Development Update :safe: 13th April 2015

Thanks to @justine_mclevy, here’s the transcript of this week’s catchup meeting.

The bulk of the effort has again gone into preparing for Testnet 3 in Rust.

Several devs were on leave last week, but those who remained have again pressed on, putting in long hours and achieving several sub-targets. The sprint (which is now finished) whilst not actually achieving its ultimate goal came pretty close.

The primary aims for the forthcoming week are to put together detailed plans for the next sprint as well as complete the missed targets from the previous sprint.


Last time i read about this topic it was said that testnet 3 would launch in c++ and rust might be something that comes after that. Now it seems like it is going to be rust from the beginning?

To me this seems like the third massive change in direction during the development to reach testnet 3.

There was a fairly large discussion about this. My understanding of it was that the rust port was happening at lightning speed, so they decided to finish the port. From Testnet 1 to now, they were able to bring the lines of code from ~550k to about 8k largely in part to porting to rust (and refactoring of routing) . Statistically as far as errors per lines of code, the opportunity to make a change like this, as well as simplify the entry barrier for new devs, I think it was an opportunity they couldn’t pass up.

Once the network is launched, it’s up. It’s autonomous. It pretty much has to be right from the start, only small changes after.


But testnet3 is not this start you are talking about.

IMO it would have been beneficial to have testnet3 using C++, testnet4 using rust port, and then launch beta with the one proving to be better.

My issue with the situation is that reading the dev updates we see that the people that were hired as c++ programmers apparently have no clue about rust and have to use their time to learn programming in it. This is not exactly utilizing the resources to the max.

I personally applaud every company that is valuing its employees by giving them training and preparing them for future technologies. Additionally they’re thinking long term here. What is the benefit of bringing out Testnet3 in C++ if a later lunch in Rust means a more stable release and an easier entry for new developers? That we as users will get to try testnet3 4 weeks earlier? I don’t think that should be a priority.


Adapt or die… I have full faith in Mr. Irvine and his team. Technology changes so quickly that it was a smart move to add RUST. As a rather large MAID holder, I do like to see that Mr. Irvine is preparing MAID as an actual software deployment, and not using the road-map just to make day traders happy; temporarily. Use this time to buy more MAID, that’s what I’m doing. Also, with MAID so close to launch, I think total transparency is NOT a good thing. This is still software, and one must be somewhat secretive when so close to launch. So I applaud (alongside hillbicks) the quiet, even though Mr. Irvine has made (MAID) himself so available on this forum to answer almost all the questions.


@reivanen I disagree with you and wonder what you base that opinion on?

Do you have any knowledge or experience in software development, design, testing, maintenance? Project and team management? Do you know C++? Have you investigated Rust?

IMO David has taken a tough, but looking to be very much the best, decision. I would not have had the balls to do what he’s done just lately. I wouldn’t even have been looking at Rust at this stage. But this is why he’s quite likely about lead a team into creating the most significant software innovation since the www and I’m here watching from my boat.

Questioning decisions and seeking explanation is fine, but opinion needs to be based on something other than emotional responses before airing it. Well that’s my opinion of opinion :smile:

So I’m curious to know and will happily debate with you if you’ve based that on something we can talk about. ?


There are quite a few reasons, and porting to a programming language that is still in beta is one of them, and quite risky. Also now there won’t be a c++ implementation to benchmark against.

Notice that i am arguing what David himself said only some weeks ago. The safe path of finishing c++ implementation and THEN go for rust port, but apparently he has been so convinced by rust that he is willing to take the risk and go all in.

I have no doubt that rust is a great tool IF things work out. As you yourself said, what David did takes balls.

I still have no idea what if anything your opinion is based on, but I think your points are valid concerns, if seemingly based on a lack of knowledge (therefore speculative) rather than informed.

In my case, when I said that I was referring NOT to his taking a risk, but him taking an unpopular decision for the right reasons: that he regards as reducing risk considerably (that is my reading and expectation from what he has said - evaluated based on my own experience and investigations).

I haven’t verified everything because what David has said stacks up based on my experience, as well as having learned he is IMO competent and of high integrity. Based on my experience and understanding of the issues here, the investigation I have done has only tended to confirm all this. So for me, less worries rather than more, though still eager and impatient to see us reach testnet3 like everyone who supports or is invested!

FYI I’m learning a bit about Rust rather than refreshing my rather out of date (though very extensive) experience of C++ (I’ve probably written several hundred thousand lines of production code in C++ alone). What I’m finding is that Rust is easy for anyone who has competence in C++, and for me coming back to these languages, I think Rust is easier than returning to and learning about the developments in C++ since I last did serious coding. Think about that. Plus Rust has seamlessly integrated documentation, testing, and testing of examples included into the documentation, as well as other features David has described but which I’ve yet to try myself. It is a tour de force, and I think this move will prove to have been a real beauty.

So far I haven’t heard criticism from anyone who claims even hobby level competence of software development, only from people who are rightly concerned, but who don’t seem to be able to assess the information presented or willing to do their own investigation before commenting.


Have you ever thought about why it is an “unpopular decision” as you say? Exactly for the reason i am trying to teach you here. But i won’t be doing that any longer since your attitude is quite rude.

ps. you might write a million LOC and still learn no deeper lessons if your attitude is wrong.

I think everyone here understands your concerns but most of us ultimately trust David’s leadership. This is his baby and he’s raising it with a competent team that he trusts. Any investment or expectation for delivery of this software is a risk based decision off of what proof we see on github, system documents, the openness of the team on these forums, etc. you catch my drift. If there seems to be some hostility I believe it’s just a defensiveness that is coming from a lot of negative posts of some folks that are on here for mainly their own self serving interests and are not happy with their perceived “lack of progress”. I believe you have valid concerns and are here for the right reason but I don’t think there is enough concern on this topic to outweigh the benefits

1 Like

I think we are quite on the same page here. I would have left long ago if David did not seem like a honest and competent person.

I just wanted to raise my concern, and honestly was a bit startled when such large and controversial decisions are made so fast and without any public announcement/discussion.

1 Like

I’ve made clear specific responses to your posts, while acknowledgng that you are right to have concerns. My purpose has been to establish whether you have any basis for them, which you have declined to provide so I’ve explained the basis I have for not sharing your so far baseless opinion.

1 Like

I think I understand your concerns, although I don’t agree with them. While initially they were going to push testnet3 with the C++ build, perhaps the rust port was smoother than they anticipated. A decision like this is huge, and the sooner the decision is made the better. It did take balls, but dirvine knows better than anybody what’s best for this project.