MaidSafe Dev Update :safe: 21st September 2015

Now the real fun is about to start, hopefully after this sprint we have the first global and truly public version of SAFE up and running!

I also really hope the sprint can be finished in the allotted time. It’s a big one again, and those tend to require extensions so far. I still have hope for test_safecoin to be up and running at the start of the new year!

5 Likes

Could anyone give a non-tech update?

  1. What is the status of the Safe App Launcher?
  2. When do we see Apps?
  3. How many sprints are required until Safe App Launcher?
2 Likes

Sure :smile:

It is currently in it’s planning phase. Launcher RFC is currently a work in progress. so questions still exist in it’s implementation design which needs answered and reviewed before it’s accepted for implementation (we hope as part of the rust-5 sprint. not confirmed yet though)

Sorry gonna be a long answer for this one. It’s quite a generic question. “soon” is probably the best answer for this. depends on what we refer to as apps. If we are talking about apps that you can use in the desktop environment showcasing a feature, well you have dns examples and sorts already to try. If we refer to third party apps interfacing with the network with the help of launcher for data storage, that should be possible once the launcher is in place with the nfs REST API. If it’s apps related to features not exposed yet(messaging, safecoin, POSIX friendly data storage…) gonna be when these features are implemented in the network layers itself. Mobile apps - when we establish the port layers for the corresponding platforms.

This should be answered when we conclude with the Launcher RFC work in this planning phase for rust-5 Sprint ( before rust-5 starts it’s implementation work). Once the RFC work is complete, the tasks required and the effort needed will be clear for us to draw the exact estimates of when these should make their way to production code.

Thats exactly the reason we havent fixed the objectives as “all of them” for rust-5. We’re hoping to have the work required laid out with RFCs in place during planning and based on the results from it, have an accurate estimation of the work we can achieve in meaningful fragments as deliverables.

10 Likes

How long will the planning for the next sprint take? I know I read how the sprints are going to be short going forward and was wondering if the next one will be 8 days like the previous?

2 Likes

Happy Sprint everyone! Rust-5 looks to be huge.

3 Likes

I´d also be interested in how long the next Sprint ist going to take.

Another thing: I feel bit stupid but can anyone explain to me what one can extract from the transcript. i suppose it´s a list with what people are up to currently, but the bulletpoints do not seem self-explanatory to me. Any help/advice?

Very exciting. The thing I see as most important to me is the hole punching with ability to connect on a global scale with what is already available on a local network.

So I can move further on hardware and mesh.

4 Likes

Currently we are estimating this planning phase for rust-5 to conclude next week(02/10/15). Unless we hit any blockers with any of the RFCs, this should be comfortable enough to finalise by and then list the final objectives for rust-5 and it’s duration for implementation.

We can almost certainly say implementation for rust-5 is going to be longer than 8 days. rust-4 was a specialisation sprint which didn’t introduce new features into the system. Ideally we’d hope to have the implementation within 3 weeks(maybe 4 as max).

5 Likes

This project needs to push forward the code bounty program very aggressively. Don´t get me wrong, I think all of you are doing a very good job, but the breadth of this project is too much huge.
Just my opinion. :thumbsup:

3 Likes

The bounty program is growing, but it can also be counter-productive if it diverts MaidSafe staff from activity they are best placed to do, so I think it needs to be balanced growth rather than pushed aggressively. This is for the core, which is indeed a big and broad project, but bounties are not needed for people to develop apps and services on top of SAFE Network, since they can be done speculatively for rewards through use, or as part of the sharing economy of open source.

4 Likes

It has been kept up because that was what we started with, so it is not so useful now we have the more explanatory Dev Updates. I still read it though, because I find it helps me know who’s who and who’s doing what which is not usually apparent from the Dev Update.

For me it still provides helpful insight into MaidSafe as a whole, new recruits, non-technical folk and so on, all of which I find useful even if the bullet technical points aren’t always understandable.

3 Likes

Thanks for the info.

1 Like

Ok, then we arrive to the point. Too small core dev team for a too wide huge task.
I think the solution for this problem is not burn the current core team. The solution is hire more people. Yes is well known that Maidsafe is looking for devs and is a hard task because very specific background is needed to fill the place. Ok, then use the bounty program to be familiar with other devs, maybe you find some good people unexpectedly. With the current system in bounty program this is not gonna happen.

That’s your opinion, not a fact. The bounty programme has been going a few weeks and is being expanded as we see in the OP. The dev team have been making great progress which and we can see things really happening, so I really don’t see the reason why your position is so negative. More bounties is good, so long as it helps, and it is already happening at an accelerating rate.

4 Likes

We will see it then. I hope yo are right, really.

That’s your opinion, not a fact.

I think the bounty system is absolutely in its infancy and we need to develop it. I would see its involvement in Rust-4 as very successful from an engagement perspective, but with hindsight it took quite a lot of our core devs time. I believe a big reason for that is that we did not spec the tasks with nearly enough detail, something that we are going to be focussing on much more in the next sprint. So, I think if we tweak the process, making continual improvements, we should be able to arrive at a place where we are leveraging the talent and passion of the community in such a way that everyone involved benefits.

11 Likes

Woot anonymous communication is on its way. When we start developing the actual network and connecting with one another will we be able to share data as well? Ie anonymous file transfers and data storage/sharing? If I understand how maidsafe works probably… isn’t it just uploading something to SAFE and sharing the bookmark via a message of some kind? Will personas be in play during the next sprint or will that come later?

Also if we upload anything to the network is it permanant or is the network going to be wiped at some point before launch?

2 Likes

Got the chance to participate in the bounty program and as you said it

Most of the tasks where written in anticipation that main core dev would be working on them and have working knowledge of what is needed. That said i like the current process so far though it would be nice to have MaidSafe have a dedicated person to handle the community pull requests.

This is something that would be fixed with having a dedicated person to handle the bounty program tasks. Most of the time you would find that when feedback is provided on the pull request am either offline or the reviewer is offline hence a lag in pace of things. I guess this is the case for other participants since contributions are done during our free time.

Been a rust newbie I do appreciate the time and patience taken by the core dev to facilitate the bounty program. Thanks.

5 Likes

I moved 2 posts to a new topic: Crust doesn’t compile

1 Like

absolutely :smile: we should be showcasing the same with the messaging examples as well as part of the messaging feature

yeh the bookmark essentially would the DataMap or DirectoryListing to the chunks you want shared and the other person can then put together the pieces to get the original content. That way you’re not transmitting the actual content across to the other person thereby also saving bandwidth and time in transferring huge file sizes across users in the network

If you mean the Vault personas like MaidManager / DataManager … they’re already in play and will need to be be for the network to function.

Probably best to assume it will be temporary for now, while we still work heavily in the core libs and iterate through quite a bit of code updates.

3 Likes