Adding new smart contracts and generated go files.
Deploy token owner function and master token address getter.
Adding deployer and privilegesLevel columns to community_tokens table.
Passing addressFrom to API calls.
Issue #11250
* rm params
* use url data proto
* feat: community channel and user data parsing; improve testing for urls
---------
Co-authored-by: MishkaRogachev <mishkarogachev@gmail.com>
Fixes an issue where if a community had an admin permission, it would be impossible to join as a normal member because the admin permission wasn't satisfied.
Prior to this commit a control node would add the revealed addresses to
the member struct on the community description, which exposes all those
addresses to the public.
We don't want that. Revealed addresses are exclusively shared with
control nodes and should stay there (although, they might be privately
shared among token masters, see
https://github.com/status-im/status-desktop/issues/11610).
In this commit, we no longer add the revealed addresses to the community
description. The addresses are already stored in the requestToJoin
database table so we can take them from there if we need them.
Closes: https://github.com/status-im/status-desktop/issues/11573
This is a bigger change in how community membership requests are handled
among admins, token masters, owners, and control nodes.
Prior to this commit, all privileged users, also known as
`EventSenders`, were able to accept and reject community membership
requests and those changes would be applied by all users.
This commit changes this behaviour such that:
1. EventSenders can make a decision (accept, reject), but merely forward
their decision to the control node, which ultimately has to confirm
it
2. EventSenders are no longer removing or adding members to and from
communities
3. When an eventsender signaled a decision, the membership request will
enter a pending state (acceptedPending or rejectedPending)
4. Once a decision was made by one eventsender, no other eventsender can
override that decision
This implementation is covered with a bunch of tests:
- Ensure that decision made by event sender is shared with other event
senders
- `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()`
- `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()`
- Ensure memebrship request stays pending, until control node has
confirmed decision by event senders
- `testAcceptMemberRequestToJoinNotConfirmedByControlNode()`
- `testRejectMemberRequestToJoinNotConfirmedByControlNode()`
- Ensure that decision made by event sender cannot be overriden by other
event senders
- `testEventSenderCannotOverrideRequestToJoinState()`
These test cases live in three test suites for different event sender
types respectively
- `OwnerWithoutCommunityKeyCommunityEventsSuite`
- `TokenMasterCommunityEventsSuite`
- `AdminCommunityEventsSuite`
In addition to the changes mentioned above, there's also a smaller
changes that ensures membership requests to *not* attached revealed wallet
addresses when the requests are sent to event senders (in addition to
control nodes).
Requests send to a control node will still include revealed addresses as
the control node needs them to verify token permissions.
This commit does not yet handle the case of event senders attempting to
kick and ban members.
Similar to accepting and rejecting membership requests, kicking and
banning need a new pending state. However, we don't track such state in
local databases yet so those two cases will be handled in future commit
to not have this commit grow larger.
* feat: proposal for collecting community metrics
https://github.com/status-im/status-desktop/issues/11152
* feat: collecting community message metrics with test
* feat: implement both strategies for fetching community metrics
* fix: review fixes
* fix: calc counts for timestamps
If a message is sent with only 1 image, the album is not generated (no albumID), so then, in the notification handling code, it didn't use the right ID, because it thought it had to use the AlbumID for the message ID
- distribute ratchet keys at both community and channel levels
- use explicit `HashRatchetGroupID` in ecryption layer, instead of
inheriting `groupID` from `CommunityID`
- populate `HashRatchetGroupID` with `CommunityID+ChannelID` for
channels, and `CommunityID` for whole community
- hydrate channels with members; channel members are now subset of
community members
- include channel permissions in periodic permissions check
closes: status-im/status-desktop#10998
This component decouples key distribution from the Messenger, enhancing
code maintainability, extensibility and testability.
It also alleviates the need to impact all methods potentially affecting
encryption keys.
Moreover, it allows key distribution inspection for integration tests.
part of: status-im/status-desktop#10998
**This is a breaking change!**
Prior to this commit we had `AddCommunityToken(token *communities,
croppedImage CroppedImage)` that we used to
1. add a `CommunityToken` to the user's database and
2. to create a `CommunityTokenMetadata` from it which is then added to
the community's `CommunityDescription` and published to its members
However, I've then discovered that we need to separate these two things,
such that we can deploy a community token, then add it to the database
only for tracking purposes, **then** add it to the community description
(and propagate to members) once we know that the deploy tx indeed went
through.
To implement this, this commit introduces a new API
`SaveCommunityToken(token *communities.CommunityToken, croppedImage
CroppedImage)` which adds the token to the database only and doesn't
touch the community description.
