FYI: apparently the #bitcoin conversion fee of @xapo is about 1.5% whereas @ecoinio is about 3%. Unfortunately those fees aren't disclosed.

FYI: apparently the #bitcoin conversion fee of @xapo is about 1.5% whereas @ecoinio is about 3%. Unfortunately those fees aren't disclosed.

I take it you do not speak Tagalog. If you did you would see that a very large majority of the discussion is on alts, pozis, sig campaigns with very little discussion on Bitcoin itself.
~BCX~

https://github.com/mailchuck/PyBitmessage/releases/tag/v0.4.5

I created a new fork of PyBitmessage, and this is the first release. There were two reasons behind it.

The first one is that the user interface of PyBitmessage is bare bones and often annoying. So I integrated the changes felidosz made, and been waiting for half a year to be merged: https://github.com/Bitmessage/PyBitmessage/pull/791

Felidosz' updates were great, but not complete, there were things missing/buggy, and some annoyances were still present. So I fixed some stuff, added other stuff and improved even more.

I also tested how a normal user (my wife) is able to use the official PyBitmessage for an email gateway. She failed to even register, even though I was sitting beside her and I asked her to read the instructions on mailchuck website. She couldn't understand it. So I also spent some time on having the UI react better to an email gateway, which is the second reason for this version. Now you can register/unregister through a context menu (right clicking on an account), and the user interface adapts automatically when using the gateway, so it looks like a normal email client.

I'm still working on it, but I think it's at a stage when others can look at it and try it out. The data and config are compatible with PyBitmessage, you can switch back and forth between the two and it will continue working. There are no changes to the sqlite database, and the keys.dat config file has some new options which the regular PyBitmessage ignores. All the code is on github, I built a windows binary, and the code has been tested on linux and windows (I have no Mac).

There is no need to use PoW to secure a block chain, let alone very strong PoW.
That design decision comes from a fundamental (but rarely discussed) design requirement Satoshi had, which is that all entities in Bitcoin must be able to join and leave at will, unannounced and anonymously.
If you want a purely decentralised system, this is a "hard" requirement, it's non-negotiable. Bitcoin simply wouldn't work without it.
If you are willing to relax the decentralisation requirements a bit so that there's some kind of explicit join/leave protocol and participants have some verified identity, or you can rely on non-standard hardware, then you don't need PoW anymore. You can just use chains of digital signatures, for example. Or you could use PoW as a means to do a randomised leader election like Bitcoin does, but where each miner just uses a single CPU because the amount of hashing doesn't matter (fraud can be punished via the legal system instead).