peers is critical to the security of the Head protocol. Initially, we considered deferring this responsibility to the transport layer, for example, leaving it to Hydra node operators to use TLS or other encrypted channels to communicate with their peers. However, this introduced an additional and unnecessary operational burden on operators.

It appeared simple enough to reuse the existing Hydra keys (using Ed25519 curve), which need to be passed to the node, to:

• sign all messages sent to peers
• verify all messages received from peers

The net benefit of this feature is to increase the life expectancy of a head, preventing some forms of denial of service attacks or wrongly configured nodes that would lead to a stalled head.

ReqSn only sends transaction IDs #728​

This feature aligns the current implementation better with the original paper, as it requests snapshots using only the transaction identifiers, rather than the full transactions.

The motivation to implement this feature came from different directions:

• Improve the performance of the protocol by reducing bandwidth usage
• Make it possible to have 'atomic' transaction sets submitted (#900),
• Prevent 'front-running' from the current leader which, in the previous implementation, could inject arbitrary (but valid) transactions that peers had never seen before into a snapshot.

ReqSn's snapshots now contain only the list of transaction IDs and each hydra-node validates the snapshot by resolving those ideas against the set of transactions they have seen (through ReqTx). A consequence is that this introduces the possibility of space leak in the event a submitted valid transaction never gets included in a snapshot, but this should not be a problem in practice.

Note that the specification has been updated to reflect this change.

GitHub security advisories​

There was a bug in the multi-signature verification which could compromise off-chain security of a Hydra Head protocol. Although the fix was quite simple, we decided to use a new GitHub feature to create the first GitHub security advisory for Hydra.

The process creates a separate GitHub repository where potential security issues could be discussed privately (between the reporter and the maintainers) since potential attackers should not be aware of any issues until they are fixed. The team scored this issue using CVSS (The Common Vulnerability Scoring System), the score was moderate (5.5), and they didn't request a CVE (Common Vulnerabilities and Exposures) for this (yet).

The process overall works well, but Hydra CI is not checking the builds in these private, one-off forks so there is a risk of merging code that doesn't compile.

Moving to GHC 9.2.7 #841​

GHC 9.2 series have been around for more than a year and the whole Cardano ecosystem is gradually moving to use this new compiler version. Moreover, GHC 8.10.7 does not work reliably on aarch64 architectures, eg Mac M1/M2.

Moving to this new version entailed some significant changes in the code and also in the Plutus scripts:

Name	After (bytes)	Before (bytes)	Δsize
νInitial	4289	4621	- 7%
νCommit	2124	2422	- 12%
νHead	9185	8954	+ 3%
μHead	4149	4458	-7%

Community​

Hydra master-class​

As already mentioned above in our starmap update, we aim to have an educational session at the upcoming Rare Evo event:

Whether you are building decentralized applications which need to be fast and scale well, or you are creating infrastructure for other builders on Cardano, we will show you how to use Hydra to make your applications and platforms utilize fast off-chain transaction processing via the Hydra Head protocol.

In this session you will learn what Hydra is all about, how to build scalable applications in a live coding session, get to know the hydra-pay and hydra-auction projects, and be guided through exercises to get your hands on these APIs and early SDKs.

We are currently putting together a detailed outline and will send out communications on how to register for this exclusive session within the next 1-2 weeks. Stay tuned.

Conclusion​

The monthly review meeting for May was held on 2023-07-26 via Google Meet with these slides and here is the recording.

Although we had planned an awesome demo for this month's review meeting, showcasing a Hydra head hack without the authenticated messages feature, we encountered some live demo problems (lesson learned: always record a backup!). While we might present it next time, this situation emphasizes the need for continuous improvement in developer and user experiences.

We haven't received many questions on our journey to further mature the Hydra software stack to support various use cases building on it. However, the recently kicked-off Catalyst Fund10 sparked a nice discussion on how to handle all these great new ideas while being mindful not to fragment the scaling efforts too much.

Open protocols mean that new ideas can be incorporated not only into the various code bases, but also in the specification and standards which ensure interoperability. We discussed means to ensure adherence to standards in a light-weight agile way and the challenges of educating potential contributors in specification-driven development.

Next month we'll be at RareEvo, but might have the monthly just before in a hybrid session from Colorado. Until then, we'll make Hydra trivial to use and get started on our new objectives.

This report summarizes the work on Hydra since May 2023. It serves as preparation for the monthly review meeting (see slides and recording), where the team updates project stakeholders on recent developments to gather their feedback on proposed plans.

Roadmap​

While there was no release this month, the team implemented several notable features, which will be released soon as version 0.11.0:

The roadmap without idea items

Notable roadmap updates​

• Realized allowing commit txs with multiple UTxO #774 as a dedicated roadmap item separate from the related commit from external wallet #215 (which grew too big).

