- Fix an issue where custom object were seeded with 2 views, and with
the wrong icon
- ACME becomes YCombinator
- Allow 2 workspaces to have different metadata seeded
- Add many seeds for messages
- Add many seeds for calendar events
- Randomize createdBy for person and companies
---------
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
- Add seeds for notes/tasks
- Adds account manager to companies
- A companies and phone numbers to people
- Add many more opportunities
TODO: add timeline activities
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
We must separate the concept of hydratation which happens at the request
level (take the token and pass auth/user context), from the concept of
authorization which happens at the query/endpoint/mutation level.
Previously, hydratation exemption happened at the operation name level
which is not correct because the operation name is meaningless and
optional. Still this gave an impression of security by enforcing a
blacklist. So in this PR we introduce linting rule that aim to achieve a
similar behavior, now every api method has to have a guard. That way if
and endpoint is not protected by AuthUserGuard or AuthWorspaceGuard,
then it has to be stated explicitly next to its code.
---------
Co-authored-by: Charles Bochet <charles@twenty.com>
BlocknoteJS requires an ESM module where our server is CJS, this forced
us to pin the server-util version, which led us to force the resolution
of several packages, leading to bugs downstream.
From Node 22.12 Node supports requiring ESM modules (available from Node
22.0 with a flag). So I upgrade the module.
I picked Node 22 and not Node 23 or Node 24 because 22 is the LTS and we
don't plan to change node versions frequently.
If you remain on Node 18, things should still mostly work, except if you
edit a Rich Text field.
I also starting changing the default runtime for Serverless Functions
which isn't directly related. This means new serverless functions will
be created on Node 22, but we will still need another PR to migrate
existing serverless functions before September (end of support by AWS).
(In this PR I also remove the upgrade commands from 0.43 since they rely
on Blocknote and I didn't want to have to deal with this)
---------
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
If you tried to add a delay in `refreshObjectMetadataItems` like this`
await new Promise((resolve) => setTimeout(resolve, 5000))`, then this
caused an issue where the user was redirected to his workspace because
the metadata was not loaded.
This happened because I had removed the call to fetch metadata
explicitly in useAuth (instead relying on the effect to fetch it because
it was done twice). I had removed it because this was causing issues in
the onboarding process where /metadata was called too early and then
cached with the wrong reply.
The correct fix is instead to change the fetch policy to `network only`
to stop hiding re-renders to the object metadata effect with Apollo's
cache mechanism. Now the [] reply isn't cached in the onboarding, the
metadata effect is only triggered during initial page load and refresh
should be called explicitely.
I also noticed a bug on the server side where sometimes the frontend was
passing a token for public requests (login token exchange request,
public domain data request). I removed the check so that the backend
completely ignores the token when it's passed on public request. The
downside is that we're losing information for logs (who did that request
to a public endpoint), but it doesn't make much sense to throw
authentication errors on that endpoint imo. Probably a better root-cause
fix would be to understand why a token is still passed on the frontend,
but that would require more investigation — the bug happened when I was
signing up and redirected from the app.xxx domain to the workspace
domain
Creating manual chunk was a bad idea, we should always solve lazy
loading problem at the source instance.
Setting a 4.5MB for the index bundle size, CI will fail if we go above.
There is still a lot of room for optimizations!
- More agressive lazy loading (e.g. xyflow and tiptap are still loaded
in index!)
- Add a prefetch mechanism
- Add stronger CI checks to make sure libraries we've set asides are not
added back
- Fix AllIcons component with does not work as intended (loaded on
initial load)
Small optimization for faster loading (gaining ~80ms - average time of a
click)
It might seem a little over-engineered but there are a lot of edge cases
and I couldn't find a simpler solution
I also tried to tackle Link Chips but it's more complex so this will be
for another PR
Changes for performance improvement.
The primary improvements include replacing GraphQL queries with
REST-based client configuration fetching and making the client config
non render-blocking
Chrome doesn't really respect preloading and was loading it before other
important assets, slowing down the app while in 99% of sessions people
don't check the REST API playground
Fix the following error:
Cannot use a pool after calling end on a pool
<img width="917" alt="Screenshot 2025-05-17 at 14 56 18"
src="https://github.com/user-attachments/assets/63081831-9a7e-4633-8274-de9f8a48dbae"
/>
The problem was that the datasource manager was destroying the
connections when a datasource cache expired.
