Please login or register.

Monday Monero Missives #16 - October 27th, 2014

Monero Monday Missives

October 27th, 2014

Hello, and welcome to our sixteenth Monero Monday Missive!

Major Updates

  1. We have made major strides in the initial database implementation (you'll recall from our last Missive that our first implementation will use LMDB), and it is very nearly ready for broader testing. Specifically: the new blockchain is working for most things, but there are bugs with certain aspects of block verification that need to be fixed before it can be more widely tested. If you are particularly intrepid you can already grab it here: https://github.com/tewinget/bitmonero/tree/blockchain and compile it, and thus assist in identifying areas where it breaks down, although such reports are probably best submitted as github issues to tewinget's repository to reduce duplication. Once these and any other major issues have been weeded out the next steps would involve a bit of refactoring, fix cross-platform nigglies, and open it up for general testing.

  2. The testing of per-kb fees on testnet, too, has gone exceedingly well. We will be adding the functionality to simplewallet (previously it required manual creation) and hope to deploy that for general testing within the next week.

  3. Kitware staff, Ben Boeckel in particular, have spent a lot of time completely reworking our CMake build system and bringing it up to best practices. The fruits of those efforts can be seen on the Pull Request currently undergoing testing: https://github.com/monero-project/bitmonero/pull/180 (feel free to checkout this PR if you'd like to test). Now that the build system is starting to come together in its final form, we are hoping to use it to tag and release 0.8.8.5 during the course of next week.

  4. In order to more efficiently deal with changes in the on-disk wallet format we are moving away from the old serialised+encrypted .keys format, and have a new format which is effectively encrypted JSON. This change allows us to note the wordlist language in the wallet format (so that the "seed" command can reflect that choice) and allows for cross-platform compatibility of the .keys file, which we are sure is excellent news for anyone that moves wallets between operating systems and architectures. You can test this in PR 179.

  5. There have been a constant string of improvements and changes to the forum software to make it more usable and useful. In particular, new comments in a thread are highlighted within that thread. Additionally, unread threads (or threads with new unread comments) are highlighted by having a green dot next to them. Both of these apply to logged in users only. If you haven't visited the forum, you are encouraged to do so: https://forum.monero.cc

Dev Diary

Core: LMDB implementation is rough but nearly working (details above). Worth testing cross-platform, least of all from a build perspective.

Core: since we have already had to perform the rather annoyingly complex task of offloading MoneroPulse checkpoint checks to a separate thread (so as not to tie anything up during checks) we have begun extending this to other parts of the core that could potentially be or currently are pain points. This does not include the flat-file blockchain saving, as that is going to be deprecated with the move to LMDB, so pools will just need to hang on and deal with that nuisance for a little bit longer.

Build: CMake is looking a lot cleaner and easier to grok. It also fixes cross-compile (see: http://www.cmake.org/Wiki/CMake_Cross_Compiling) which means that binaries for all our major supported platforms can be built on a single system.

Account: multilang wordlists are now inherent to the wallet/account, so that RPC and CLI calls that retrieve the mnemonic do so in the correct format. This has, in turn, necessitated moving away from the horrible serialised data format for account data. Since epee's JSON library is beyond redemption, we have opted to use RapidJSON instead (which is headers-only and thus straight in the source tree).

Until next week!

PS. this Missive has the very great honour of being the first one to be finalised and very nearly posted from the air;)

Replies: 3
davidlatapie edited 8 years ago Replies: 1 | Weight: -852 | Link [ - ]

Am I the only one to consider this very newsletter will show some positive signs to concerned holders that things are actually going well?

Reply to: davidlatapie
just-mike edited 8 years ago Replies: 1 | Weight: -851 | Link [ - ]

Yes. This M3 sounds like most of the other ones to me: "we are working but can't prove it to you". Since I am not a coder, looking at pull-requests does not help me.

I realize the changes are not easy. Is there any way to make improvements more visible? Can I go on the testnet and see the per-kb fees in action? How do I even get on the testnet?

Reply to: just-mike davidlatapie
fluffypony edited 8 years ago Weight: -849 | Link [ - ]

Sure - you can checkout this branch: https://github.com/tewinget/bitmonero/tree/per_kb_fees and run your own mining daemon on testnet, then create a manual transaction using simplewallet in RPC API mode to test (non per-kb daemons won't mine it anyway as it's lower than the fee they expect).

Getting on testnet is easy - just run anything with a --testnet flag and you're sorted:)