Current behaviour : you can enable / disable channel configuration

Suggestion : a developer could simply remove the config if it's desired to be removed

There are a couple of things that I think should be addressed:

1. The list of channels is hardcoded in the BP repository (https://github.com/botpress/botpress/blob/sm-extmsg-7-fix-hitlnex/src/bp/core/config/config-loader.ts#L176). This list should be provided by the Messaging process. Once the messaging process is registered (https://github.com/botpress/botpress/blob/sm-extmsg-7-fix-hitlnex/src/bp/orchestrator/web-worker.ts#L61), it could query messaging to get a list of available channels, then update the botpress configuration (adding new channels to the file).

2. Channel configuration schemas should be moved from the main repository (in channel-*/assets/config.schema.json and into this one.
   A. There should be an endpoint on messaging that can be called to obtain this schema (to use in code editor)
   B. There should also be an endpoint where you can get the default configuration from that schema (to generate the default file)

3. Messaging displays a lot of verbose and useless logs for the user.
   A. The list of channels can be shortened to a single line. Space is wasted to display them on separate lines
   B. Endpoints are all listed but usage is not clear, and even disabled channels are displayed
   C. Logs should fit nicely with logs of other processes

4. The process of enabling and adding channels is really not clear. You need to update the botpress.config.json file, then somehow go on the code editor to create a channel-name.json file, and copy the config from somewhere else.
   A. When you enable a channel in botpress.config.json, there should be some process which would automatically query messaging, get the default config for that channel, and create a channel-name.json file

5. Maybe we could simplify the channel creation process. We could either merge all channel-*.json files into a single channels.json file, or in a channels property in the bot.config.json file. We could even add an "environment" variable to ease development. What do you think? This would make it really easy to develop the bot once and move the archive between dev and prod

bot.config.json or channels.json

channels: {
  prod: {
    slack: {}
    telegram: {}
}

// or 
channels: [
  {
    name: 'slack',
    enabled: true,
    env: 'prod',
    authToken: ''
  },
  {
    name: 'slack',
    env: 'dev',
    enabled: false,
   }
]

6. Once a channel is enabled, messaging should make a basic test to let the user know if the channel is properly configured, and it should give meaningful information to the user on how to do that.
   A. When channel is enabled, display a log indicating so
   B. Perform a small test to see if the channel configuration is actually correct
   C. Either add a log or an endpoint that a user can query to get the complete URL required to configure the webhook on different services (ex: the full event / interactive URL required by slack)

7. I think we should add a "legacy" route handler on messaging, and redirect all requests done to channel-* to the messaging process (except for channel-web). Since we removed channels from the main BP repository, all the logic (including legacy logic) should be in messaging. This will make it much easier to update channels and fix issues by having all the logic at the same place.
   A. Proxy all routes channel-* to messaging process
   B. Print a warning to the user indicating that the old route is deprecated (/mod/channel-slack/events) and display the correct route

I got this error message when my server restarted while discussing with the bot on slack. This log is way too long (axios traces of like 300 lines). Please just print the error message, or pick useful stuff from the axios error and only display that

On a side note, shouldn't it write [Messaging] Channel-Slack instead of "launcher" ?

When messaging has crashed and will not start again, the error message displayed is a bit misleading.

What do you think writing "Messaging service unavailable" instead ?

Describe the bug
When configuring Vonage to use the testing API (sandbox), we can easily exceed the rate limit defined by the sandbox.

To Reproduce
Steps to reproduce the behavior:

1. Configure Vonage and make sure "useTestingApi" is set to true.
2. Send multiple messages back to the user in one go.
3. You will see a 429 too many requests error in the console.

Expected behavior
The messaging server should make sure that we respect the one-second rate limiter that Vonage's testing API enforces.

Screenshots

Environment (please complete the following information):

• OS: Linux (Arch)
• Browser Chrome
• Browser Version 93.0.4577.82
• Botpress Version 12.26.2
• NodeJS 12.13

Additional context
N/A

nvm. missed the logs line

There is web_user_map to map channel web's visitor ID to Messaging UID, and you use msg_kvs to store channel-teams user ID to Messaging UID.