This PR implements a global PostgreSQL connection pool sharing
mechanism.
- Patches pg.Pool to reuse connection pools across the application when
connection parameters match, reducing resource overhead.
- New environment variables allow enabling/disabling sharing and
configuring pool size, idle timeout, and client exit behavior.
WorkspaceDatasourceFactory will now use shared pools if enabled, this
will avoid recreating 10 connections for each pods for each workspace.
---------
Co-authored-by: Charles Bochet <charlesBochet@users.noreply.github.com>
As discussed with @Weiko
Even though we cache the datasource, the connection expire after
10minutes in TypeORM, that might be the reason why our app is spamming
the proxy asking for connections. Also lowering the pool size.
Revert changes in #12006 as it might still be handy to have the DB
auto-created (e.g. for test or self-hosting users), but if there is a
permission exception we will just ignore it and assume the database
exist in that case
## Architecture Detail
The goal is to merge the two TypeORM schemas.
Having two schemas prevent doing things like fieldMetadata.workspace in
TypeORM, and create useless debates since there is no clear line (is a
serverlessFunction core or metadata? What about events? etc.)
### Before
```
┌───────────────────┐ ┌───────────────────┐
│ core schema │ │ metadata schema │
├───────────────────┤ ├───────────────────┤
│- User │ │- ObjectMetadata │
│- Workspace │ │- FieldMetadata │
│- UserWorkspace │ │- RelationMetadata │
│- etc. │ │- etc. │
└───────────────────┘ └───────────────────┘
```
### After the Migration
```
┌───────────────────────────────────────────┐
│ engine schema │
├───────────────────────────────────────────┤
│- User - ObjectMetadata │
│- Workspace - FieldMetadata │
│- UserWorkspace - RelationMetadata │
│- etc. - etc. │
└───────────────────────────────────────────┘
```
## Strategy
1. During 0.53 we backfill the *_typeorm_migrations* table of the core
schema with all metadata migrations
2. That way in 0.54 we can move the metadata migrations from the
metadata folder to the core folder. We will also edit the migration
files to reference "core" instead of "metadata". For people doing a
fresh install this will run smoothly and create the tables in Core
directly. For people on an existing install, this migrations will not
run because they were added to the *_typeorm_migrations* in 0.53
3. In 0.55 we will rename "core" to something else (for example
"engine")
Note: if someone jumps version, for example skips to 0.54 directly
without having run 0.53 then this could cause issue. In 0.54 we should
consider gating the "migrate:prod" in the docker file so that it's
controlled and run by the upgrade command (and not run if the command
wasn't executed properly)
No need to audit log workflow runs as it's already a form of audit log.
Add more audit log for other objects
Rename MessagingTelemetry to MessagingMonitoring
Merge Analytics and Audit in one (Audit)
---------
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
isEmailVerified was set to false which was annoying in the staging
environment
Also updated password for tim@apple.dev from AppleCar2025 to just
tim@apple.dev since the joke is outdated
We didn't get much complaints on Localization so I guess we can expand
it more and make it the default behavior to use the browser's locale
when you signup
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
This is hard to test without merging PRs unfortunately
Goal of this PR is to replace the action I had introduced since there
was already a similar one in the codebase
We want to have fewer base path for routing.
We will have:
- /files
- /webhooks
- /graphql
- /metadata
- /rest
- /auth
- /healthz
I'm moving /open-api under /rest, and centralizing the webhooks
(removing /stripe and /cloudflare)
Now the source of truth for the version is set during the build process.
We set it as an environment variable from the tags.
We could add it back to the package.json during the build process (from
the git tag), but there is not use for it at the moment since it's not
npm packages.
Let's deprecate Sentry Release and use APP_VERSION instead.
It'll make it more clear in the interface to use named version for bug
analysis, than commit sha
Recoil-sync was causing issues with Firefox, replacing it with a simpler
mechanism to hydrate variables on page load
---------
Co-authored-by: etiennejouan <jouan.etienne@gmail.com>
A small PR but a big step towards making Twenty easier to self-host and
upgrade!
Now changing the tag and pulling a new version should be the only step
to upgrade as migrations script will be ran automatically upon starting
the containers. It was already the case for typeorm migrations, but not
for standard objects migration and data migration scripts. It is still
possible to disable this behavior for the most complex deployments such
as our own cloud.