• Revisited off-chain performance by doing benchmarks #186. Identified a bottleneck, groomed and planned a follow-up performance improvement feature for Event sourced persistence" #913

• Plan to release 0.11.0 without Authenticate network messages #727 to deliver enabling featues earlier.

• API improvements and exploring batched transactions on L2 showed that ReqSN only sends transaction IDs #728 is in demand, front-loads further API changes and is estimated to be a low-hanging fruit.

• Reprioritized items in <= 1.0.0 column, to do items with on-chain protocol impact earlier and not much feedback has been received on snapshotting items. In fact, if Support larger # of UTXO via split-fanout #190 is done slightly differently, that should avoid some of the problems of impossible to finalize snapshots.

The latest roadmap with featuresand ideas

Development​

Issues and pull requests closed since last report

This month, the team worked on the following:

Commits with multiple UTXO #774​

One of the early adopting projects is exploring how to move scripts from layer 1 to layer 2. For that purpose, it was necessary to not only commit the actual script UTXO, but also a "regular" UTXO holding only ada to be used as collateral (the cardano-ledger always requires collateral although it would not be necessary on a layer 2).

To enable this, the specification and on-chain protocol needed updating. Before a protocol participant could commit zero or one UTXO, which changed now to a list of UTXO. As the specification is now part of the repository, it could be kept consistent within the same pull request.

Despite being a breaking change, leading to new Hydra script hashes to be published and used starting with version 0.11.0, this change was suprisingly easy to do and demonstrated the amenability of the Head protocol and the system architecture.

Commits from external wallet #215​

The team started to mark fuel some time ago as it was an easy workaround to distinguish UTXOs that can be committed into a head apart from regular outputs holding ada to pay for fees - the so-called fuel. However, this required users to "send funds" they want to commit first to the hydra-nodes internal wallet and involved additional steps in tagging such outputs with a special datum hash

To commit from external wallets, a new API endpoint was introduced for the purpose of drafting a commit transaction. The clients would request such draft transaction by sending a POST request to /commit and the hydra-node would respond with a transaction already authorized by the internal wallet. If the commit involved user funds (empty commits are still possible), then the client application would need to sign the transaction using the corresponding signing key. Also, submitting this transaction has shifted from hydra-node to the client.

This removes direct custody of hydra-node over user funds since clients can now use whatever key they own, not known to the hydra-node, to do a commit step and no single hydra-node has access to user funds used in the Head protocol.

Within this work package, marking fuel became deprecated and all UTXOs owned by the internal wallet are considered fuel. Fuel marking wil be completely removed in the future. Furthermore, a good old HTTP-based API is now used for the new query (so far it was only WebSocket-based), which prompts a potential shift to using OpenAPI as API specification since AsyncAPI does not describe synchronous requests well.

Benchmark performance of Hydra Head #186​

Low confirmation time is a key feature of Hydra Head protocol and is needed for some use cases. The existing end-to-end benchmarks have been revived and tailored for a baseline scenario, one with a single hydra-node and a single client. While this scenario is not representative of target deployments, it's interesting as it gives us an upper bound on the performance of a Hydra Head network.

The results have been somewhat disappointing, and further investigations lead to a few insights:

• There was a misconfiguration in the RTS of the hydra-node executable which was therefore not taking advantage of parallelism
• We identified the main bottleneck to be persistence of the node's state, which is currently done by overwriting a single file with the JSON content of the full state on state change which is pretty inefficient. As a consequence, the team will work on improving the persistence strategy as described in this ADR
• More benchmarks:
  • The aforementioned End-to-end benchmarks,
  • Ledger Micro-benchmarks as a comparison basis.

Operating hydra nodes​

As a "dogfooding" exercise, the Hydra team is operating a Hydra Head on mainnet on top of which is running our hydraw application. The Head protocol got stuck a couple of times and post-mortem investigations lead to a few improvements in how to operate a hydra-node and network:

• Better reporting on the version of executable hydra-node and hydra-tui which now report the Git SHA1 at which they were built
• Reduction in the volume of logs emitted by hydra-node by removing some chatty network-related logs and using ids to link Begin/End pairs
• Rework the log-filter tool to provide timings for various events and effects keyed by transactions

Community​

Hydra hackathon / workshop​

The Hydra team is considering holding a workshop at/around Rare Evo at the end of August and is in contact with the event teams at IOG and the Rare Evo organizers. The concrete format, scope and agenda is still a bit unclear as we are contemplating whether to do a workshop/tutorial style or rather a introduction + challenge setting event. If it does not work out for Rare Evo, we will find another event or do it ourselves.