Channels each have their unique ID to identify a visitor. Each entry in msg_users represent an association userID - botId, why don't you add the channel's internal ID to this table instead ?

We could also add a "metadata" field which could be used to store additional data if required.

I see absolutely no reason to have a table "msg_kvs" in this service.

When running Messaging in standalone, we should display a warning if EXTERNAL_URL is undefined.

Can anyone update https://github.com/botpress/messaging for dummies like this tutorial: https://github.com/botpress/botpress/discussions/12393 I can add to that page how to install the messages module and activate etc.
I have also read https://v12.botpress.com/messaging-channels/direct-integrations/facebook-messenger it does not tell you how to install the module in a way i can understand.

I just realized that there's no trace of the messaging client in the diagnostic report, this might be useful to investigate weird issues or race conditions. We just had an issue where there were some "rogue processes" and this could have helped.

Also, can you add a name to the redis connection, with a similar naming than the other processes? The master process of the orchestrator generates an ID (process.env.SERVER_ID) which can help us associate processes together.

Having those information help us pinpoint the problem more easily

I know that repositories should have more autonomy, but they should still be able to be built together with all other repositories. I don't think we have a lot of clients who are able to install multiple versions of nodejs on their development computer just because we have different requirements.

I am simply unable to build everything without messing around with node versions (.nvmrc is not supported on windows). Therefore, It makes more sense to stick to the same node version for all related repositories.

Either you find a way to scale down the version requirement of messaging to the version we use everywhere else (including the builtin server in the binary), or you find a way to update the version in each repositories at the same time (including related tests).

At first glance, the module eslint-plugins-jsdoc version 36 requires a higher node version, is there any reason for that ? Can't we simply use the same version we have everywhere else and skip this requirement?

Apart from that, there is also an higher requirement for "parcel". What is this package used for? If it's for frontend, why not using webpack like every other repositories we have ?

Having independence over the codebase doesn't mean that you have to reinvent the wheel and to make things incompatible between other codebases, we should be able to work seamless between workspaces

WhatsApp CloudAPI now is publicly available.
https://developers.facebook.com/blog/post/2022/05/19/whatsapp-cloud-based-api/

using the official meta service can reduce some third party costs..

I already had a slack bot configured with the old URL. I tried to update the domain (new one for ngrok), but I forgot to update the second bot ID in the URL. Maybe needs an additional check to handle that gracefully?

https://4b1799194b02.ngrok.io/api/v1/bots/welcome-bot/mod/channel-slack/bots/main/events-callback

(theres "welcome-bot" and "main" there). The error could be more user-friendly:

When playing around with different configurations, I used to change the name of the channel so it wouldn't be loaded by messaging. Looks like some additional validation was added and it prevents the bot to load... do you think we can remove it ? I don't see any downside by having that, and it gives more flexibility to play around with different configurations. If messaging doesn't support an option, it could just ignore it instead of throwing an error

When you don't follow the "happy path", it's incredibly hard to get messaging to work.

Creating a bot in one instance and move it elsewhere

1. Spin a standalone messaging server.
2. Spin a botpress server and set MESSAGING_ENDPOINT to that server
3. Create a bot
4. Export the bot
5. Kill the server
6. Spin a different server, with the same messaging endpoint.
7. Import the same bot there

It should work correctly. Right now, it fails because the bot ID must be unique, but it shouldn't be the case in the future.

Reset messaging credentials

If you create a bot on a server, then remove messaging.id / messaging.token, and restart the server, it will receive a new id / token combination, but it doesn't seems to update the messaging database, because the bot doesn't work anymore.

Deleting and importing the same bot

If you export a bot, delete it, clear the id/token and re-import the bot with the same name, it will fail

When using an external messaging service (not spinned by botpress), there seems to be an additional configuration which is not specified.

