What happened with the AMA questions on the last Dev Meetup with pods?
1 Like
Nothing happened. I was surprised no questions from this thread were asked by the Pods. There was just an example demonstration and program language discussion then it ended abruptly.
It might be a better idea to take the entire list of questions and to let them be answered one by one by David or someone else from the dev team.
So a new video for every question and just show the question in screen at the beginning. If youâre going to do this during a Skype call, the sound or connection might be bad.
What do you think?
2 Likes
I would happily do that. @ioptio is coming over soon and one idea was she interview me. This may be a good idea and we could use a proper camera etc. rather than skype etc. which is always bad. We could do a question at a time like that and perhaps even persuade the whole team to say hi 
8 Likes
To the Pods or by the Pods?
I didnât want to repeat what I posted a day earlier, so I didnât ask any questions.
If I did, I would prefer to type my questions so that they can be replied in text as well and saved for sharing. Itâs still difficult to search for words in audio and video recordings.
@Melvin The âinterview (face to face)â format would provide better audio and video result.
@russell this could also double as material for the documentary, if youâre still interested in starting one.
@Shona Hopefully, youâll have your Project SAFE shirts in time?
@janitor no worries, I suspect the meetup was very focused on testnet1 and it did take up a good amount of time. So the AMA would be better served as part of an interview.
Weâll get it done, eventually.
2 Likes
Thanks. For my personal feeling, thereâs already far too much explanation about this project hidden in videos. I donât like videos in general, and especially for explanation and tutorials, because I canât search in them to go back to that particular information I remember was explained somewhere, and I canât skim them to get a quick overview. I really have a very hard time trying to understand why video tutorials are so popular nowadays. 
@phrphz Videos are not intended instead of traditional documentation - thatâs a big work in progress - but rather as an easy to digest source of information for those that do like them, and we know there are a lot of people who do. Also to help those (me included) who want to get the gist but not necessarily start coding. I also like them because I get a feel for the people, community, MaidSafe etc. So personally I try to watch them even though Iâm not really looking for the factual stuff they contain.
Heâs saying that in his view the effort (time invested) should be tilted a bit more towards documenting rather than recording the videos (assuming the same people can do the both types of work). Letâs say these activities consume 10% of their time and itâs spent 10% on docs and 90% on vids/interviews. Maybe 15-85 would be better.
I also like videos, but that doesnât tell us about the overall marginal utility of the participantsâ time. I think the P man is right.
But then there are dependencies. If the code isnât ready, itâs not necessarily useful to constantly create outdated and incorrect documentation. Still, Iâd like to see existing docs fixed and updated (in a word, maintained) rather than more vids. Just my opinion.
Iâm not sure if @phrphz is saying that. Best to let people speak for themselves. You are saying that which is fine.
The decision to produce these videos is a result of an earlier discussion of exactly this issue, though it seems to have moved from âwe can make them as part of the PoD hangoutsâ back to âletâs have David sit down and run them offâ. I agree it is a valid question about best use of Davidâs time, but heâs best person to judge that I think.
From a dev point of view, documentation is an enormous time consumer. A new Boost C++ Library can expect to invest about 25% of total development time on documentation if it expects to pass peer review, and that is all time not spent writing or debugging code because only devs can write documentation for a company as small as MaidSafe. In fact there is an argument going on right now in the Boost dev mailing lists about upping still further what we demand of documentation for a new library, and the bar is already set so high that no large library has passed that bar in years.
There are many high level non-video documents written on SAFE already. People perpetually seem unable to either find these or read them or understand them, which is fair enough. However there are also dozens of articles in the press which also deliver even higher level overviews. All appear readily in Google searches.
Iâll be honest here: I didnât find a problem with the quantity or quality of the documentation when I was evaluating whether to sign on with MaidSafe, and I had a good enough understanding to say yes after only a few hours of searching. The technology is very straightforward, almost conservative.