If you are reading this and would be interested in joining such an event please drop us a line on discord or DM @ch1bo! Ideally along with some thoughts on preferred format or what you would be interested in.

Hydra for Auctions contributions and closing of project​

One of the Hydra lighthouse projects is slowly coming to an end. The collaboration project between IOG and MLabs on using Hydra for auctions is currently finalizing documentation and creating demonstration running the whole thing on a public testnet. Although the demo video was not available at the copy deadline of this report, watch this space for more about this next month.

The project yielded multiple github issues containing ideas and sparked great discussions on improving Hydra like reported last month. It is also the first project which demonstrates how to move smart contracts from the layer 1 (L1) to the layer 2 (L2)! Overall it is a great example of establishing crucial state on L1 and achieving scalability through Hydra as L2.

The code is fully open source and available on Github hydra-auction.

Conclusion​

The monthly review meeting for May was held on 2023-06-21 via Google Meet with these slides and here is the recording.

It was a fairly straight-forward month and consequently unexciting review meeting this month. Unfortunately, we could not release 0.11.0 yet, but we wanted to get the quite significant change of supporting commits from external wallets done first. This feature in particular was more involved than expected, but as the demonstration in the meeting showed, we are in the final stages of getting this over the line.

Besides some nice findings to potential performance improvements for layer 2 transaction processing, there was not much to announce this time around. Behind the scenes, however, there have been great progress on the Hydra for Payments project and the next updates are going to be more interesting again.

Despite holiday season approaching, we will march on, steadily adding features and enabling more and more use cases to scale through Hydra.

This report summarizes the work on Hydra since April 2023. It serves as preparation for the monthly review meeting (see slides and recording), where the team updates project stakeholders on recent developments to gather their feedback on proposed plans.

Roadmap​

This month the team released version 0.10.0 which includes many important features. The project roadmap was only slightly updated this month and already saw one more feature completed:

The roadmap without idea items

Release 0.10.0​

• First mainnet compatible release of hydra-node including technical changes and documentation

• Mainnet exposure currently limited to 100₳ per participant

• Added some missing features to the client API (short-term solutions)

• Some quality of life improvements in running the hydra-node

• Full release notes and a list of delivered features

Notable roadmap updates​

• Got new input and a concrete scenario for benchmarking (off-chain) performance of a Hydra head. As it is a very basic one, we moved the item from an idea to a feature and started work on it.

• Added a new feature item to add Hydra as a tool to developer platforms within our agenda of promoting Hydra as an open-source platform for scalability on Cardano.

• Already completed a first feature for 0.11.0 and considering whether to release it as early as possible with latest advances.

The latest roadmap with features and ideas

Development​

Issues and pull requests closed since last report

This month, the team worked on the following:

Timed transactions #196​

To realize the Hydra Head protocol, a hydra-node observes and follows the chain progress on the layer 1 (L1). Until now, this component was only observing protocol related transactions. Now, on every block roll forward, a Tick event carrying the notion of time from L1 is observed by the hydra-node and is used to validate transactions in a Hydra head. This means that clients can submit time bounded transactions to an open head and expect them to be validated using the same slot that would be used on L1. It is important to mention that time only advances on L1 when a block is produced.

Timed transactions in a Hydra Head state channel

This feature will make the Hydra L2 ledger now en-par with UTXO features available on the Cardano L1. A logical next step in this direction could be to make time in the state channel advance every configured slotLength (e.g. every second) using the system clock in between blocks as a form of dead reckoning.

Removing --ledger-genesis #863​

The hydra-node had two command line options to configure the ledger used for validating layer 2 (L2) transactions: --ledger-protocol-parameters and --ledger-genesis. The former option, which is self-explanatory, allows configuration of the Cardano protocol parameters to be used. Often, the same parameters as those on L1 are configured, or similar parameters with reduced fees, for example. On the other hand, the second option required the genesis-shelley.json file previously used to initialize the Shelley era by the cardano-node.

When we started using the current slot in the L2 ledger (see above), we realized that only the start time and slot length are effectively used from that configuration file. Moreover, it would be quite surprising if those were different and slots would be longer or shorter on L2 (unless explicitly configured). We opted to remove the option altogether and have the hydra-node fetch the genesis parameters from the Cardano network. This makes the system easier to configure and more isomorphic to L1.

Improving CI runtime​

The Hydra project embraces Test Driven Development and implements Continuous Integration (CI), checking all these tests pass and more.

The CI could sometimes take as long as an hour or more to run, which has a negative impact on the project's workflow. For instance, all the project's branches have to be fast-forward with master before being merged. So if all the tests on a branch are green and the pull request has been approved but is lagging a bit behind master, it has to be rebased, so the CI has to run again, incurring a one hour or so delay in this case before being able to merge it. The situation becomes worse when several pull requests are ready to be merged.