Using the built-in messaging service (works)

1. Start BP binary
2. Create a bot
3. Add channel telegram to the bot.config.json file
4. Restart server
5. Speak on telegram, it works

Using a different messaging endpoint (doesn't work)

1. Start messaging binary (clean install)
2. Set MESSAGING_ENDPOINT=http://localhost:3100
3. Start BP binary (clean install)
4. Create a bot
5. Add channel telegram
6. Restart server
7. Getting errors on messaging: No address associated with hostname / 401 when calling BP

When a bot is configured with teams with a non-working appID it crashed the messaging app and no bot can be talked to unless Botpress is restarted.

I fixed this by removing the config in the affected bot and an old global channel-teams.json file that was present on the server containing this appid.

Here's the error in the logs:

This is happening using 12.26.10

The migration seems a bit flaky, we can expect invalid entries in the database, and it should be able to handle it.
I can provide my core.sqlite file if you want to try it out

Describe the bug
Carousel Cards are sent in reverse order to WhatsApp (through Smooch)

Environment (please complete the following information):

• 12.26.4

Closes DEV-1889

When a user clicks a button in a single choice, the event.type sent to Botpress is 'text'.
This should be of type 'quick_reply'.
'text' should only be sent if the user wrote something instead of picking an option.

Sometimes I get this weird error message. I think telegram sends events (maybe some kind of update) which doesn't have a userId and displays that. No idea how to reproduce, I just have my messaging server running and rebooting stuff / playing around

Not sure how to reproduce, but a missing conversationId causes the whole server to restart

Okay, so hmm not sure how I managed to get there the first time, but I was able to reproduce easily:

1. Create a bot from "welcome-bot", open the studio and write something to the bot (keep the tab open)
2. Open a new tab, go on the admin panel
3. Delete the first bot you created
4. Create a new bot with a different name
5. Open the studio, and write something to the bot

Basically, I ended up with those errors, and an unresponsive server which needed a reboot

And it ends up by

Not sure of the root cause (bot id validation or else).

Starting up my server with invalid config causes BP to spam this error message every 15 seconds

Currently: When we capture a user's message in the webhook we receive something like the following

{
 channelId: 'asdasdas',
 clientId: 'asdasd',
 /*....*/
}

Then if you want to reply to that message you need to specify the channel and no where it says that we should use the channel name and not the id which can become confusing. It's also super tempting to use data.channel.

From what I understand, channel-id is for internal usage so it shouldn't be exposed to the user

v12.26.1

I was trying to use the voice composer after activating the google speech module, and I face the below issue in the logs:

09/13/2021 23:03:35.996 Mod[channel-web] [test]
HTTP (post) URL /media
Received "This type of file is not allowed: audio/webm"
Request failed with status code 400

The discord channel doesn't seem to be working properly.

The problems (I think) are:

• discord.js is outdated
• in server/src/channels/discord/conduit.ts, at line 21, you are not specifying the bot's intents, so discord isn't sending the bot any messages. Try using this.client = new Discord.Client({ intents: [Discord.Intents.FLAGS.GUILDS, Discord.Intents.FLAGS.GUILD_MESSAGES] }) after updating discord.js
• I'm not entirely certain, but I think that discord-buttons is wholly unnecessary and all the buttons can just be used via discord.js

I tried fixing this myself, but my lack of skills in node.js prevented me from doing so. If you want to see what I attempted to do, I will gladly share. The main problem I encountered was that the updated discord.js presumably uses modern javascript syntax (such as ?? and ?.) and the javascript compiler in the botpress standalone server didn't seem to like that.

I'd appreciate it if someone with more brains than I could fix this.

Messenger was working fine but stopped working after the migration

So hm, if you simply change the messaging endpoint and restart the server, when you open the webchat there's a ton of endless queries. This definitely would need some sort of protection to prevent a user to ddos the service

1. Download botpress v12.26.1
2. Create a bot (welcome-bot)
3. Download messaging v0.1.13 from the official release
4. Start the process
5. Set MESSAGING_ENDPOINT and restart botpress. This happens:

If you create a bot with a remote endpoint and then remove it, it does the same thing locally

I got this error message and had to spend some time to find the issue.

