Hello, Gene

I have question regarding device functionalities in Tinode. So I notice that both store.Devices.Delete() & store.Devices.GetAll() are currently not called anywhere on the main.

So where should we call these functions by design?

Thanks

Hi, after i made git pull when i am trying to sign up and i get the error:"Error: user not found or offline (404)", i didnt change any server code in go, also i reinstall database 2 times, so maybe i find a bug.
And in console i have the next msg:
2018/01/07 19:58:44 Server v0.13: pid=3596 started with processes: 4
2018/01/07 19:58:44 Using config from: '/home/bogdan/godev/src/github.com/tinode/chat/server/tinode.conf'
2018/01/07 19:58:44 Running as a standalone server.
2018/01/07 19:58:44 plugins: no active plugins found
2018/01/07 19:58:44 Serving static content from '/home/bogdan/tinode/example-react-js/' at '/x/'
2018/01/07 19:58:44 Listening for client HTTP connections on [:6060]
2018/01/07 20:04:47 Session.dispatch got '{"hi":{"id":"80026","ver":"0.13","ua":"TinodeWeb/0.13 (Linux x86_64) tinodejs/0.13"}}' from '[::1]:39220'
2018/01/07 20:04:47 Session.dispatch got '{"acc":{"id":"80027","user":"new","scheme":"basic","secret":"dGVzdDM0MjM0Mjp0ZXN0MzQyMzQy","login":true,"desc":{"public":{"fn":"test342342"}},"tags":["email:test342342"]}}' from '[::1]:39220'
2018/01/07 20:04:47 Session.dispatch got '{"sub":{"id":"80028","topic":"me","get":{"what":"sub desc"}}}' from '[::1]:39220'
2018/01/07 20:04:47 Sub to 'me' from 'usrsBPAj0q002k'
2018/01/07 20:04:47 hub: user's account unexpectedly not found (deleted?)
2018/01/07 20:05:49 sess.readLoop: websocket: close 1001 (going away)
2018/01/07 20:05:49 serveWebsocket - stop
2018/01/07 20:06:00 Session.dispatch got '{"hi":{"id":"116964","ver":"0.13","ua":"TinodeWeb/0.13 (Linux x86_64) tinodejs/0.13"}}' from '[::1]:39284'
2018/01/07 20:06:00 Session.dispatch got '{"login":{"id":"116965","scheme":"basic","secret":"dGVzdDp0ZXN0"}}' from '[::1]:39284'
2018/01/07 20:06:00 basic auth: invalid login
2018/01/07 20:06:26 sess.writeLoop: ping/websocket: close sent

Hello,
I get this error while I try to install from source:

# go get -tags rethinkdb github.com/tinode/chat/server && go install -tags rethinkdb g ithub.com/tinode/chat/server

Yields:

package context: unrecognized import path "context" (import path does not begin with hostname)

How can I fix it?

Hi, I just build from source and got server, when I run tinode-db tool, error says:

2018/03/17 20:23:13 Failed to init DB: store: attept to Open an unknown adapter 'rethinkdb'

rethinkdb I just installed from brew and using:

rethinkdb --bind all --daemon

to started it. I think tinode-db tool is using for init that database right?
But there was an error. I dont know why I am sure these are right:

• I installed rethinkdb success, and run it;
• I build server with both mysql and rethinkdb enable, and conf file using default rethinkdb.

Any suggestion about this?

I signed up from this, but nothing happened when I click "SIGN UP" button. The chrome development tools showed that no data had been sent.

Hello, Gene

I have question regarding {del what=topic} & {leave unsub=true}. Are these both functions basically done something equivalent?

I noticed that these both functions basically deleting user subscription. Maybe the only difference is when we use {leave} we must already subscribed to the topic while {del} is not. So is my understanding correct?

Thanks

Hi,
I've got the error below whenever I execute the command:

go install -tags rethinkdb github.com/tinode/chat/tinode-db
/go/src/github.com/tinode/chat/server/db/rethinkdb/adapter.go:16:2: cannot find package "gopkg.in/gorethink/gorethink.v2" in any of:
/usr/local/go/src/gopkg.in/gorethink/gorethink.v2 (from $GOROOT)
/go/src/gopkg.in/gorethink/gorethink.v2 (from $GOPATH)

Do you have some clue about what's generating that error?