Analyzing the run from May, 10th we can see that:

• Building and testing takes 19 minutes for the longest
• Generating haddock documentation and running the benchmarks takes 28 minutes for the longest
• Documentation (which will need artifacts generated in previous steps) will take 14 minutes
• In total, this run took 1 hour and 16 minutes.

Let's focus on the build and the testing stage first. We're expecting the plutus-merkle-tree to run fast but it took 8 minutes and 52 seconds , 7m and 11 seconds being spent setting up the build environment. In other words, 81% of the build time is downloading binary dependencies from some nix cache.

Compiling the code requires cabal and other nix dependencies that will trigger massive downloads. Instead, let's, straight on, run the tests with nix and let nix decide what needs to be compiled. Sometimes, most of the code will not have changed and the test binary will already be available in some nix cache to be run without any compilation step.

This has been done in #867. Merging this PR, the build on master took 45 minutes and, specifically, the same plutus-merkle-tree job only took 1 minute and 44 seconds, only 20% of the time observed before.

Then, let's focus on documentation (14 minutes). Most of the time spent by this process is website optimization. It makes sense for the documentation that will actually be published on the website (master or release). But what about all the builds in branches for which the website will never be published? We need to keep this documentation step in branches because it gives us sanity checks, preventing broken links, for instance. But we can definitely remove the optimization step.

This has been done in #880 and it saves 10 minutes from this step.

Our first goal was to reduce CI execution time when pushing on branches and this has been improved. We're now having execution time significantly under 30 minutes where it used to be 45 minutes or even an hour.

We had some issues with compilation output, obfuscated by nix, which have been solved by #889.

Every morning, we rebuild our master branch and we can observe the whole execution time on the following graph (in seconds):

C.I. execution total execution time

Although we observed performance improvements on branches CI execution time, master execution time has not been reduced that much. We may save 10 minutes from the documentation step because optimizing the unstable documentation is maybe not worthwhile. But what's more problematic here is the process variability.

This variability can be explained by nix cache misses. That's something we need to investigate. It's hard to optimize a process with buffers, especially here where in case of a cache miss, recompilation has to happen and takes time. So every cache miss introduces several minutes of overhead which explains why we observe so much variation between two days with this master execution time.

Next steps:

• Inspect why we have an unexpected random cache miss
• Improve haddock generation time (15 minutes)
• Reduce bench time (we probably don't want to run the whole benchmark suite for every single commit, or a smaller one)
• Focus on the changed area (do not compile everything to generate the monthly report)

Community​

Hydrozoa #850​

This month, the project saw a new kind of contribution from the community. @GeorgeFlerovsky has written a research piece about an adaptation of the (Coordinated) Hydra Head protocol into a new flavor - Hydrozoa. The article is currently being examined and discussed on GitHub. Feedback so far has been positive. Of course, one does not simply change the Hydra Head protocol (pun intended), but the ideas contained could drive evolution and be followed up with concrete protocol extensions. Very much like the ideas presented in the original paper.

Hydra for Payments project update​

In this collaborative effort between IOG and ObsidianSystems, we are pushing the frontiers of using Hydra in payment use cases. It will lower the entry barrier for developers and ultimately users to leverage Hydra L2 for sending and receiving ada and Cardano native assets with very low fees and sub-second finality.

The project is in its second phase where the open-source hydra-pay library is being expanded with necessary features while the team pushes for building a dedicated mobile application.

With the recently released hydra-pay version 0.2.0, an end-to-end workflow of the Android App nicknamed "HydraNow" can be realized. This app will act as a layer 2 wallet quite like a Bitcoin Lightning Wallet and drives feature development in both hydra-pay and hydra in the background.

Two instances of HydraNow (in browser) connected via a hydra-pay channel

Spanish translation​

Last, but not least, we would like to thank @Agustinblockchain for their contribution of a Spanish translation #866 of the hydra.family website 🇪🇸 🎉

Conclusion​

The monthly review meeting for May was held on 2023-05-24 via Google Meet with these slides and here is the recording.

The monthly review meeting was lively and we had stakeholders from IOG as well as interested community members join. They were well served by two very nice demos shown by ObsidianSystems and IOG teams.

After having dealt with some issues last month, we are happy to have the first mainnet-capable version released and have already another feature completed this month.

It is great to see that more and more contributors from the community are stepping up to become advocates of the technology. A community is forming and we realize that the project could need some help from someone dedicated to keep track of all the great things happening. What a great problem to have!

This report summarizes the work on Hydra since March 2023. It serves as preparation for the monthly review meeting (see slides and the recording), where the team updates major project stakeholders on recent developments to gather their feedback on proposed plans.

