Service receiving requests?

Is there a sense of how a “customer” might make a request that is visible to a “service”.

There’s a good sense a lot of data is static… I wonder there’s been talk in the past about messaging and even inboxes… but unclear to me if there’s a clear sense of what options there are for a request to be visible to those who might respond to it.

I wonder is a public option to write to a bucket that is only accessible to one key, is possible… I suppose that falls under the obvious public key signing but if there’s something necessarily one step beyond just putting encrypted files in public, that might be ideal??.. if private space is more than that.

Then I wonder too if there can be a known location that is owned by one and writeable by many.

I can imagine a whole host of services but cannot quite see how I wonder know there is a request. I’d tempted that safe urls might have the first element as not ownable, so, everything subordinate to that could be anyone writing their own domain/request#123 perhaps but are there better options already understood?

I’m wondering that it’s ideal is a service does not need to be online to catch messages and just responds to what is collected to their bucket, when they prefer to look.

?.. I would search the forum but it’s all a bit generic and not sure what to look for… I’m trying in part to wonder about our not just replicating the past and what exists on the normal internet but drawing something from the difference of what Safe Network is. Some opportunity to do everything differently??.. :thinking:

1 Like

If I remember correctly, in Safe Network resources should be available without dependency on online status of some users.
Which means no servers (services) in traditional form.

Yes, obviously no servers but what options in the absence of them?..

So, you can have a bucket into which requests are put… call that an inbox, if you want to be traditional… but any sense of what else is possible, might be useful to dream on what can be.

I’d be happy if it is that the network is a data resource without any service as there are option for that too… but there is no point creating what is possible from the network itself.

So, locations owned by one and writeable by many, I wonder is one simple option…??.. and also push of what difference when that arises. Thinking this on the back of the network does not want unnecessary traffic forever checking a location, where the simplest options for catering for normal functionality are worth the difference.

I do not know correct solutions, sorry.
One more addition: there were some problems with notifications about data change in Safe Network.
Do not know if this is still the case with nowadays state of SN development.

Yes, I don’t expect one person to know all the answers… hence the thread.

The question is more where is the answer for services best landed… as there are to be expect many layers and the first layer might not be the place for all functionality.

Reason obviously I’m asking is that even at an early stage, an option on two way or prompted communication, release alsorts of options for real world applications. If you do not know where to look for what someone wants, how to provide it…

So how do we enable a watchdog for all assets owned by an account – something that scans all the assets (NRS urls, wallets whatever) associated with each of your accounts?
Then when there is a comment on your site all about cylinder head bolt metallurgy in 1950s French petrol engines OR your pal belatedly returns ( in MAID) the £20 you subbed him in the pub last week you get a notification (where and how?)
Each “service” is totally independent of the other - the only common factor is that you have the passphrase/password combo ( is that still a thing?) for both.

Seems to me there may be room for a watchkeeper app/service that is opt-in but highly recommended for each account/asset you create.

I gather this is just a setting in the permissions for the data.

This is the only method I’ve seen from the development. The only messages otherwise are the responses to requests. Any other method, from my understanding, would require your client to be always connected since there is no server you’re connected to that will proxy (in some way) for you.

I always thought that all stored data on the Safe Network is “public” in the sense if you know the XOR address you can read the encrypted data. Its the decryption that that is needed.

Perhaps what can be developed as YAAPP (yet another App) is one that knows where to look for messages to you from all the various sources out there. Maybe the Apps out there will write a message to your “account” (maybe many IDs) notification inbox and this YAAPP will scan it and give you an ordered & curated (your specs) list of the notifications sent.

2 Likes

A diff there of naive user encryption and what Safe will do to ensure it is only available to recipient … so, a hole into which you throw your request and pray the owner knows to pick up on it.

1 Like

What you say reminded me of the change to allow private data to be deleted and that was to use the owner ID to restrict access. But for an APP to have another person notify you by posting on a forum or other such thing then it cannot be private data. A service trying to notify you (or send you data) may not be from an App you are running but another user, so then cannot be private data.

Another statement of the bleedin’ obvious here but…

“Unresponsive” apps/websites will fall by the wayside if the owners do not interact in the necessary manner with data that is sent.
Successful apps will be those where the operator puts mechanisms in place to ensure incoming data requests are dealt with in a timely appropriate manner.
Maybe we will just have to accept a far more hands-on approach for certain things we have previously taken for granted… The price of security…

Yes, I’m looking for that base network functionality that is not app and perhaps is not there…still I do wonder it could be possible for the network to alert owner as it writes that new data to private space… as an option. If inboxes are possible, then surely that should be possible?.. and to note the diff is huge. So much of privacy in a private network is predicated on authority and this follows as a query for that to be possible. The interaction between a user and another is not subtle enough for real world interests in many kinds of profile interacting. Those kinds flavours of interest call on others as trusted service to confirm xyz… and ideally that is prompt in order service knows there is a request.

Considering only the owner can write to their private data, I am unsure the point of it.

If you mean sending messages to the owner (ID) then its much the same as the owner looking at their “inbox”. The client would still need to look somewhere for the notification. If you mean the node a client connects to when logging on then its the node looking somewhere for the notification. In the current web, its the server responding to a request with that notification, like in browser notifications do now.

EDIT: The App wanting to notify the ID would I think be sending a message to the ID’s inbox. Be similar to allowing notifications as to whether the App sends the message.

1 Like

As it stands … but surely anyone could write to a private space they cannot read but the owner can, by signing with the public key.

Whatever works but the question is on the back of the option for it not being clear atm.

It simply needs a known location to send requests AND the owner being aware there is a request, which tempts a single location to look to that they have privileged access to, beyond what the public would do.

So, many to one write location, along with perhaps a proactive ping to the owner, if they are online - or expect they will check their mail when they are back online. Just that to avoid perpetual checking… services that are busy would be rewarded but services that are less busy will just spawn too many get requests, where something simple might be better for everyone.

Messaging seems the simplest system to me. One thing to check for the user.

Maybe a ping to the user’s client if the user is logged in somewhere. Guess we have to wait and see the messaging system that gets implemented.

2 Likes