Hello,
As users logs in, they encounter with a blank chat page and need to add user/groups, which is not very pleasant for a public chatroom were people don't necessary know who else are in the room.
I'm wondering how can change tinode to show the list of available users (or a public default group) to them, as they log in?

Thanks for your quick replies.

Hi, wondful project , any plan for iOS client ?

I found my Cloud Messaging API key at the following URL:

https://console.firebase.google.com/u/0/project/*project*/settings/cloudmessaging/

The current URL in the install document loads a blank firebase console.

Hi,gene sorry for disturbing,i download the latest react.js and tinode.js version 0.14 also has latest version on server, but i still get "Error: user not found or offline (404)".Maybe i am doing something wrong?Can you help me please?

Hi, Casbin is an authorization library that supports models like ACL, RBAC, ABAC.

Related to RBAC, Casbin has several advantages:

1. roles can be cascaded, aka roles can have roles.
2. support resource roles, so users have their roles and resource have their roles too. role = group here.
3. the permission assignments (or policy in Casbin's language) can be persisted in files or database (Gorm or Xorm or any popular DBs).

And you can even customize your own access control model, for example, mix RBAC and ABAC together by using roles and attributes at the same time. It's very flexible.

Casbin can provide more flexibility and security than the current level-based access control. Let me know if there's any question:) Thanks.

Hello, Gene

I would like to ask regarding setting private value on P2P permission. Why it is prohibited to change private value on P2P subscription? source code

Maybe there is another way if I want to set private value to P2P subscription?

Thanks

Hello, Gene

I have question regarding deleting message on Tinode. Is it possible to selectively delete message / delete only one specific message? Because currently the functionality which exist is behaving like clear messages, right?

Thanks

Hello,
I'd like to show a list of all available groups to the users who have just logged in. I appreciate if you elaborate how this can be achieved?

Thanks

Hi guys,this is the most awesome open-source product that i saw, but i found a bug, i cant create more then 3 tags on 'acc' it responds with error 500

Hello, Gene

I have question regarding current cluster design. I found that user’s topic always sticked to a certain cluster node. Why?

In the event of node failure this will lead to users (which their topics sticked to that node) won’t be able to access their data nor contacted by others, right?

Hello,

How trivial would it be to implement another database store. My use case would involve a lot of traffic, so being able to use a database like redis or cassandra would be quite beneficial.

How trivial would it be to implement redis support, while taking advantage of features like redis pub-sub, which are important for an application such as this? Is this something I can tackle and create a pull request for without refactoring the entire application? (Bear in mind, I haven't gone through the code indept) .

EDIT:
ON more thought. Having the messages stored in a Cassandra database, and the meta data stored in Postgres, is the ideal scenario for me. Would this be possible?

First time I launched the server it was configured to use :8080 port same as rethinkdb's admin interface, yet http.ListenAndServe did not produce any errors.

while i am writing command "go get -tags rethinkdb github.com/tinode/chat/server && go install -tags rethinkdb github.com/tinode/chat/server"
i get next error: src/github.com/tinode/chat/server/ringhash/ringhash.go:79: undefined: fnv.New128a

Hello, Gene

Currently I’m facing several problems when trying to implement following push message use cases in client app:

1. FCM token refresh in the middle of user session
2. User logout in app

So as far as I understand, it is possible for FCM token to be refreshed. When this event occurs, we need to update our existing token on the server.

Yet in Tinode, FCM token (called "deviceId" in Tinode) could only be updated in {hi}. So in case of token refresh occurs in the middle of user session, we need to break the session first to register new updated token.

So maybe there is a better way to do this which I'm missing?

The next issue occurs when single device is being used by multiple user. So for example User A & User B use single device together.

In Tinode, deviceId is associated with uid (link), yet store.Devices.Delete() is only called when deviceId is invalid (link).

Since User A & User B use the device together, then the deviceId would be same for them. So in case User A has been logout & User B logged in, then message for User A would be still delivered to User B.

So I think it's better if we also offers delete device functionality for user so the client could notify server which deviceId still used for certain user & which is not.

What do you think?

Also if I may suggest I think it's better if we differentiate between deviceId & fcm token though. The reason is because essentialy a device may have multiple tokens due to rotation / refresh.