For those wanting docs on SAFE application development, I do feel for you because there are none apart from the source code itself. Such is the price of being in the vanguard of a new technology platform, and it is no different anywhere else. If you want to be the first in, you pay the price of a high barrier to entry.
Niall
2 Likes
Can somebody respond to questions from this thread?
Mine are here (above):
I see how I can install the s/w, but I canât find instructions on how to use it. Should we read the source code for that too (Iâm not trolling, just asking), or wait for how-to instructions?
It would be nice if before the next testnet some instructions are provided. I am not asking for a detailed explanation of everything and as I said above I agree that itâs a huge effort (as stated by @ned14 as well), however without any instructions whatsoever itâs going to be really tough.
If that would be easier, could a ready-made VM without IDs be put online for our download? Then weâd just need to create the accounts and restart services (or reboot the VM if theyâre set to autostart).
The examples (only one currently) show you the basics of running your own test network, and you can also run tests âmake ExperCommonâ etc. Not very exciting but worth doing just to ensure youâre ready for when the networks open up and we can all join together. At that point, Iâm sure Viv and others will put out instructions as to how to join the network, how to make and run the sample apps etc. Its all very exciting and not long to go, though I have accused @Viv of teasing to build up interest for the launch, the old âcuriosity dollarâ heh!
Thanks.
That is a strange place to put âgetting startedâ info, but okay, at least now we know.
Link here:
On this page - Build Instructions ¡ maidsafe-archive/MaidSafe Wiki ¡ GitHub - there could be a link âNext stepsâ or âConfigure and runâ.
1 Like
Something called network coding or random linear network coding which aims to replace TCP/IP, which sounds a bit like soft radio tech for wired lines had the lead developer in trying to describe it saying, as paraphrased by journalistâŚ" he believes communications is really storage, because a buffered packet is, in fact, stored." Which sounds like Project SAFE
@janitor thanks for posting that, I was looking for that as well. I agree it should be added to the build instructions - it talks about being a dev there, why not also mention how to fire the thing up as well Thatâs the first thing I always want to do. Anyway, it was exciting to see my test results on the public dashboard and Iâm looking forward to joining the network whenever I can.
1 Like
âNetwork setup completed successfully.
To keep the network alive or stay connected to VaultManager, do not exit this tool.â
4 Likes
Yep I got to there too.
Now Iâd love to know what I can do next to actually connect and be more useful in testing some of this great initiative.
Assume I know nothing of C++ and not too much of anything else.
1 Like
In case you missed it, from the Docker thread @dirvine replied to our question of what next?:
We made some really big changes two weeks ago in a mad rush to get a testnet example for a google hangout. We bypassed every Qa rule we had and broke the testnet. So we have stepped back and are making the Qa process way more rigid with some pretty large investments. @BenMS was telling me there will be over 80 automated builds on a commit now for pushing so this will hammer our code and make sure we do not do that again. I have also set up another 20odd CI machines remotely waiting for a push to master and we are incorporating jenkins/gerrit for test and review.
With the new devs starting now we need to be super skilled at this process, but its looking pretty strong and will allow us to keep creative. I suspect this week and next will bed this all in place. Then you will see some pretty managed and fast releases of hopefully many examples and mini projects for folk to work on. @Viv has is team starting to look at this now with some devs in India doing language APIâs for the codebase. It should allow us to start bringing out apps very quickly as well as the portable apps that @happybeing mentioned.
Should be a great few weeks though with a lot of change and the background work will be manic as ever with a much more managed front end. rUDP is looking pretty amazing now, so I expect to see a lot of testnet activity again. It has been a completely frantic two weeks for sure with all the back end workload being enough to swamp a tam 10 times the size of us. We now look to be in a terrific shape though.
There should be a much simpler dev environment soon as well which will make is a one click affair to get started and get apps coded.
1 Like
THank you - now have local_network_controller running fine in a Docker container.
Now for the visualiser⌠
1 Like