Roadmap​

The project roadmap was only slightly updated this month:

The latest roadmap with features and ideas

Notable roadmap updates​

• There are still many 💭 idea items on the roadmap, however, not on the current and next planned release columns. The process involves clarifying and refining each idea item into a 💬 feature before starting work on it. This includes giving equal consideration to new user 💭 ideas and requests.

• Temporarily postponed the Hydra heads explorer #696 item until there is more clarity. While there aren't any major blockers,the explorer just requires further detailing, especially since we're currently in talks with existing Cardano explorer platforms.

• Converted the aggregated multi-signature #193 from the concrete roadmap into an idea discussion #787. This is based on feedback from the community, and we believe that discussing it there would be more beneficial.

• The main focus for the 0.10.0 release is implementing mainnet compatibility. This is mostly done and only requires some documentation and disclaimer updates. We're about to release 0.10.0, which will be the first mainnet-compatible version.

• Meanwhile, work on the configurable API #380 was completed, which gave rise to new ideas and follow-up feature requests from users. One of them (Add HeadState/Snapshot to Greetings #823) was fairly straightforward and necessary to deliver a consistent, usable increment on the API with the upcoming release.

• Prioritized Support timed transactions #196 higher as yet another use case would benefit from this.

The roadmap without idea items

Development​

Issues and pull requests closed since last report

This month, the team worked on the following:

Configurable API​

The API evolved a bit, driven by the issues our users reported #823 #813 #800 #789. Related changes were added to the API server so now our clients can:

• Control the historical messages output. History messages can be displayed upon re/connection or not depending on client needs.
• Snapshot UTXOs can optionally be disabled.
• Transactions can be displayed in CBOR or JSON format.

Our clients can also have a nice insight into the current Hydra node state and head UTXOs that are now displayed as part of a Greetings message.

Next steps on the API level are to further fulfill user needs by grooming and implementing needed changes related to filtering, pagination etc.

Versioned docs and specification​

Over the last couple of months the Hydra specification became an important artifact to use in discussion, review and potential audit of the Hydra Head protocol implementation. The document was now moved from overleaf into the Hydra repository, where it is properly versioned and built on each CI run. Changes can be proposed using our regular pull request worfklow and the final PDF is built and published to the website automatically.

Note that the above link points to the new /unstable version of the documentation, which holds the bleeding edge user manual, specification and API reference (which got a new sidebar) built directly from master. The normal, non-unstable version of the website is always referring to the last released version.

Specification on the Hydra website

Fixed scripts, Plutonomy and custom script contexts​

As we made the specification use a more direct way to represent transactions (instead of the constraint emitting machine formalism), we realized that our scripts are not correctly ensuring script continuity. We identified these 'gaps' as red sections (see above) in the specification and worked on fixing them.

While the actual fix #777 was fairly straightforward and could easily be covered by our mutation-based contract tests, the script size increased and we could not publish all three Hydra scripts in a single publish transaction (which allows for a single --hydra-scripts-tx-id parameter on the hydra-node).

To mitigate this, we looked into the UPLC optimizer Plutonomy. Applying it was fairly simple, our tests did also pass, script sizes and costs also became lower. But, script size does not matter so much as we are using reference scripts and using a (not really maintained?) optimizer which introduces yet another question mark after compilation from plutus-tx to uplc was not our cup of tea right now at least (and we might pull this out of the drawer later).

There is an alternative: decoding ScriptContext involves quite some code, but we don't need everything in all validators. So we introduced a customscript context that only decodes the fields we need.

scripts	@0.9.0	fixes	fixes + plutonomy	fixes + custom ScriptContext
νInitial	4621	4727	3672	4300
νCommit	2422	2513	1816	2068
νHead	8954	9492	7579	9456 (no custom SC)
μHead	4458	4537	3468	4104

As part of this process, we also updated dependencies #826 to the latest cardano-node master. Although it didn't have an impact on script sizes, it's a crucial step towards preparing for upcoming hard-forks.

Rollback bug hunt​

Opening our first head on mainnet failed. We didn't lose any funds, except for some fees, but the head just did not open. Exploring the logs, we figured out that a rollback happened while opening the head and there was a bug.

This is our test pyramid. It already contained some tests about the rollback but we decided to enrich our model-based tests to simulate rollbacks (before that, it used to run on a perfect blockchain). You can find more about our model-based test strategy in Model-Based Testing with QuickCheck.

Testing pyramide

The new property headOpensIfAllPartiesCommit helped prove the issue. At the end of the day, the problem came from a concurrency issue introduced while implementing ADR 18.

In the figure below, the DirectChain processes a new block, updating the chainState stored inside the headState. This also leads to an event being published to some event queue. Later, the HeadLogic (called Node in the picture) will process this event, updating the headState.

