SAFE Network Client Testing Commences

same for me. took me four login attempts to get in and now won’t load the demo App. Says ‘Could not connect to launcher’. A bit glitchy at the mo. Worked like silk for me for the past few days.
Maybe work is being done (late for Scotland)?

It’s one of the bugs being fixed, client silent disconnects. AFAIK it is already fixed in routing, but not 100% sure.


We don’t need an escrow site. I am willing to pay you 10 bucks in BTC at the course of 1 Jun 2016 if the Safecoin tokens are implemented in the network.
Important: I am referring to Safecoins, not TestSafecoins!

You can set a reminder, I will do the same. I am a long time user of this forum, so I will not dump my identity and run away :smile:. If you accept and will do the same, let me know, we can for sure tie the knot.

I’d say MVP is proved at this stage. Well done.
You should switch off. It’s 11pm!

1 Like

You sir, have a deal. I will take your bet. NLT 1 Jun 16, Safecoin
tokens will be implemented in the Safe Network. The winner receives the
equivalent of $10 USD (valued on 1 Jun 16) payable in BTC. I am certain
that both of us, for the sake of liberty, are hoping that I win this
bet. In any case, good luck my friend!



It’s my time now, 6 am last two nights, but makes for helluva long days so it does. We will move forward again this week and the next few till we have vaults in everyone’s homes and start the messaging and safecoin rollouts as soon as all looks well :wink: We creep along now with regular solid updates to the codebase. Nice thing is more folk will see some of the work at least. Most 80% plus is all behind the scenes stuff, but now even there we should see speed and feature improvements.

All in a decades work :smiley:



I would appreciate so much if anyone could help me out on how to install the SAFE client (and wallet) to Linux Ubuntu 14.04.

I tried the Source code (tar.gz) from here:

but could not make it run.

I already installed the libsodium version 1.0.8.

Thank you guys!


How to download and install SAFE Launcher


would be great that safelauncher requests PIN each time that you want to grant permission to an app.

1 Like

It depends.

If the app is just an exe and nothing prevents me from patching or replacing the exe of a permitted app, why bother with getting a permission at all?

Welcome to the world of trusted computing :slight_smile:

The launcher should securely store hashes of permitted apps and linked code like libs/DLLs, otherwise it’s not very meaningful to ask for the creds every time.

It does AFAIK, well at least this is the plan. Too busy in lower layers to confirm, but for sure it will. It is also how we recognise upgraded apps etc. as well. @ustulation or @Viv will hopefully confirm, but this is the release mode for sure,

By extension, then, it would also be possible to create app reputation ratings, although one has to be careful with that (both when creating and using).
But at least few core apps (including the launcher, but that one with PGP keys) could be signed by the Foundation or something like that, so that one doesn’t have to start from 0 in terms of trust.

1 Like

If Safecoins are implemented : price rockets, you make 10$ like pocket money.

If they’re not : you win, you make 10$

This bet doesn’t look very fair.

1 Like

Yes this is an interesting angle for sure. The issue will be the ability to check the rank. perhaps showing somehow this is the in use version by most folk or similar. I have not thought it out yet. I would prefer all portable apps where we know the hash of the software and it goes through some kind of beta test then have a huge number of folks form a group we all know sign they are happy. We can use multisig for this. But again who watches the watchmen ? I have the same issue with reproducible builds though.

I am sure we can do something a bit better than today but not sure how much. Kernel loaders etc. would be great to also confirm, it seems better than building an OS and hope everyone uses it, Then again we can prob confirm win kernel seems OK but may be already corrupt, probably we could do this with only open source many build reproducibly and then sign?


Yeah if one wants to go deeper, there’s always more to do. But at least for this immediate “bootstrap” level of code (the launcher and the browser, for example) it would help a lot to be able to validate the launcher and then allow access by a validated browser.

If someone screws up after that, or has a compromised OS or low level firmware, that’s seems very out of scope, at least initially, and if (for example) Tails ISO is used to boot a VM, then the next two pieces (launcher + browser) could help everyone get to SAFE … safely.


Regardless of the technological means, in the human end of the equation it is always about basic trust in some sort of authority (“they know what they are talking about it, and I deem their character worthy of trust”.) If a dozen people of good reputation sign something, I’ll probably decide it’s legit. Especially if they are coming from different circles.

The point is, now I can decide whom I want to trust, instead of having to depend on a faceless corporate entity.


Well, the previous implementation of the launcher used to store the hash of the binaries securely to prevent/detect binary swapping.

But in this new approach, it is a know limitation for now. We have few areas to improve, like the one listed in the unresolved questions in the rfc. We are yet to integrate the app reward mechanism. Once that is place, it would enforce a secure process for validating the application/vendor. But the reward mechanism will take some time to be rolled out. We can expect that only after the SafeCoin integration and testing.


I like that this has happened!

In a way this gives us a date without putting any pressure on maidsafe.

The next time anyone asks when SAFEcoin is due we can just tell them there’s a bet for +/- that date. Simples! They have the satisfaction of thinking they have some vague idea and no one is responsible for it, nor does it add any pressure to the devs; they’ve wisely avoided making any predictions themselves at this stage.


1 Like

Serving an audio file from the test network.


If you happen to test a listen of it, I had to make it fairly short, (current file-size limit) and being tired at the time put a very long fade off at the end, so there’s nothing wrong with how it plays, just not well made :slightly_smiling:

Oh, and when I say “served”, it’s just that, it won’t be streaming from safenet, it’ll be download-and-play.

By way of feedback, it took quite a few attempts to get this up, I had to keep coming back to it, so things are still fairly variable and patchy right now (or at least they were last night). When the upload failed, things seemed to get very slow and then time out or something and I noticed at the same time that another site I have up wouldn’t load, so, that would seem consistent with the network coming and going I guess?

Anyway, none of this detracts, very keen to try some video served as soon as the file size limit is extended a bit. Great fun, cheers all.