^This guy’s the real deal
5 Likes
Well, that’s why I’m coding a web app instead of a local one. The browser loads it from SAFE every time the user opens it. It currently keeps its Launcher authorization token in HTML5 Local Storage. Of course, this relies on the browser being secure.
9 Likes
This is how remoteStorage.js apps work too. I have a couple almost ready but there are some spanners in the works due to CSP/CORS 
2 Likes
Hey I like local apps and storage. I grew up without internet and there are still times I have to make do without it. Not being dependent on the net is a good thing. I get what you’re saying but I’d suggest what we need would be akin to an airlock of some kind, kind of like how if you’re in a preasurized space you don’t “trust” a depresurized space else you’ll be eating vacume. It’s not like going out EVA or whatever is a bad thing, it just has risks attatched to it that need to be accounted for.
If an app was uploaded to SAFE and the local copy was left unchanged, would the hash change if it was uploaded a second time? What I’m thinking is one could upload a clean version of one’s app/document/files whatever and have it vetted. Then use that as a control. So if your files got infected or corrrupted later on you’d have a clean copy to compare them against.
Personally I don’t fully trust apps that only function online because that requires that one BE online all the time in order to run them. Not everyone has a permanant high speed internet connection. I know plenty of people who don’t even have internet. So being able to use SAFE at point X, remotely take away your data or run your apps and then reconnect at time and location Y is important. It would be no good for farming but farming is only requied to actually upload data to SAFE. So say an app required to upload data. You could set it aside in a local file(s), estimate how much it would cost in safecoin, and then check, thus updating the price, and ask again to upload when one reconnected to the network. Again that would require the need for an airlock.
2 Likes
I’m not sure if this is possible, but if OfflineFirst apps could load from SAFEnetwork and if you later don’t have internet, could load them from your browser cache, you would get the best of both worlds.
OfflineFirst apps continue to work if you lose your connection and automatically sync with your server - SAFEnetwork in our case - when you next connect. This is what I’m building with RemoteStorage.js and it’s neat. Haven’t tried what I just suggested tho. Yet 
8 Likes
locked & compressed data chains #177
A potential issue with data chains is the ability for old keys to be re-introduced. Old keys are analegous with “toxic waste” in snark type bootstrapping, i.e. they need to be destroyed. In our case these keys need to become irrelevent as destruction is not provable with digital information (copies can always exist and should be assumed to). Another issue is data_chain size, which does grow and may become cumbersome even though it splits across the network. This RFC addresses these points and more.
10 Likes
This rfc (if I understand correctly) seems it will speed up operations of the network on the client side? Any laymen explanations are welcome 
3 Likes
Haha, I just made a screenshot for this one but you beat me up to 7 minutes ;-). Thanks for sharing. We have to make several new topics on this forum about the new RFC’s as well. Lot of action this area.
5 Likes
This is how SAFE responds to criticisms with actual action, instead of words 
4 Likes
A decade ago if someone asked if you could make a decentralized internet they’d have told you it wasn’t possible. We are making history here folks. Don’t ever forget that. The worlds full of people saying things aren’t possible until somebody actually does it.
8 Likes
We are TRYING (the maidsafe team is)
Let’s not forget this is still unproven and something that hopefully materializes in the future. We need to be objective and talk about things in the proper terms. The last thing this project needs is being at the receiving end of accusations like onecoin etc.
1 Like
I can’t remember who said this but it was something like “Idealists invented the airplane. Cynics invented the parachute.” Consider that even with each test net, each bug report, succeed or fail we learn something new and advance. Overcoming the impossible is largely an engineering problem not a matter of physics. If we wanted to create an airlock between offline apps and apps hosted on safenet I’m sure we could do it. The question is are we willing to put in the time to actually figure it out? Same with the project at large. It’s not a matter of if but when.
4 Likes
The issue is the autonomous management of a p2p network with no outside influence. That is something never done before, and may prove impossible. The other aspects have been proven and tested to a basic degree, but a functioning autonomous network is yet to be seen.
I am not trying to be a downer, at the very worst this technology must transform into a managed node based “cloud service” with redundancy via trusted nodes.
But the original goal, of having something that works, resists abuse, autonomously, is still totally unachieved, and saying so should not be a heresy.
2 Likes
True it hasn’t been done before but when I see SAFE I remember and think of things like this:
THAT wasn’t ever done before and everyone thought it was impossible.
1 Like
As i said, i don’t want to be negative, to halt the effort or anything like that. Just that the discussion stays realistic and no persons invest in maidsafecoin thinking that the final rollout is just around the corner.
I wholeheartedly support this project and wish my small investment will bear fruits, but at the same time i want we stay true to what is happening and don’t try to capitalize on people who don’t know the whole picture.
No, certainly not sure. I have not even heard about this project so i may very well be mistaken. However i cannot research it in depth right now so i must assume you argue it can resist abuse to at least same degree as bitcoin? Meaning for all practical situations it is infeasible to attack it.
If so, and if it is comparable to safe implementation, then there should be no issue.