At the end of the process, we can see that the new headState points to a previousRecoverableState which contains the same chainState, chainState 1 instead of chainState 0. If a rollback then happens, the headState will be reverted to this previousRecoverableState and the fact that it contains chainState 1 instead of chainState 0 makes some on-chain observations impossible.

Race condition

This explains the issue we had when opening our head:

1. a commit A is observed on-chain;
2. a rollback happens so that the headState forgets about this commit but not the chainState (remember, it's the wrong chainState);
3. the commit is observed again on the chain but ignored by the chainState (because it has already seen it, it just ignores it);
4. the headState will never hear about this commit again and so will never open the head, waiting forever for the missing commit.

We decided to implement the following solution:

• A local chain state is re-introduced in the chain component, not shared with the head logic.
• A copy of the chain state is kept in the head state to keep the benefits of ADR 18 regarding persistence.
• The rollback event is removed from the API until #185.

Possible solution

Rollback management is quite a tricky business. It might be the case that we've tried to be a bit too smart. So we're doing a rollback in our way of handling rollbacks until we focus on this topic again when dealing with this roadmap item: Handle rollbacks II.

Community​

Hydra for Voting​

The project is advancing and a basic vote tallying scenario in the Catalyst use case was demonstrated in the review meeting. The project is driving the API discussions as it is not using any Haskell tooling, but an application in Java with Aiken as the validator scripting language. Besides the catalyst use case, other scenarios like the ballot voting for the summit are also explored now.

Hydra for Auctions​

A new demo was recorded in the wake of an upcoming Twitter space discussing auctions and NFT marketplaces with the community. The feature set currently includes starting the auction on L1, bidding on L1 or (and this is the novel thing!) transferring the auction from L1 to L2, such that it can be bid on L2.

Kupo x Hydra​

In a good old pairing session between IOG and CF engineers, the integration of Kupo with Hydra was explored. This seems to be promising and work started here kupo#117. This will make it possible to run kupo natively connected to a hydra-node, very much it would run with cardano-node or ogmios. Kupo is a lightweight indexer of chain data like unspent transaction outputs and allows its clients to query information on-demand. 🐹

CBIA meetings​

Hydra Team is now a regular participant to Cardano Blockchain Infrastructure Alliance meetings which happen more or less on a monthly basis. Hydra was featured during the meetup that happened while the team was on a workshop in Feldkirch and through this participation we hope to contribute to growth of the Cardano eco-system and position Hydra as a key infrastructure for builders.

Twitter space on scaling Cardano​

This month we took part in a Twitter space about scaling Cardano and how Hydra can contribute to this. Thanks for conducting this @thepizzaknight_ 🙏

Conclusion​

The monthly review meeting for April was held on 2023-04-26 via Google Meet with these slides and here is the recording.

Although it has been a busy month we could not cut a release, unfortunately.

We've experienced several setbacks due to the commits vs. rollbacks bug and oversized script sizes, which have slowed down our progress. Additionally, the back and forth on the API, which at times required creative and unconventional solutions, has also been time-intensive. However, we view this feedback as incredibly valuable, as it will ultimately make hydra-node more user-friendly and capable, albeit through a step-by-step process.

Associated projects in the greater Hydra community are moving ahead nicely due to the collaborative approach and tight loops of interaction between the individual teams.

All things considered, the project can be considered on track. We are very close to cutting our first mainnet-compatible release and the rising number of user requests and interested developers are good indicators that Hydra adoption is increasing.

This report summarizes the work on Hydra since February 2023. It serves as preparation for the monthly review meeting (slides/recording) , where the team updates major project stakeholders on recent developments to gather their feedback on proposed plans.

Roadmap​

The project saw one release this month and several items on our roadmap were updated.

Released version 0.9.0​

• This release brought in on-chain and off-chain changes that are now also fully reflected in our specification.

• We managed to decrease the costs of our plutus scripts by using a new error code framework and made also the head script a reference script.

• The mutation test suite is improved. Now, if tests fail they show the correct reason.

• The contestation deadline is now pushed out by each contesting party, so the contestation period parameter can be chosen irrespective of number of parties.

• Added a tutorial contributed by @perturbing.

• Full release notes and list of delivered features

The latest roadmap, with many items marked as idea.

Notable roadmap updates​

• Now that 0.9.0 is released, the focus is on mainnet compatiblity. The next planned version 0.10.0 will be the first hydra-node to be mainnet compatible.

• Also prioritized API configurability #380 higher and planned it into 0.10.0 as more users were requesting this.

• We aim for several more 0.x.0 versions to incorporate user-requested features before reaching a 1.0.0, which will be a fully maintained release where features can be considered stable and won't be removed without appropriate deprecation cycles.

