Data Chains Implementation Poll

That’s not true, I think.
The low probability of breaking safe is not related to safecoin directly.
Safecoin is a reason to keep storage, but on testing it’s gonna be by itself.

Safecoin is not a solution for DDOS.

All I’m saying is what you have to loose while testing?
If you worry about testing feature priority you can have parallel test as I mentioned.

Yes it was tried, I also mentioned what you said about bad experience, but SAFE got better since then.

If you put this test net parallel, you can always review the progress and have the TODO sooner.
ב-17 באוק׳ 2017, ב-JPL <safenetwork@discoursemail.commailto:safenetwork@discoursemail.com>, 22:47 כתב:
[https://avatars.discourse.org/v2/letter/j/a9adbd/45.png] JPLhttps://safenetforum.org/u/jpl
October 17
[https://avatars.discourse.org/v2/letter/a/7993a0/40.png] Antifragile:

. But non of them except me tried alpha2. It is shame. It should be much easier to use. For example sell invitation for 1 USD paid in btc or via CC. or let people verify via SMS, or risk the attack and let people play for free.

Would be nice but a few testnets ago some malicious person or people spammed the network and brought it down within a couple of days of launch. Unfortunately there are plenty of people who would like to see the project fail for whatever reason, so until SafeCoin is in place, which would make such DOS attacks too expensive, then some sort of limiting mechanism is necessary for protection.

Selling invites for a dollar or so is not a viable solution as any attacker with relatively deep pockets (who knows who they are?) could easily decide that a few hundred dollars is well worth the investment for the reputational damage and delay caused. Make them any more expensive and the barrier to entry would be higher than gaining trust level 1 status IMO.

[https://avatars.discourse.org/v2/letter/t/e495f1/40.png] Tal500:

Make everyone hostile and test how it behaves

That was tried, inadvertently, before. We know what happens - the network slows down then falls over. Not a good advertisement for those who want to try the Alpha!

The requirements right now are much easier than when I joined. I’m in favour of keeping the current membership process.

Anyone can spend $1.00 …
I think at this stage, paying with your time is a more acceptable payment to join the early stages, which isn’t much of a bother when you want to register 1 account.

Keep the requirements as is. It’s an excellent filter to keep spammers out.

7 Likes

I see this as “proof of staked resource” so it doesn’t bother me. Why should a node that has just joined not be suspect, whereas one that has been around for a while and performed well should not be trusted more?

4 Likes

The speed bump to joining is a testing parameter that will likely remain in place through early launch, as I figure it. It’s a scale matter. Once scaled to a certain (UNcertain, actually) size, especially with safecoin implemented, swarm attacks will be mitigated. Till then we have to deal with what we’ve got. I’d love it if everybody could easily check it out, but till it’s ripe that would do no good in any case.

We’re still REALLY early adopters who, somehow or other, got intrigued enough to study enough to get the picture. Broad adoption of Bitcoin is still just beginning to become possible after years of a fairly stable no-internal-compromise network. SAFE adoption should be much faster, but we’re still pre-launch.

7 Likes

Yes I agree and 1 to two hours is hardly an impose for those with the right intentions. A lot of the game alphas/betas are a lot worse to join, you have to ask, then be vetted then wait and contribute and so forth. Having alphas/betas open to all with no restrictions is actually rare in the development world, for obvious reasons. So until SAFE has the anti-Spam mechanism in place we have to have something.

5 Likes

Fergish you miss my point.

You should run a parallel test net, call it X.
In X, people are welcome to try make troubles. (Even the team)

This way everyone can diagnose security issues.

If there are problems with the system, it should be tested ASAP, not should be waited until a stable release is coming.

I believe that allowing everyone to kill you shows how strong you are. If you’ll die, don’t worry, try again.
ב-17 באוק׳ 2017, ב-John Ferguson <safenetwork@discoursemail.commailto:safenetwork@discoursemail.com>, 23:28 כתב:
[https://sjc3.discourse-cdn.com/business5/user_avatar/safenetforum.org/fergish/45/3030_1.png] fergishhttps://safenetforum.org/u/fergish SAFE Crossroads
October 17
[https://avatars.discourse.org/v2/letter/t/e495f1/40.png] Tal500:

I(and I think the team too) HATE trust level concept.

It started after some DOS attack on test 8 or such.

I see this as “proof of staked resource” so it doesn’t bother me. Why should a node that has just joined not be suspect, whereas one that has been around for a while and performed well should not be trusted more?

I’m not a tech but I’m pretty sure the “try as you will” will happen at Alpha 3 stage.

1 Like

The problem with that was that a storage spammer would fill all your vaults and the system becomes unresponsive and basically unusable for anything other than JUST downloading the files that were stored before the spammer filled it up.

It isn’t much of a test bed if the lowest common denominator can essentially cripple the network in no time at all.

Before something should be presented for hackers to break, it must first have the obvious holes plugged. It is rather pointless otherwise.

11 Likes

The thing is we are not near that stage, for instance I would be shocked if members of the team could not take that down in several ways very easily. Now the time for us to look at these and discard them as already known is substantial. So right now we know too many things that need done, or we would be launched :slight_smile:

There will be a time for that and we will not slouch as we move as fast as we can, but right now it would just waste everyone time. especially the app devs we want building apps.

tl;dr This would just be a time sink right now.

22 Likes