Posts made by girish

@fbartels Just to throw some obvious ideas:

• Does it hang in other browsers? And also check with another device and another network.
• I would then try to see if curl http://container_ip:xxxx/wp-login.php is hanging on the server. The actual container_ip:port, you have to take from the nginx file in /etc/nginx/applications/appid.conf (there will be a proxy_pass).

If all the above still hang, I guess we can then assume it's something inside WordPress.

For inside WordPress:

• Disable LDAP plugin and others one by one and see if it has any effect. I suspect some plugin is slowing things down, it's the most common issue.

User teams are not propagated. I have reported this upstream - https://github.com/espocrm/espocrm/issues/2120

[1.11.1]

• Update metabase to 0.41.1
• Full changelog
• Better approach for column ordering in exports (#18450)
• Tools for fixing errors problems with Postgres semantics of limits (blank display of error table) (#18432)
• Raise MB_DB_CONNECTION_TIMEOUT_MS to 10000 as default (#18354)
• Allow caching of fonts and images (#18239)

@marcusquinn Also, the fallback email is in the Profile view of the user . But of course, it's impossible for a normal user or even admin to know all what I said....

@marcusquinn I think this is a bug in our flow, will get this fixed.

Looking into the flow a bit:

• When a user is created (by the admin), it just says "Email". What we do is to set the primary and the fallback email to this email address.

• Then, when a user signs up, he is not asked for any email. He has to know to go to Profile page and change it. Not ideal.

I think we should fix the flow to ask the admin for both the emails. And also warning them if fallback email is hosted on Cloudron. If fallback email is left blank, we should ask the user when they sign up.

@marcusquinn the are two email address on Cloudron - the primary email (exposed to apps) and the fallback email (sent for password reset). The fallback email is not supposed to be hosted on Cloudron. This is a bit hard to "impose" because at any time that fallback email address can become hosted on Cloudron... What can we do to prevent this?

@jfrey I have updated it all the way to the latest release now. Can you check?

[2.11.0]

• Update Taiga to 6.4.2

[2.9.0]

• Update Taiga to 6.2.2

[2.8.0]

• Update Taiga to 6.1.1

@loudlemur as @jdaviescoates pointed out, the closest we have is the WIP tags but even those get a bit out of date.

We are a small team, so our priorities unfortunately keep shifting to the next "fire". In general, cloudron.io issues, support tickets (both via mail and on this forum), security updates, app updates get higher priority than any new development.

@jfrey didn't realize the repo has moved. Looks like the package got a bit out of date because we didn't get notification. Looking into updating it.

[2.15.0]

• Update Confluence to 7.14.0
• Release notes

It's updated now.

[2.14.0]

• Update Confluence to 7.13.1
• Release notes

[2.13.0]

• Update Confluence to 7.12.5
• Release notes

You should get something like this:

$ curl 'https://my.domain.com/api/v1/groups?access_token=xxx'
{
  "groups": [
    {
      "id": "gid-02fc23f0-c0df-4276-a54a-e3958b53ffd8",
      "name": "people",
      "source": "",
      "userIds": [
        "uid-0a7dc777-c010-459e-a8db-ab853ea4f742"
      ]
    }
  ]
}

[1.5.0]

• Update Trilium to 0.48.2
• Full changelog
• Set default icon for clipped notes by @sigaloid in https://github.com/zadam/trilium/pull/2246
• hide tooltip after clicking on a button, fixes #2228
• fix backend API's getNote, getBranch, getAttribute, closes #2230
• fix restoring note revision, closes #2232
• allow async scripts in runOnBackend(), fixes #2243

@sam_uk I guess you are looking for the API call to add a user to a group ?

So, you make a PUT request on /api/v1/users/:userId/groups with groupIds in the body. groupIds is an array of group ids.

Group ids are not the same as group name. You have to get the ids via the GET /api/v1/groups call. This is because group names can be changed.

@Sam_uk @jdaviescoates that's a good idea, indeed! Can you make a feature request out this? It's quite easy to implement.

@teamcrw Do you think you can try with a new install? Maybe there is some update issue, I can check after you can verify the below.

This is what I did

• Installed a new package
• Setup user and checked the Diag view.