• Marked multiple features as 💭 idea, meaning they are up for discussion. Many have been on the roadmap for a long time without user demand or input, while other ideas are incubating as github idea discussion. We are planning to convert these idea issues to GitHub discussions as these provide better dicussion tools and allow to gauge interest in topics. In turn, by cleaning up we provide more room for the most popular ideas to be planned onto the roadmap in a more timely manner.

• In short, if you want to see some feature added, show your support on the corresponding idea discussion.

• The 💬 feature items remaining are the currently identifed "must-haves" or too vague to remove 🔴 red items.

The roadmap without idea items.

Development​

Issues and pull requests closed since last report

This month, the team worked on the following:

• Making Hydra mainnet compatible. Besides making it technically possible to run on mainnet, this is about safeguarding our users and preventing them from shooting themselves in the foot with a mainnet gun. That is why we limited the amount of ada you can commit to a head on mainnet. Our smoke tests should be running on mainnet also so we made sure to return the leftover funds back to our faucet. There was also work on our CI that enables running the tests on mainnet using a dedicated github runner.

• Improving the Hydra UX. We noticed a possible pitfall when restarting the hydra-node using different parameters than in the persisted state. Now, the node would prevent this kind of misconfiguration.

• Optimize the on-chain scripts. Reduced the cost of opening/closing a Head by reducing size of scripts via error codes #748 and also having the head script as a reference only #701).

  The mutation test also got improved by making all cases expect the corresponding error codes #705 and new golden test suite ensures the script hashes don't change accidentally.

  Furthermore, we addressed a problem discovered in our property tests #724 by preventing committing outputs with reference scripts to a Head #766. This is still a documented known issue, but a workaround is available.

Community​

• Team workshop in Austria. Meetup of core contributors in Feldkirch, Austria for retrospective, grooming and discussions on future plans for Hydra. Part of the agenda was also a presentation of Hydra to members of the Cardano Foundation and a meeting with builders from the Cardano community.

  The highlight, though was the common effort of going the last mile to open a head on mainnet for the monthly review meeting! 🎉

  JUST IN: We have a Hydra Head live on the #Cardano Mainnet 🚀 pic.twitter.com/6kDKq7T7no

  — Emmanuel 𓂀 🍕 𓅓🇬🇭🦄🟣⚡️ (@thepizzaknight_) March 22, 2023

  We demonstrated our good old hydraw application on that Hydra head and we saw thousands of pixels painted by hundreds of users (no detailed metrics).

• Next step in Hydra for Payments. Also announced in the monthly meeting was the next phase on the Hydra for Payments project. The scope of this will be to extend hydra-pay and build a mobile payment channels app that makes direct use of it - working title: HydraNow.

Conclusion​

The monthly review meeting for March was conducted on 2023-03-27 via Google Meet - slides/recording.

This month was very important for the project and culminated in the first demonstration of a Hydra Head on mainnet! The demo was well received and we received positive feedback in the meeting, on following twitter announcements and on the published recording on youtube. Inspired by this, we saw multiple people offer to help and collaborate on communicating, educating and ultimately spreading the love. The invitation via the new discord category worked well - we had about 40 community members in the call - and we will continue with this workflow. Make sure to follow the Hydra #announcements on the IOG Technical discord server.

While being mainnet compatible is a major milestone for the project, there are still many known issues, shortcomings and requested features. The roadmap changes this month should make it clear that we are serious about the latter - Hydra will only reach 1.0.0 if it is used by an application on mainnet. Hence, we will focus on adding features required for payments, voting, auctions, ... and eventually, your use case.

Hydra Head is ready to be used on mainnet. Are you ready to use it?

This report summarizes the work on Hydra since January 2023. It serves as preparation for the monthly review meeting, where the teams update major project stakeholders on recent developments to gather their feedback on proposed plans.

Roadmap​

Looking at the roadmap, there are a few things that changed this month:

• In general, there will be several items in flux as we align the technical roadmap to support our high-level objectives for the year.

  • New items for example: create a use case-centric roadmap #710 and open specification with change process #693.

• Also note that we have not had a release in a while and we want to have smaller faster releasable items again.

• Aligning implementation with HeadV1 specification #452 was more work than anticipated. There was quite some back and forth with the writing of the specification and discussions with researchers leading to scope creep.
  Most importantly, we made on-chain behavior consistent with specification.

  • Follow-ups for some off-chain protocol logic:
    • Authenticate network messages #727
    • ReqSn only sends transaction ids #728

• Version 0.9.0 will be the first release to be developed against a full technical specification. We expect a release in early March.

• Version 0.10.0 will be the first mainnet compatible release. For this, we created and groomed the main feature item Mainnet compatibility #713:

  • Not only making it technically compatible but also adding guard rails to limit risk (“not shoot yourself in the foot .. with a real gun now”) and documenting known issues.