Since basically store.Devices.Update() is defined as "upsert", then every unused fcm token would be just pile up on database. But if we differentiate between deviceId & fcm token then we could just update the associated token for the deviceId leaving more compact space on the database.

What do you think?

Thanks

Hi.
I am going to implement the react native realtime chatting mobile app and I am very interested with this backend.
It seems that it is available with react.js and I am not sure if I can use it for react native.
Could you tell me if I can use it for react native?
Best wishes and regards.

Hello,

How does to run server on cluster mode? I tried to execute 2 process of $GOPATH/bin/server which in first process use cluster-one.conf (listen to localhost:6060) and the second one use cluster-two.conf (listen to localhost:6062).

But after I accessing both of the web consoles (localhost:6060/x/example-react-js/ & localhost:6062/x/example-react-js/) I cannot login. I received error 500 instead.

So is this bug or the functionality is not yet done or I misconfigured both of my servers? I tried with devel branch but I also face the same issue.

Thanks

Hello, Gene

I have question regarding {pres} packet, I found that scenarios 3 & 7 doesn’t work. Are these features removed or maybe these are bugs?

Thanks

Hello, Gene

I have question regarding MsgDelRange. Why it is having exclusive end rather than inclusive one?

Maybe there is some design consideration behind this?

Thanks

hi,rethinkdb not support binary file over a MB. if a video 's size exceed. How to handle it?

Hi, i made a last build version 0.14 and now when i make login or signup i got next error: Error: command out of sequence (409).
Is this on my side only?Did you get this issue?

Hello, Gene

So if I would like to implement hard delete for P2P topic. Is it good idea to just include D permission on P2P default permissions?

Thanks

Hello,

While following the install guide, I had to manually install an additional package. I'm not sure if this is specific to my install, but I figured you might want to know so you can update your instructions.

root@instance-1:~# go get github.com/tinode/chat/server && go install -tags rethinkdb github.com/tinode/chat/server
go/src/github.com/tinode/chat/server/db/rethinkdb/adapter.go:16:2: cannot find package "gopkg.in/gorethink/gorethink.v3" in any of:
        /usr/local/go/src/gopkg.in/gorethink/gorethink.v3 (from $GOROOT)
        /root/go/src/gopkg.in/gorethink/gorethink.v3 (from $GOPATH)

go get gopkg.in/gorethink/gorethink.v3 fixes the error

Hello, Gene

I have question regarding hangup signal (syscall.SIGHUP) in Tinode. What are the possible causes for Tinode to receive hangup signal?

So I was trying to run Tinode cluster on EC2 instances behind load balancer. On each instances I was using nohup to run Tinode on the background. When I tried to use it, sometimes only one instances shutdown randomly, sometimes both. When I check the output on nohup.out I notice following consistent output: Signal received: 'hangup', shutting down.

Before I was running Tinode on cluster, I also tried to run Tinode on single instance & sometimes I get the same output. But in cluster setup, it is occurred more often. Especially when I tried to send large payload from another instance.

So what are the possible causes for this issue on Tinode?

Thanks

Hello, Gene

I have question regarding purpose of access control on P2P topic. So what is the purpose of it anyway?

The reason why I ask this question is because I found that the purpose stated in documentation is somewhat contradict with the implementation written on code. So with this question I would like to validate what I found because I think it is very fundamental.

In documentation (see here) I found that the purpose of it is for managing presence + banning user from initiating or continuing P2P conversation. So if there are user1 & user2 which have P2P connection, then after sometime user2 would like to block message from user1, then user2 would be just only need to remove “W” permission from user1.

But based on current implementation it is not possible to achieve that purpose. The reason is in current implementation it is not possible to change others permission on P2P topic, it is just for setting user own permission for filtering the data retrieved by user itself (which I personally think is somewhat useless on this context).

So in above case, user2 would only able to manage its own permission, so does with user1, so they cannot block each other in case they want to do it.

So which purpose is true? Managing the P2P connection or just for filtering data retrieved from the context? Or maybe you have another design decision?

Thanks a lot, Gene.

Firstly, I am very impressed by this system and I appreciate all the effort you put into it!

I'm trying to simply create a user through websockets. I'll list all the messages I send and receive.