• Then, I go to File Manager and edit the /app/data/php.ini like below:

; Add custom PHP configuration in this file
; Settings here are merged with the package's built-in php.ini


post_max_size = 100000M
upload_max_size = 100000M
upload_max_filesize = 100000M

• Restart the app

• Check the diag view again:
  

Possibly final update: the manual tests went fine. We are just getting the e2e tests to pass and then we are good.

@chymian-0 oh I didn't see your screenshot at all, missed it completely. That screenshot looks like rainloop and not roundcube.

@potemkin_ai ah, i see. Do you think the defaults you have provided are reasonable? If so, I can put them as part of the default install.

@loudlemur the app is pre-setup with a username/password for admin. The default username is admin and password is changeme

@loudlemur On Cloudron, apps are run in containers. The paths which are mentioned in app docs and messages are inside the container. They are not easily found on the server (and not meant to be directly edited by ssh'ing).

/app/data is the path in the container where apps write to. The easiest way to access this is using the File manager. The file manager exposes only the /app/data directory of the container and nothing else.

If you want to see the full container file system, you have to use the Web Terminal. You can always cd into /app/data there as well. There you will see the same files as the file manager.

https://github.com/roundcube/roundcubemail/issues/7556 is the issue to make the scope configurable

@chymian-0 Are you referring to the search options ? If you click the "arrow" on the right of the search text box, it will drop down with options. One inconvenient thing is that Scope has to be changed manually each time to All folders. IIRC, there was a task for this in roundcube somewhere.

We are still working on the EspoCRM 7 update. We can look into making Cloudron authentication optional afterwards.

@teamcrw there was a bug in the package. I have pushed an update now. Can you update and check?

[2.6.1]

• Update Lychee to 4.3.4
• Full changelog

@garyhost Did you manage to figure it out?

https://www.cloudron.io/console.html#/userprofile?view=referrals has your referral link

@molinaire the IP address there resolves to 86.180.81.68 . Is that the IP address of your server? The server seems unreachable, I cannot ping it.

@brutalbirdie I have investigated this in the past. If we disable open registration, then you cannot register the initial admin user. So, it's a catch-22. See https://git.cloudron.io/cloudron/rocketchat-app/-/blob/master/start.sh#L22 .

In effect, even when you pre-setup LDAP, opening rocket.chat will show the initial Rocket.Chat wizard . But you cannot complete the wizard because open registration is disabled. The only thing we can do is put a message in post install message.

I guess this has to be reported upstream, not sure? In the past, there used to be a way to add an admin user programmatically but now we have a wizard (which I guess they want to collect info on the setup).

Not yet, we have to start looking into apps next week. This week we are working on the Cloudron release (almost there!).

@molinaire said in Apps disapeared:

I get this error message when accessing the IP:

You cannot access the dashboard by IP after installing apps. Can you try accessing via https://my.domain.com ?

@sam_uk What does export mean in this context? You mean like some json file (what's inside the json?) ?

@potemkin_ai said in Automated server reboot:

-> $TOKEN==
<** 535 5.7.8 Authentication failed
-> AUTH PLAIN AGFsZXhAcG90ZW1raW4uY28A$TOKEN==
<** 535 5.7.8 Authentication failed
*** No authentication type succeeded

Something wrong in the script. It's sending $TOKEN

@khadanja probably something wrong on our side, can you drop us a mail at support@cloudron.io with your cloudron.io email and we can get this sorted out?

@bbm_noahgilbert on it!

nextcloud has a weather app. Always makes me smile when the file sharing app shows the weather

You can use swaks like this:

First get the IP of the mail container:

# docker inspect --format '{{ .NetworkSettings.Networks.cloudron.IPAddress }}' mail
172.18.0.2

Then, you need the "token" which is the auth to send mails:

# docker inspect --format '{{ .Config.Env }}' mail | tr ' ' '\n' | grep CLOUDRON_RELAY_TOKEN | sed 's/^.*=//'
9e000daa7ba34f29ed975d54b880f30794c2cebcd795f124

Now, you can send using swaks (replace IP and token below):

# swaks -s 172.18.0.2 -p 2525 --au no-reply@cloudron.space --ap 9e000daa7ba34f29ed975d54b880f30794c2cebcd795f1241 -f 'no-reply@cloudron.space' -t 'test@cloudron.io' --h-Subject "Test mail"

I think you can use swaks to send mail from Cloudron. Let me dig up the exact way to use it.