• For a fully maintained version 1.0.0 we still need to address some things, but we aim for an increased rate of planned releases in the next months as we march toward it.

The latest roadmap with multiple new items and starting to reflect Hydra objectives of 2023.

Development​

Issues and pull requests closed since last report

This month, the team worked on the following:

• Aligning specification and implementation. This major feature item was the topic we worked on over the last couple of months. This time we could finish the major, must-have parts and can continue now on other features. Notable things from this month were:

  Pushing out contestation deadlines: The original paper defined that all participants need to be contesting within the contestation period (after close, before the deadline). In discussions with researchers last year, we identified better semantics: the contestation deadline gets "pushed out" on each contest. This has the benefit that the period needs to only accommodate the minimum time required to get a single contest transaction into the blockchain. Before it was depending on the size of the Hydra head - each participant can contest once.

  Align off-chain protocol logic: While most concerns addressed were about the on-chain scripts of the Hydra protocol, this is about the layer two protocol logic. That is, which messages are exchanged between Hydra nodes and how a node reacts to them. Here we realized that the implementation was slightly different from the formalism used in the paper and specification. This makes it non-obvious whether the behavior is exactly as specified or whether the specification can even be implemented into a working system.

  To clarify this, we put in some effort to align the implementation with the specification and the other way around. This allowed us to reason through our behavior tests whether the protocol as specified is sound as well. Below you see an excerpt of the Haskell code against the LaTeX pseudocode formalism:

  

• Profiling Hydra transactions. We will want to maximize the capability of the Hydra Head protocol. This includes writing the on-chain scripts efficiently. Besides having transaction cost benchmarks, we worked on a profiling workflow for Hydra protocol transactions and published a tutorial on the Hydra website. This allows any contributor to Hydra to profile the scripts used in these transactions and identify code that uses a lot of the memory or CPU budget in flame graphs like this:

  [Interactive]Head not closable/limiting script size discussion + acceptance property

Community​

• Audit guidelines. Hydra is an open-source project and, as such, should be freely used, reviewed and audited. The goal of audits and peer reviews is to increase confidence in the security of a system. However, conducting an audit is also very expensive and not in the scope of our current budget. All the while, the worth of an audit is not the same to all users, some will not even trust an audit they have not paid for themselves, while others do not care at all and would blindly trust the word of the maintainers.

  While the Hydra project continues to seek for funding to eventually pay for an audit of the protocol specification and the hydra-node implementation, it is a prerequisite that the necessary artifacts for such an audit exist. As we have been creating them in recent months, we also worked on guidelines for auditors, now published on our website. These guidelines should serve as a starting point and scope description of what would be most crucial to audit.

• New tutorial. This month we received a great contribution from @perturbing, who created an end-to-end tutorial on using a hydra-node and opening a Hydra head on the testnet. We gladly adopted this tutorial and aim to keep it up to date with upcoming versions of the hydra-node. It will provide a great starting point for new users and makes the project more approachable. Thanks! 💚

• Hydra for Payments demo. As we are about to kick off the next phase of the payments use case with Obsidian Systems, they demonstrated the use of hydra-pay in a new backoffice-like view, listing multiple payment channels and providing a great teaser of using payment channels for pay-per-use APIs or cashback use cases (micro-transactions).

  

• Hydra for Voting. The project with Catalyst and Cardano Foundation on voting with Hydra (or rather tallying with Hydra) showed some progress. They have contributed Merkle Trees to Aiken, which will be useful to prove the correct tallying of votes in the L2. They started work on a hydra-java-client and provided valuable feedback on the hydra-node API over the last month.

• Hydra on Demeter. The folks from @txpipe have been working on integrating Hydra into their platform-as-a-service offering, Demeter. This is a great way to get started with running a hydra-node as a service and we plan on integrating one instance into our long-running hydraw demo. It's still not fully released, but the teams are working together to make this integration great. Thanks for providing this service! 💚

• New discord category. The Hydra project will be using Discord a bit more in the coming weeks and months. For that, a new category on the IOG Technical Discord server has been created. Look out for news and, most importantly, invites for the next month's review meeting in the new Hydra #announcements channel!

  

Conclusion​

The monthly review meeting for February was conducted on 2023-02-27 via Google Meet - slides / recording.

This month's installment featured two demos, one about the Hydra Payment channels, and another about the progress on the Hydra for Auctions project. It was great to showcase not only core development progress on the specification and protocol implementation in the hydra-node, but also to see contributions from the Community on tutorials and tooling (Demeter integration).

An exciting, but also challenging next step will be to aim for mainnet compatibility and at the same time deliver quick releases to enable these lighthouse projects and the increasing number of use cases explored by the builders of Cardano.