Refactor query runner to improve the import method for upserts, we now
take into account any unique field and prevent any conflict upfront.
Previously, we would only update if an `id` was passed.
https://github.com/user-attachments/assets/8087b864-ba42-4b6e-abf2-b9ea66e6c467
This is only a first step, there are other things to fix on the frontend
for this to work.
Fixes#11310
Note:
- I changed "system to advanced" because I think we will link that to
the Advanced Mode in the future
- Not sure when to use useCallback / useMemo, AI added some useCallbacks
which I removed but I left some useMemo... Not sure what should be the
rule
- It had to use MenuItem because this sub-menu behavior wasn't available
in the Standard select component. We should probably rename the
"MenuItem" elements to something more generic. I didn't do it in this PR
because I'm not sure about the strategy and it would change a lot of
files.
Mostly renaming objects to avoid conflicts (it was painful because names
were too generic so you could cmd+replace easily)
Also refactoring `useBuildAvailableFieldsForImport`
Advanced mode toggle was in `twenty-ui` which doesn't support Lingui.
I removed lingui from the global package json and moved it to the local
package.json instead to prevent that kind of error from happening again
CSS was loaded in a global context (full screen which might be re-used
for other use cases in the future) instead of a local context.
\+ small update on .env.example
Prepare for better version upgrade system + split admin panel into two
permissions + fix GraphQL generation detection
---------
Co-authored-by: ehconitin <nitinkoche03@gmail.com>
Workspace Member will get their own record page in the future.
This PR lays backend changes to prepare for this:
- Settings most fields on WorkspaceMember as system fields
- Renaming workspaceMember/workspaceMemberId to
forWorkspaceMember/forWorkspaceMemberId as it conflicts with the morph
relationship, if we want to be able to add a workspace member as
favorite
---------
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
This pull request focuses on improving localization by replacing
hardcoded strings with translatable strings using the `Trans` component
from `@lingui/react/macro`. Additionally, it introduces locale support
to several email components. Here are the most important changes:
### Localization Improvements:
* Replaced hardcoded strings with `Trans` components in various email
templates to support localization.
(`packages/twenty-emails/src/emails/clean-suspended-workspace.email.tsx`,
`packages/twenty-emails/src/emails/password-reset-link.email.tsx`,
`packages/twenty-emails/src/emails/password-update-notify.email.tsx`,
`packages/twenty-emails/src/emails/send-email-verification-link.email.tsx`,
`packages/twenty-emails/src/emails/send-invite-link.email.tsx`,
`packages/twenty-emails/src/emails/warn-suspended-workspace.email.tsx`)
[[1]](diffhunk://#diff-ca227a03c0aa66428daff938c743435e8a4dc3ffa960c0952f2697a23e280fdbR6-R25)
[[2]](diffhunk://#diff-ca227a03c0aa66428daff938c743435e8a4dc3ffa960c0952f2697a23e280fdbL42-R45)
[[3]](diffhunk://#diff-523cd37f5680ce418450946f62b7804b6586158efb190ced73920ef0fdf96bc8L1)
[[4]](diffhunk://#diff-523cd37f5680ce418450946f62b7804b6586158efb190ced73920ef0fdf96bc8L23-R23)
[[5]](diffhunk://#diff-cf16aa55d3eeb6be606bbe93de4c83b6f146c49b60d6f512d4b87e49fe14338cL29-R29)
[[6]](diffhunk://#diff-cf16aa55d3eeb6be606bbe93de4c83b6f146c49b60d6f512d4b87e49fe14338cL46-R46)
[[7]](diffhunk://#diff-16b613160f937563ec108176f595d8f275a1d87a5b8245d84df60d775f3efebeL1)
[[8]](diffhunk://#diff-16b613160f937563ec108176f595d8f275a1d87a5b8245d84df60d775f3efebeL22-R22)
[[9]](diffhunk://#diff-0da62e7cc5cfcb32cc25f067fa1d50123047c239af210398f065455ab6700886L1)
[[10]](diffhunk://#diff-0da62e7cc5cfcb32cc25f067fa1d50123047c239af210398f065455ab6700886L42-R41)
[[11]](diffhunk://#diff-0da62e7cc5cfcb32cc25f067fa1d50123047c239af210398f065455ab6700886L57-R56)
[[12]](diffhunk://#diff-483346065c074946a43c18492334bd680422a1d4cb994dc8c3cd39d0208e6016L1-R21)
[[13]](diffhunk://#diff-483346065c074946a43c18492334bd680422a1d4cb994dc8c3cd39d0208e6016L28-R31)
[[14]](diffhunk://#diff-483346065c074946a43c18492334bd680422a1d4cb994dc8c3cd39d0208e6016L53-R55)
### Locale Support:
* Added `locale` prop to email components to dynamically set the locale.