[hi]

{
  "hi": {
    "ver": "0.14"
  }
}

[acc]

{
  "acc": {
    "scheme":"basic",
    "user": "new",
    "secret":"dGVzdDp1c2Vy",
    "tags": ["Test User"],
    "desc":{
      "public":{
        "fn":"Test User"
      }
    },
    "cred": [{"meth":"email","val":"[email protected]", "resp":"123456"
    }]
  }
}

However after doing this it seems that while the user is created, I simply get a 500 internal server error when trying to create a new p2p message subscription from a default user (alice) to the newly created test user.

{
   "sub": {
        "topic": "<created users id ie: usrhHVAzk9Owww>",
        "get": {
            "what": "desc sub data"
        }
    }
}

response:

ctrl: {topic: "usrhHVAzk9Owww", code: 500, text: "internal error", ts: "2018-04-16T05:57:23.462Z"}

I am using docker, and the MySQL extension.

If you need any more information let me know!

Hello, Gene

I have question regarding server license. If we run Tinode on our own servers to serve our public users, does this count as act of "distribution"?

So we would like to integrate Tinode with our own existing system. To do so we need to modify Tinode in the way that it would expose our internal architecture.

As far as I know in GPL we need also to make our modified source code public whenever the distribution of the software is made. Yet in our case it is impossible to do so after the modification is done.

So we would like to confirm first what is the meaning of "distribution" here for Tinode.

Thanks a lot, Gene

Hi,

The spec states that timestamps will be formatted according to RFC 3999. But it should say RFC 3339, as 3999 doesn't exist.

Hello, Gene

I’m facing a weird bug when trying to set private field of user’s fnd topic which apparently located on different cluster node. It always return following error message:

cluster: forwarding request to node 'two’
cluster: call failed to 'two' [gob: type not registered for interface: []interface {}]

The culprit is this line: https://github.com/tinode/chat/blob/devel/server/cluster.go#L135. But I don’t understand why it throws error while for forwarding normal message it works.

Maybe you have idea why this error happens, Gene?

Thanks

Do you plan to increase the possible group size, or is it going to stay at 16? If so, why?

Thanks

Hello,
I have followed the instructions but get blank page at http://127.0.0.1:6060/x/ with 404 page not found while the server is running:

me@pc:~/Desktop$ $GOPATH/bin/server -config=$GOPATH/src/github.com/tinode/chat/server/tinode.conf -static_data=$HOME/tinode/example-react-js/
2018/01/20 18:28:06 Server v0.14: pid=8559 started with processes: 8
2018/01/20 18:28:06 Using config from: '/home/me/go/src/github.com/tinode/chat/server/tinode.conf'
2018/01/20 18:28:06 Running as a standalone server.
2018/01/20 18:28:06 plugins: no active plugins found
2018/01/20 18:28:06 Serving static content from '/home/me/tinode/example-react-js/' at '/x/'
2018/01/20 18:28:06 Listening for client HTTP connections on [:6060]

tionde client file structure:

me@pc:~$ tree /home/me/tinode/
/home/me/tinode/
└── example-react-js-master
    ├── css
    │   └── base.css
    ├── drafty.js
    ├── img
    │   ├── ignasi_pattern_s.png
    │   ├── logo32x32a.png
    │   ├── logo32x32.png
    │   └── logo.svg
    ├── index.html
    ├── README.md
    ├── tinode.js
    ├── web-desktop-2.png
    ├── web-mob-chat-1.png
    ├── web-mob-contacts-1.png
    ├── web-mob-info-1.png
    └── web-mob-new-chat-1.png

How can I debug/fix it?

Hello,
I followed the instructions but could not see the web interface.
I run
/home/me/go/bin/server -config=/home/me/go/src/github.com/tinode/chat/server/tinode.conf -static_data=/home/me/Desktop/example-react-js/

But when I visit:

http://localhost:6060/home/me/Desktop/example-react-js/

I get:

{
ctrl: {
code: 404,
text: "not found",
ts: "2017-09-25T08:55:41.132Z"
}
}

I have extracted the react files and to example-react-js and copied tinode.js there.