Somehow I had 2 different configurations:

When validating the schema using Joi, please consider setting the option "ignore unknown properties" instead of rejecting the payload completely. The correct fields were also present, so the bot would have worked properly.

I have disabled 3 bots, however you can still see the webhook for hotel-template-v1

Teams is also disabled for welcome-bot, but it looks like its enabled on messaging:

To use messaging and slack, I have an hostname which redirect to my bp instance, but messaging refuses to connect because it's a self-signed certificate. Is there a way around?

Yes, I could use ngrok... but no, I don't want to have to reconfigure my channels everytime I want to test the bot

I was trying to setup my slack channel, and I doubt everything I provided was correct, but it shouldnt crash messaging.

There is a lot of new tables.
Can you make a database schema explaining the link between them, their purpose, and the source of the information? This schema could grow over time and would be very useful to understand the flow of information for new users (ex: me)

Ex:

When I start 2 servers and I speak with the bot on slack, I receive no answer and I get this stack trace on the first server I started (plus, messaging crashes and restarts).

Also, I think this log should be a debug log instead, since it has no use for the end user

I have configured a new app in the facebook developers account and it forces me to use v15.0 API. But botpress is using v12.6. How can I upgrade my code to use the v15.0. While using the v12.6 graph API it gives the error 400 Bad request:

{
    "error": {
        "message": "Unknown path components: /me/messages",
        "type": "OAuthException",
        "code": 2500,
        "fbtrace_id": "Ay0lBm-yBJucULHbD0R5QdZ"
    }
}

At first I had a missing field error which crashed the whole app

After adding the field, it crashed again for another reason...

I had to manually go delete the entry from the database... Please make this more reliable and error-resistant. Such errors can happen and should NOT crash the whole server

Botpress core version: 12.26.3

All though titles are not required for other channels when using the image content-type, if you don`t provide, it will not work in the Slack Channel and an error will be thrown:

[ERROR] web-api:WebClient:0 failed to match all allowed schemas [json-pointer:/blocks/0/title]
[ERROR] web-api:WebClient:0 missing required field: text [json-pointer:/blocks/0/title]
09/28/2021 14:20:49.466 [Messaging] app:slack Error occurred when sending a message. Error: An API error occurred: invalid_blocks
at Object.platformErrorFromResult (C:\snapshot\messaging\node_modules@slack\web-api\dist\errors.js:51:33)
at WebClient.apiCall (C:\snapshot\messaging\node_modules@slack\web-api\dist\WebClient.js:156:28)
at processTicksAndRejections (internal/process/task_queues.js:97:5)
at async SlackCommonSender.send (C:\snapshot\messaging\packages\server\dist\channels\slack\senders\common.js)
at async SlackConduit.sendToEndpoint (C:\snapshot\messaging\packages\server\dist\channels\base\conduit.js)
at async InstanceService.sendMessageToInstance (C:\snapshot\messaging\packages\server\dist\instances\service.js)
{

to fix this you need to specify a title:

Since titles are not required by the content type, I believe the renderer should auto-generate one.

Hello,

I can see checked checkbox for video type, file type and location type.
on https://github.com/botpress/messaging/tree/master/packages/channels/src/telegram

But, I'm unable to receive any of these types on telegram.
It is working as expected on web channel.
Does it need extra configuration or is it a bug?

We have an issue when running E2E and security tests with the latest version of ts-jest (27.1.3 - kulshekhar/ts-jest@v27.1.2...v27.1.3)

See this run for the error message.

Follow up on botpress/v12#370
As this now has its own project, I would like to request support for the creation of a Mattermost plugin.

I understand this might be low priority for the Botpress Team, but maybe you could reference documentation or a tutorial to support a contribution for Mattermost and similar messaging vendors?

Hello everyone,

I'm trying to start developing a custom channel based on the custom-channel repo and from what I can understand I need to get this working locally so I can connect my custom channel and start building, but I'm having some issues making it work.