(`packages/twenty-emails/src/emails/clean-suspended-workspace.email.tsx`,
`packages/twenty-emails/src/emails/warn-suspended-workspace.email.tsx`)
[[1]](diffhunk://#diff-ca227a03c0aa66428daff938c743435e8a4dc3ffa960c0952f2697a23e280fdbR6-R25)
[[2]](diffhunk://#diff-483346065c074946a43c18492334bd680422a1d4cb994dc8c3cd39d0208e6016L1-R21)
### SnackBar Messages:
* Replaced hardcoded SnackBar messages with translatable strings using
the `t` function from `@lingui/react/macro`.
(`packages/twenty-front/src/modules/auth/components/VerifyEmailEffect.tsx`,
`packages/twenty-front/src/modules/auth/hooks/useVerifyLogin.ts`,
`packages/twenty-front/src/modules/auth/sign-in-up/hooks/useHandleResendEmailVerificationToken.ts`,
`packages/twenty-front/src/modules/auth/sign-in-up/hooks/useHandleResetPassword.ts`,
`packages/twenty-front/src/modules/object-record/record-field/components/LightCopyIconButton.tsx`,
`packages/twenty-front/src/modules/object-record/record-field/meta-types/display/components/PhonesFieldDisplay.tsx`)
[[1]](diffhunk://#diff-551f2f94eacd8856d22bab7e63dd3ad693f87e9fa9b289864802ebc387f72b42R7)
[[2]](diffhunk://#diff-551f2f94eacd8856d22bab7e63dd3ad693f87e9fa9b289864802ebc387f72b42L24-R29)
[[3]](diffhunk://#diff-551f2f94eacd8856d22bab7e63dd3ad693f87e9fa9b289864802ebc387f72b42L43-R51)
[[4]](diffhunk://#diff-428199461992a01325159f5fbf826d845f05f3361279eccd3f1ce416e0114845R7-R15)
[[5]](diffhunk://#diff-428199461992a01325159f5fbf826d845f05f3361279eccd3f1ce416e0114845L24-R26)
[[6]](diffhunk://#diff-cde42d6abfed63e52c2bda09d537a6577148d0baf957fde75ceaa8657ed58403R5)
[[7]](diffhunk://#diff-cde42d6abfed63e52c2bda09d537a6577148d0baf957fde75ceaa8657ed58403L16-R17)
[[8]](diffhunk://#diff-cde42d6abfed63e52c2bda09d537a6577148d0baf957fde75ceaa8657ed58403L28-R33)
[[9]](diffhunk://#diff-9332c1988864863f12516c2fb77e814af60bedb37c36ffa094f49afc335d5457R5-R17)
[[10]](diffhunk://#diff-9332c1988864863f12516c2fb77e814af60bedb37c36ffa094f49afc335d5457L27-R33)
[[11]](diffhunk://#diff-9332c1988864863f12516c2fb77e814af60bedb37c36ffa094f49afc335d5457L42-R44)
[[12]](diffhunk://#diff-8d64afa825b47ab71d18e3e284408e2097f5fd2365eae84d9d25d3568c48e49cR7)
[[13]](diffhunk://#diff-8d64afa825b47ab71d18e3e284408e2097f5fd2365eae84d9d25d3568c48e49cR20-R28)
[[14]](diffhunk://#diff-6e4361ded2b5656afaeb1befa8b1d23a45b490a1118550da290e27cdb8ebcdceR6)
[[15]](diffhunk://#diff-6e4361ded2b5656afaeb1befa8b1d23a45b490a1118550da290e27cdb8ebcdceR19-R20)
[[16]](diffhunk://#diff-6e4361ded2b5656afaeb1befa8b1d23a45b490a1118550da290e27cdb8ebcdceL29-R38)
There are a lot of requests being done in parallel when you load the
page, and the support chat is responsible for some of them. I don't
think it's critical to have it upon loading so delaying it by 2 seconds