@timconsidine Don't know if it matters, but iirc mattermost does not have access control in the community edition that we have on Cloudron. As in, all the users are "equal" and any user can archive channels etc. There is no fine grained permissions.

@loudlemur said in Canvas LMS - modern, open-source LMS:

But Sakai seems cool too - we should add an App Wishlist post for it!

I could create one, but our utterly glorious developers here must be feverishly busy preparing other applications for us. Would another wish be a burden or distraction?

Not at all, please do create threads. Gives an opportunity to collect feedback and votes.

[1.54.1]

• Update Mattermost to 6.0.1

@loudlemur yeah, some apps like Ghost/LimeSurvey/osTicket don't make it immediately obviously where an admin user has to login.

@potemkin_ai yes, there is no use case to deploy 100s of cloudrons. For the big providers, we have 1-click images but those don't require automation of the configuration.

Thanks for clarifying.

While it's easy to implement, have to think more about this one. After all, no VPS provider I know has this functionality. Maybe an idea is to help the sysadmin somehow automate this via API call (say via n8n).

[1.7.0]

• Update Uptime Kuma to 1.9.1
• Full changelog
• Add custom HTTP methods, body and headers (Thanks @bertyhell)
• Add an ability to clear old monitor data automatically (Thanks @chakflying)
• Add Steam Game Server Monitor Type (Thanks @Revyn112)
• [SMTP] Add custom Email subject/title (Thanks @NixNotCastey)
• Add PM2 config (Thanks @GhostSlayer)
• Notifications: Feishu (Thanks @xjoker), Aliyun SMS (Thanks @xjoker), DingDing (Thanks @xjoker)

[3.188.1]

• Update Ghost to 4.19.1
• Full changelog
• Fixed handling of "incomplete" subscriptions - Fabien O'Carroll
• Fixed "Send" publish type not being reset when closing publish menu - Kevin Ansfield

@robi We also tinkered with automating it in cloudron.io, but that requires the user to give us ssh access to the server (atleast, for the duration of deployment) to setup stuff.

Would that be interesting? Login to Cloudron.io, add an IP address and some deployment parameters and it will go deploy. Underneath, it's probably only web-ifying the deploy script.

I can't speak for Element since I use it sparingly. We have been using Rocket.Chat for 3-4 years now and like it. We mostly only use it for text messaging. When we started out, we also used it for Live Chat on the website, though we have removed now. The mobile app can be flaky wrt push notifications but we don't need it.

The jitsi integration inside Rocket.Chat often broke. This led us to using meet.jit.si directly. But during the start of COVID, their hosted version had trouble keeping up (I guess lots of traffic). We switched to a selfhosted Big Blue Button at around that time and have stuck to that since. We still use jitsi for all customer calls.

I think Rocket.Chat is probably not a great fit for personal messaging though you can make it work if the individuals involved are "accomodating". Rocket.Chat targets teams and business and has gazillion features targeting them. For individuals, I think what's most important is the client apps and having them butter smooth and flashy stuff which Rocket.Chat mobile apps are not.

@potemkin_ai Did I understand correctly that the reboot is tied to the ubuntu security updates requiring a restart? (i.e this is not a generic automated server reboot feature).

A while ago, we wrote https://git.cloudron.io/cloudron/cloudron-sysadmin which is basically an unattended install. That script most likely won't work / is obsolete.

If there's interest, I can write a quick doc on how to write this script. It's really just a bunch of API calls to automate everything. Fee free to automate in the language of your choice.

I will lock this one since the other one already has 10 upvotes.

@roru2k20 Let's Encrypt recently changed the root certificate since the old one expired. The joplin client is being updated to accomodate this change per https://forum.cloudron.io/topic/5771/sync-error-certificate-has-expired . Not sure about the bank program, maybe you can check if it needs some update. Does the nextcloud client work?