First, I don't not if it's an issue of the documentation but I couldn't start the services using Tilt the way it's documented - had to run a yarn install first (which sounds logic but maybe the step should be explicitly mentioned?)

This yarn install gives me some warnings about dependencies not satisfied, but after it Tilt can start the resources successfully:

❯ yarn install
➤ YN0000: ┌ Resolution step
➤ YN0002: │ @botpress/messaging-board@workspace:packages/board doesn't provide @parcel/core (p8ec10), requested by @parcel/config-default
➤ YN0002: │ @botpress/messaging-components@workspace:packages/components doesn't provide babel-loader (p6cd3a), requested by @storybook/addon-essentials
➤ YN0002: │ @botpress/messaging-components@workspace:packages/components [d721a] doesn't provide babel-loader (p7a0bc), requested by @storybook/addon-essentials
➤ YN0002: │ @botpress/messaging-server@workspace:packages/server doesn't provide @opentelemetry/api (pc6b23), requested by @opentelemetry/instrumentation-aws-sdk
➤ YN0002: │ @botpress/messaging-server@workspace:packages/server doesn't provide @opentelemetry/api (p07603), requested by @opentelemetry/instrumentation-express
➤ YN0002: │ @botpress/messaging-server@workspace:packages/server doesn't provide @opentelemetry/api (p1080b), requested by @opentelemetry/instrumentation-http
➤ YN0002: │ @botpress/messaging-server@workspace:packages/server doesn't provide @opentelemetry/api (p6bb08), requested by @opentelemetry/instrumentation-ioredis
➤ YN0002: │ @botpress/messaging-server@workspace:packages/server doesn't provide @opentelemetry/api (p6bef0), requested by @opentelemetry/instrumentation-knex
➤ YN0002: │ @botpress/messaging-server@workspace:packages/server doesn't provide @opentelemetry/api (p2e0b5), requested by @opentelemetry/instrumentation-pg
➤ YN0002: │ @botpress/messaging-server@workspace:packages/server doesn't provide prom-client (p34f6f), requested by @promster/types
➤ YN0002: │ @botpress/webchat-inject@workspace:packages/inject doesn't provide @parcel/core (p5db79), requested by @parcel/config-default
➤ YN0002: │ @botpress/webchat-inject@workspace:packages/inject doesn't provide typescript (pb00c9), requested by @parcel/transformer-typescript-tsc
➤ YN0060: │ @botpress/webchat@workspace:packages/webchat provides mime (p5bc44) with version 3.0.0, which doesn't satisfy what @botpress/messaging-components requests
➤ YN0060: │ @botpress/webchat@workspace:packages/webchat provides react (p33191) with version 17.0.2, which doesn't satisfy what @botpress/messaging-components and some of its descendants request
➤ YN0060: │ @botpress/webchat@workspace:packages/webchat provides react (p4b3ac) with version 17.0.2, which doesn't satisfy what mobx-react and some of its descendants request
➤ YN0060: │ @botpress/webchat@workspace:packages/webchat provides react (p61e8a) with version 17.0.2, which doesn't satisfy what react-ga requests
➤ YN0060: │ @botpress/webchat@workspace:packages/webchat provides react (p18844) with version 17.0.2, which doesn't satisfy what react-intl requests
