Rough idea about names of urls inspired by lbry

Features
1

So I start by telling about how lbry does it, anyone can have the same name (url) but there is a hash next to it and the most popular in the end gets the actual name without a hash.

So for e.g.

There are:
safe://example#3e and safe://example#4a

And in the end the most popular gets the safe://example for ever (until the winner deletes the claim about that name)

[Disclaimer I dont have expert understanding of lbry naming system, I might be wrong in some part in what I said above]

So I have a rough idea:

Why don’t we have something similar;

Many will try to impose a known website. So in safe right now is: first to lock a name owns it!

Why not allow many people to get a name and then figure a way of dealing with it?

My early thoughts on that:

Maybe have a system that goes like that:

  1. First instance Safe://sitename1#by:user1#hash:5g

  2. Safe://sitename1#by:user2#hash:f7

That will allow many websites using the same name. One then could bookmark one website that he likes in these ways:

  1. Safe://sitename1#by:user2

  2. Safe://sitename1#hash:5g

Why I think its usefull

A website named safe://wikipedia might be malicious but if we allow many people to make their own version of the website then people will choose whichever they like.

Also with the logic of lbry there might be a way to globally choose one implimentation of the website that is not malicious and is the best implementation so when one navigates to safe://sitename1 he gets to load the best implementation that people have ended liking and accept.

Result?

No more problems that will rise by people claiming a really popular name like safe://wikipedia.com and making a poor implementation or malicious one

I am really looking forward to hear your comments ideas on that or even an offical PR to implement something like that!!!

Edit: another idea when one enters safe://sitename1 he gets a list of all the available websites with info and comments by people

8 Likes
2

first off - i love it :slight_smile: :hugs:

maybe this here (it’s a simlilar concept suggested by seneca way back) could help to refine some thoughts =)
instead of likes the version number of the AD containing the link on the network could represent a value and therefore cound as ‘like status’ - the only pity here is that people cannot ‘donate’ likes … they’d need to donate safecoin to the site owner who then increases the version count … the popularity-ranking you suggest would ofc be perfect … i just atm don’t know how one can express it on the network …

1 Like
3

Interesting idea, but my first thought would be that legitimacy isn’t a popularity contest. If there were strong incentives to pretend to be a legitimate site, people may use bot nets to make a fraudulent site popular, for example.

This feels more like a search technique, rather than name resolution. I could see it having value by giving popular sites a rank boost in results, with the caveat above to consider.

As a replacement for NRS unique names, having an ambiguous URL, by not always going to the same place, does not seem suitable, IMO.

2 Likes
4

Its like emails, if you send info@google.com or info@yahoo.com

Likewise you would go to the site http://info#by:user1 and info#by:user2

If you go to http://info you are presented all the websites

1 Like
5

We can brainstorm and find the perfect solution

6

if making popular involves a cost this is not a problem - isn’t it …?

tbh i think this is a cleaner approach than having different versions of the same name where a newer name might go somewhere completely different suddenly (while you rely on urls being unambiguous)

everyone can sell every name to a fraudster … (or be infected by malware + loose their domain)

7

So it’s the same kind of idea like mentioned in the other topic?

8

I would argue that the other threads are old, and now that we are at the last line to releasing the working beta network we should figure out this thing.

I read from @c0dr also about command based naming

maybe “safe://example#by:user1” or “safe://example#hash:5o” could be also:

“safenet example @user1” and the command will see: website “example” by “user1”,
and for the hashing: “safenet example #5o” would be website “example” with hash “#5o”
and maybe by version: add after the name a ! with a version like !2 or !last or !first or !1 or !last-visited or if there is a verification (with an extention maybe) !verified
so it would look like: “safenet example @user1 !last”

I would propose we look into these things, maidsafe please gather all solution and make a PR where we will find the best solution!!!

9

maybe increment the cost for a name eveytime one uses it?

maybe increment it logarithmicly? so fist one buys it at 1 safenet next one at 2 next one at 6 and so one so someone looking to use a name would try to find a unique name to be cheap or if he wants one name he can pay as much as it costs.

from the other side I feel that anyone should have any name one wants so if 100 people get the wikipedia name and the 101th person is a really good programmer and wants wikipedia and makes a great implementation of wikipedia shouldnt he be able to use the wikipedia name without paying a fortune?

1 Like
10

Hmhmm when having a donation system that automatically increases the reputation of a name for the other person this might solve the issue :face_with_monocle: …?

A donation basket (with proof of spent safecoin in it that anybody can fill but nobody delete) maybe a append only data which anybody can modify?

1 Like
11

perfect!!! if people vote with their money is good, but it may be gamed!!! I propose there should be a verification initiative that you donate to someone that does security verification and a second one system of donating into a pool if you think that the implementation is good, but again it can be gamed.

the security verification should be detailed and a proof of non malicious of a website…

thoughts thoughts thoughts!!!

1 Like
12

Sorry for quoting myself :innocent:

But if it is possible to determine the unique visitors, wouldn’t this be better than rank based on money?

13

no just no. imagine a malicious website in a url name that is very rare, one guy that offers a service that people access few times would be not in the top results based on unique visitors in relation to a service that is malicious and people visit a lot out of ads spam or even by accident or by missguided clicking on a link based on false data!!!

do you understand what I am saying here? there needs to be a verification process of non malice and a way of services that dont get lots of unique visitors but the service is actually what a user entering the url wants to use!

and as I said before the url itself should list all the services that use that name and have a way of assisting the user know what is what, maybe a description maybe stats about visits in numbers of unique visitors or time spent or comments or rating or anything that will help a user entering a plain url go to the service that is what he wants or is non malice or best implimentation

14

yes, but I visit most websites by bookmarks. So I might visit a malicious site once, but that’s it. After that I bookmark the good one, and visit that one each time. More people will do this, so slowly the good site rises on top.

15

again if a service is a low unique visitors, maybe its a service that you use once a month then you and all that have it in a bookmark can access it. but imagine someone goes to the plain url for the first time because one heard about it by people then he should not go to the service with most unique visitors but he should be presented with a list of all the services and see stats and info for each and find the one his friends have referenced!!!

edit: for e.g. safenet

if you say safenet there are other services that have that name,

you would say to your friend safenet by maidsafe

when he searches safenet he gets results and when he sees somewhere maidsafe he knows thats it

the same with the urls!!!

16

So, something like this?

1 Like
17

yes!!! the only thing I wanna add is that it would have lots of stats info and maybe comments and ratings and anything that would help a visitor of the plain url get to where he wants or discover the best implementation/non malicious

18

I got a question why is there two urls one referencing the other?
are these “shortcut” urls for the url on the right side?

also in my proposition with the command like address:

safenet news

result

-news @user1

-news @user2

19

It’s explained in that particular post

1 Like
20

ow damn I am bored to even look up the whole post, or rather I didnt think of that, sry - now I know what to do from now on :smile: !!!

1 Like