The server output in the terminal:

me@pc:~/go$ /home/me/go/bin/server -config=/home/me/go/src/github.com/tinode/chat/server/tinode.conf -static_data=/home/me/Desktop/example-react-js
2017/09/25 10:55:22 Server v0.13: pid=24955 started with processes: 8
2017/09/25 10:55:22 Using config from: '/home/me/go/src/github.com/tinode/chat/server/tinode.conf'
2017/09/25 10:55:22 Running as a standalone server.
2017/09/25 10:55:22 Serving static content from '/home/me/Desktop/example-react-js'
2017/09/25 10:55:22 Listening for client HTTP connections on [:6060]

Any idea why this error happens?

As I further known that RethinkDB company were disbanded. Does RethinkDB is a long term choice to be used in such as a good library? (Just my confusion, not about library bugs or other issues).

Hello, Gene

I have question regarding {ctrl} packet. Can we treat this packet as end of response packet from server?

Thanks

Hi,

When I exec this command
sudo docker run --rm --name tinode-init --link rethinkdb tinode-init-db
face with this error
docker: Error response from daemon: Cannot link to a non running container: /rethinkdb AS /tinode-init/rethinkdb.

When I want to start tinode-init-db with follow command
sudo docker run tinode-init-db
face with this error

2018/02/11 12:30:21 Initializing DB...
2018/02/11 12:30:21 Failed to init DB: gorethink: dial tcp :0: getsockopt: connection refused

Please help me to solve this problem.
Thanks

Hello, Gene

I found that in current push implementation, if a user doesn’t subscribe to topic, even though he is online, he will still receive push notification when there is new message on the topic.

You could see the details here, which shown topic only exclude device ids which subscribed to current topic.

I think this behavior will annoys user though. Because even though user is already online in app he would still receive notification.

What do you think?

Thanks

Hi. I have a problem installing tinode server on windows 10 using docker.
On point 5 - $ docker run --rm --name tinode-init-db --network tinode-net tinode-init-db in console i've got this:

standard_init_linux.go:195: exec user process caused "no such file or directory"

On stackoverflow i don't found decision for solving this problem. Do you have any ideas?

I follow your docker installation instruction. When I build initializer image from Dockerfile provided， I get this error "error: RPC failed; curl 56 GnuTLS recv error (-54): Error in the pull function." in "RUN go get -tags rethinkdb github.com/tinode/chat/tinode-db && go install -tags rethinkdb github.com/tinode/chat/tinode-db". And is there any way to solve this? Thank you.

Hello, Gene

I have questions regarding hard delete message. So how does offline user supposed to know when the message(s) in certain topic has been hard deleted?

I thought that supposedly offline user receive me notification when message(s) is hard deleted (see here). But apparently it is not, because {pres} would be silently dropped (see here).

So how does client for offline user supposedly to know if it should update its messages cache for the topic?

Thanks

I'd like a simple way to sort my subscriptions (conversations) when first receiving the information on them. Currently there is only the updated timestamp which is only changed when the subscription is actually changed. The only way (I can think of) to get this information is to get sub info to get the last message, then grab that using the sub again for the topic with a single ID for the message:

sub: {
    topic: "usr3424rwefsdf",
    what: "get",
    get: {
        since: "<lastId>",
        limit: 1
    }
}

It would make sense to receive the last contact date in the info for the initial me subscription. I don't have much experience with Go but I'd like to contribute as much as possible. I am writing an angular4 implementation for your chat and I plan on open sourcing it once completed.

sub: {
                topic: 'me',
                get: {
                    what: "desc sub data info", // info would also include last contacted timestamp.
                }
            }

server not found after running in dockerized environment .

localhost:6060
gives : 404 not found

Hello, Gene

I have question regarding Tinode ACS.

So I'm still not fully understand why would we need both ModeWant & ModeGiven when defining user access to the topic. Shouldn’t be ModeGiven only is enough to define it?😅

I mean ModeWant is only useful when user want to mute the conversation, right? But for the most of the use cases, ModeGiven should be enough, right?

For example when group owner want to promote a member to become admin, it would give the member "AS" permissions. Naturally, user would become admin directly.

But since there is a user's ModeWant, user need to elevate itself to admin first before it able to become the admin (set its ModeWant to also include "AS"). So why must we take this two steps when just once should be enough?

Maybe there is a design consideration behind this implementation?

Thanks

Hi

When I install the tinode server whit docker, face with this error:

docker build --tag=tinode-init-db init-rethinkdb
unable to prepare context: path "init-rethinkdb" not found

Thanks.

Any plans to create a docker image that's in docker hub?