➤ YN0060: │ @botpress/webchat@workspace:packages/webchat provides react-dom (p50820) with version 17.0.2, which doesn't satisfy what @botpress/messaging-components and some of its descendants request
➤ YN0060: │ @bpinternal/trail@npm:0.1.0 provides @opentelemetry/api (pa21f3) with version 1.1.0, which doesn't satisfy what @opentelemetry/auto-instrumentations-node and some of its descendants request
➤ YN0002: │ @mdx-js/loader@npm:1.6.22 doesn't provide react (p12550), requested by @mdx-js/react
➤ YN0002: │ @parcel/optimizer-image@npm:2.2.1 doesn't provide @parcel/core (p48c9f), requested by @parcel/workers
➤ YN0002: │ @parcel/reporter-dev-server@npm:2.2.1 doesn't provide @types/express (p57ed2), requested by http-proxy-middleware
➤ YN0002: │ @parcel/transformer-image@npm:2.2.1 doesn't provide @parcel/core (p7bfaa), requested by @parcel/workers
➤ YN0002: │ @parcel/transformer-js@npm:2.2.1 doesn't provide @parcel/core (pe40b6), requested by @parcel/workers
➤ YN0002: │ @parcel/transformer-postcss@npm:2.2.1 doesn't provide postcss (p0b5da), requested by postcss-modules
➤ YN0002: │ @parcel/types@npm:2.2.1 doesn't provide @parcel/core (pfd728), requested by @parcel/fs
➤ YN0002: │ @parcel/types@npm:2.2.1 doesn't provide @parcel/core (p9b3cc), requested by @parcel/workers
➤ YN0002: │ @parcel/types@npm:2.2.1 doesn't provide @parcel/core (pe2210), requested by @parcel/cache
➤ YN0002: │ @parcel/types@npm:2.2.1 doesn't provide @parcel/core (p7a308), requested by @parcel/package-manager
➤ YN0002: │ @promster/express@npm:7.0.5 doesn't provide prom-client (pf18c2), requested by @promster/metrics
➤ YN0002: │ @promster/metrics@npm:9.1.5 [45095] doesn't provide typescript (pcf955), requested by ts-essentials
➤ YN0002: │ @promster/metrics@npm:9.1.5 [471c6] doesn't provide typescript (p7c2a9), requested by ts-essentials
➤ YN0002: │ @promster/server@npm:7.0.7 doesn't provide prom-client (p399ec), requested by @promster/metrics
➤ YN0002: │ @vonage/server-sdk@npm:2.10.7-beta-2 doesn't provide @babel/core (p3f8da), requested by @babel/eslint-parser
➤ YN0002: │ @vonage/server-sdk@npm:2.10.7-beta-2 doesn't provide eslint (pf6ac9), requested by @babel/eslint-parser
➤ YN0002: │ @vonage/server-sdk@npm:2.10.7-beta-2 doesn't provide eslint (pe1fde), requested by @babel/eslint-plugin
➤ YN0002: │ sunshine-conversations-client@npm:9.5.9 doesn't provide @babel/core (pf0dc1), requested by @babel/cli
➤ YN0000: │ Some peer dependencies are incorrectly met; run yarn explain peer-requirements <hash> for details, where <hash> is the six-letter p-prefixed code
➤ YN0000: └ Completed in 0s 649ms
➤ YN0000: ┌ Fetch step
➤ YN0000: └ Completed in 1s 381ms
➤ YN0000: ┌ Link step
➤ YN0005: │ core-js@npm:3.20.0 lists build scripts, but its build has been explicitly disabled through configuration.
➤ YN0005: │ core-js-pure@npm:3.20.0 lists build scripts, but its build has been explicitly disabled through configuration.
➤ YN0000: └ Completed in 1s 848ms
➤ YN0000: Done with warnings in 4s 144ms

