All me old fashioned, but I like the idea that all data is equal. @jim costs the same as @mij but it’s what I do and say and create with it that builds its value.
1 Like
But what about the question of DNS name parity to improve user experience and prevent fraud? Isn’t this a bigger issue than just getting a ‘good’ name first?
I’m not saying NRS is the only system we need, but it’s one component, and it doesn’t need to solve all problems in one go. It’s got a purpose.
Nor am I saying we’ll ever have the perfect network… it’ll always be improving.
But we don’t need to get stuck in a centralising mindset because that’s where we’re coming from. We’ve got more tools to play with here, and some great foundations.
5 Likes
I’m both lost and tired. But could somebody explain to me why these “services” are needed and why they cannot be replaced by just paths? Maybe in a new thread, if it’s complicated?
I was thinking only trustworty sale by 3rd person would be kind of opensource smartcontract, where 3rd. person owner does not see key and new owner will see it straight after succesful payment.
1 Like
I didn’t want to say that we must resolve the issue right now and find the perfect solution… But I think we should be well aware that this issue most certainly will arise and we probably should be prepared to test some naming systems after launch that tackle the issue of name squatting (imho it should be a client side solution and the network shouldn’t know about names to keep stuff simpler - but I can be wrong ofc ; )
For Fleming and as first version of it the current system is probably just perfect (simple and easy to communicate) +we’ll see the same problematics as with more complex naming systems too (broken links because they don’t include the explicit version)
2 Likes
Before reading this thread I would never though that list of names can be such a complicated thing, but here we are 
Maybe we should pretend DNS doesnt exist and look for examples elsewhere, this problem existed before internet. I will borrow the Joe Plumber guy - I live in small town, I know Joe, he is unique to me. My friend lives in another town and there is also a Joe Plumber, he is unique to him, so far everithing works. If either of them decides he wants to operate in both towns, it leads to name conflict. When this happens people tend to start calling them Joe Plumber from A, Joe Plumber from B. When you say Joe Plumber, you will get question: which one of them? So if Joe wants to be the first one who comes up, he has to either invest heavy in propagation, or market himself under more unique name. An if Joe wants to be THE Joe on whole Earth, he has to be unique, or invest more than any other Joe.
My idea of the naming system is not completely new, it is more like a combination of some ideas that are already here:
-
Names are not unique, you can register whatever you want (maybe some minimum length)
-
There is a fee for registering a name, you can pay some predefined minimum or more.
-
If the name already exists, the registration fee is anything higher than the last price. This should limit spam-naming popular names, keep the NRS database in reasonal size and still allow anybody register whatever they want.
-
When you enter non-unique name in the browser, in first version you will get simple list of all pages with this name. In later versions there is lot of space for various metrics how to order/filter/search the list.
-
What about google, bankXY? They are probably not going to come first, but they can raise the registration price very high (which can be used also be used in serch results) and new scammers will be discouraged by high price. Without ne scammers, the initial ones will be quickly known and can be flagged and hidden from the listling, so you will know that this is the right bankXY.
-
The principle of perpetual web is not broken. In 2198 your “superman” site is still there. If you called it “superman2019myvillage”, no problem, if it was simply “superman”, it will be probably pain in the ass to find, but not impossible.
2 Likes
You think you’re tired now? There’s a whole thread on that over here Protocol://Domain.Subdomain/Path
3 Likes
I think the main reason to differentiate is to have a clean way to offer different functionality.
Most people so far use them just as subdomains for different websites, which is fine, but if you want to offer a different service such as LDP (for Solid), WebDav etc, instead of HTTP (for a website) using subnames is a clean way to offer this. I’m not sure how we’d do this if we used the same mechanism for both directory paths and services.
2 Likes
cc @Sascha
I think subdomains in the current form are useful because there is many for e.g. web mail providers. its easy to just type mail in your address bar and if you have visited mail.google.com and mail.yahoo.com then you are presented with these options, so its kind of a categorization of services cause you look for the service first cause you use a service from few providers, where if you use all google services and it was the case that first goes the organisation (google) and then all the services you would type in the address bar google and be presented with all google services you use! which are a lot if you use them all!!!
visually:
if you use mail from two providers but you use ALL the services from two providers
typing mail will give you
if you want to go to your mail by google in the other senario (first provider then service)
typing google will give you:
google/search
google/images
google/news
google/translate
google/maps
google/docs
google/mail
given that you use the services in this order of frequency!
It has a downside. They introduced fishing. I can easily create mail.yahoomailservice.com
Lazy people scan the first few characters and are caught. Skipping the order to yahoo/mail is more safe
I agree on the fishing, I am just giving a thought on why it was implemented like that!
1 Like
But this is just a question of how the browser searches for suggestions. I don’t see it as a reason to have an extra entity for service/subpath. I often forget the exact URL, so I just type the part I remember, even if it’s not the first part, and currently I get suggestions for all URLs containing that string in any position.
This sort of makes sense to me, I guess. Maybe. Hmm. But I was under the imression the protocol was stated first and the structure was the following:
protocol://service.name/path/subpath
(or maybe “service” would be called “subname”)
Are you saying there could/should be
safe://ftp.sascha/files/pictures
on the Safe network?
No, I’m still lost. By the way, I don’t think hiding the protocol by default in the address bar makes much sense.
1 Like
I used ‘protocol’ generically, so you are right to point that out. Sorry to confuse. My point stands, but instead of protocol substitute ‘web services such as LDP, WebDav etc.’
EDIT: I’ve reworded my reply so hopefully it is clearer.
1 Like
This is looking very good, hope they keep it up!
1 Like