Once I have those resources started and in "ready" status if I open the messaging resource details I can see in the logs that the server is listening:

12/01/2022 14:01:00.058 Launcher ===========================================================================
                                                             Botpress Messaging                            
                                                                Version 1.2.7                               
                                 ===========================================================================
12/01/2022 14:01:00.191 Admin ===========================================================================
                                                         ADMIN_KEY IS NOT SET                           
                                                     ADMIN ROUTES ARE UNPROTECTED                       
                              ===========================================================================
12/01/2022 14:01:00.220 Launcher Server is listening at: http://localhost:3100/
12/01/2022 14:01:00.220 Launcher No external URL configured. Messaging Server might not behave as expected. Set the value for 'EXTERNAL_URL' to suppress this warning

But if I go to http://localhost:3100 I get this:

Also, if I try to open the webchat (which is located at http://127.0.0.1:3543/) I only see the "Example website
This website integrates the botpress webchat!" text without any webchat, and these errors on the console:

Live reload enabled.

127.0.0.1/:1 Uncaught SyntaxError: Unexpected token 'c', "connected" is not valid JSON
    at JSON.parse (<anonymous>)
    at ws.onmessage (runtime-79e8b212159ad102.js:109:12)

runtime-ea4cd6bdd08c17f2.js:95 WebSocket connection to 'ws://localhost:8080/' failed: 
goWOo @ runtime-ea4cd6bdd08c17f2.js:95

useObserver.ts:59 [MobX] You haven't configured observer batching which might result in unexpected behavior in some cases. See more at https://github.com/mobxjs/mobx-react-lite/#observer-batching
useObserver @ useObserver.ts:59

runtime-ea4cd6bdd08c17f2.js:95 WebSocket connection to 'ws://localhost:8080/' failed: 
goWOo @ runtime-ea4cd6bdd08c17f2.js:95

runtime-ea4cd6bdd08c17f2.js:171 undefined
ws.onerror @ runtime-ea4cd6bdd08c17f2.js:171

runtime-ea4cd6bdd08c17f2.js:175 [parcel] 🚨 Connection to the HMR server was lost
ws.onclose @ runtime-ea4cd6bdd08c17f2.js:175

index.4fdcc17c.js:12947 Uncaught (in promise) Client not found

The same thing happens if I drop Tilt and try to start it using yarn dev.

Is there anything that I'm missing? If so please let me know, I'm eager to make this work and hopefully contribute to the project in the future since we may use it intensely at work.

Thanks,

Hi, i am trying to build the messaging server binaries, to test with the botpress docker image.

1. Is there any way to build the messaging server (and tools) binaries in the botpress building process when is present in the workspace?
2. Can you post the steps and intructions to build the messaging server binaries?
3. Can botpress use a messaging server in other container, or as remote service API as you have done with language server, NLU and duckling?

Thanks in advance.

I set up all channels on the previous version, then copied my data folder over to master, and started it. I had those issues. I couldn't test smooch because I don't have an account, but I hope it will be tested

1. Telegram was working fine in the previous version, but it doesn't work anymore after the migration. I don't use a standard port, but it should still work fine, can you use the same version/setup that was used by channel-telegram so the migration is seamless?
   

2. Smooch fails on the first startup after the migration and requires to be rebooted again to start working
   

3. Migrations are executed after modules are loaded, which generates a ton of error messages... Is there a better way we could handle this ?
   

There is already an issue opened for the database schema, but can you make a schema like the event engine, explaining the layers and paths that a message will follow through messaging? Something like the event engine (https://botpress.com/docs/main/overview) would be amazing.

You can use draw.io to make such a schema. For example, here's the xml schema for the event engine:

Event Enginev2.2.zip

On Botpress you can configure the base URL to something different than / and it looks like messaging is not taking this into account.
There are often users who changes that to /botpress, so it changes the base url of the whole server, so it would be http://localhost:3000/botpress/api/v1/chat/receive

WhatsApp now supports interactive messages (buttons)!

This should be supported rather than resorting to the old way of sending the options in as a text message.

https://developers.facebook.com/docs/whatsapp/guides/interactive-messages

Migration went fine while using the old URL, but as soon as I use the URL that messaging recommends me, Teams doesn't work anymore.

I changed it from /api/v1/bots/welcome-bot/mod/channel-teams/api/messages to /api/v1/messaging/webhooks/welcome-bot/teams then I started getting these error messages:

When I speak to my bot, I get this error message continuously. I have no idea how to debug that, I tried to add LOGGING_ENABLED: true in messaging-server, but it didn't help.

Previously I would just add some logging in messaging source code, build it, then set DEV_MESSAGING_PATH, but it doesn't work since the migration to node 16 (which supports natively optional chaining, while 12 doesnT). How do you develop and debug currently?

Twilio was working fine before the migration, but now it says that the token is invalid, but the configuration in the bot config looks fine, so maybe it's a setup issue