The `AddCommunityToken` API is then changed that it's exclusively used
for adding an already saved `CommunityToken` to the community
description so it can be published to members. Hence, the signature is
now `AddCommunityToken(communityID string, chainID int, address
string)`, which makes this a breaking change.
Clients that used `AddCommunityToken()` before now need to ensure that
they first call `SaveCommunityToken()` as `AddCommunityToken()` will
fail otherwise.
* chore: make the owner without the community private key behave like an admin
* Add test for the owner without community key
* chore: refactor Community fn names related to the roles
If user followed onboarding flow to recover his account using seed phrase or keycard,
then `ProcessBackedupMessages` property of node config json object should be set to
`true`, otherwise it should be set to `false` or be omitted.
- Fixed redundant permissions check. If community is set to auto-accept,
then permissions would be checked twice, in
`HandleCommunityRequestToJoin` and `AcceptRequestToJoinCommunity`.
Mitigated it by returning from `HandleCommunityRequestToJoin` immediately
in case of auto-accept.
- Extracted `accountsSatisfyPermissionsToJoin` to remove code
duplication and simplify the logic.
* feat: add api to remove private key and separete owner from private key ownership
For https://github.com/status-im/status-desktop/issues/11475
* feat: introduce IsControlNode for Community
* feat: remove community private key from syncing
* feat: add IsControlNode flag to Community json serialisation
* Update protocol/protobuf/pairing.proto
Co-authored-by: Jonathan Rainville <rainville.jonathan@gmail.com>
---------
Co-authored-by: Jonathan Rainville <rainville.jonathan@gmail.com>
* mute and unmute all community chats when community mute status changes
* unmute community when atleast one channel is unmuted
* fix: save community, extend the function to save muted state and mute duration
chore:
- add CommunityEventsMessage
- refactor community_admin_event to accept a list of events and patch a CommunityDescription
- save/read community events into/from database
- publish and handle community events message
- fixed admin category tests
- rename AdminEvent to Events or CommunityEvents
Adds airdropAddress to the request to join params and a is_airdrop_address flag in the communities_requests_to_join_revealed_addresses table.
This airdropAddress is used by the owner to know which address to use when airdropping
I have encountered a crash in the app after syncing, looks like a
community has been created without a Config.
This crashes the app after the user logs in.
This commit prevents the app from crashing, but does not fix the
underlaying issue (that's something I will have to investigate).
Added tests to validate the behavior.
Check community exists
* feat: don't remove sent mutual state messages on accepting a CR
* fix: don't send mutual state message for a new contact
* chore: move mutual state messages to `addContact`
* fix: use one chat for mutual state messages and contact requests
* fix: change `added` mutual state updatede messages to `accepted`
* feat: Use different content type for each mutual state event system message
* chore: use constants for mutual event system messages test, review fixes
* chore: fix tests related to local contacts map
Modify API to handle also ERC20 tokens.
Modify community_tokens table - keep supply as string since string is easly convertible to bigint.BigInt.
Use bigint.BigInt for supply functions and fields.
Issue #11129
This is the second step of improvements over keypairs/keycards/accounts.
- `SyncKeycardAction` protobuf removed
- `SyncKeypair` protobuf is used for syncing keycards state as well as for all
keycards related changes
- `last_update_clock` column removed from `keypairs` table cause as well as
for accounts, any keycard related change is actually a change made on a related
keypair, thus a keypair's clock keeps the clock of the last change
- `position` column added to `keypairs` table, needed to display keycards in
the same order accross devices
This is the first step of improvements over keypairs/keycards/accounts.
- `SyncKeypairFull` protobuf removed
- `SyncKeypair` protobuf is used for syncing all but the watch only accounts
- `SyncAccount` is used only for syncing watch only accounts
- related keycards are synced together with a keypair
- on any keypair change (either it's just a keypair name or any change made over an
account which belongs to that keypair) entire keypair is synced including related keycards
- on any watch only account related change, that account is synced with all its details
This commit extends the `AddCommunityToken` API to also expect an
optional `CroppedImage`, which will be used instead of the `ImageBase64`
path provided by `CommunityToken`, to calculate the actual base64
encoded image.
* feat(share-links): Add protobuf and encode/decode url data methods
* feat(new-links-format): Adds generators for new links format
* feat: add parsing for new links format
* feat: add messenger-level pubkey serialization and tests
* feat: fix and test CreateCommunityURLWithChatKey
* feat: impl and test parseCommunityURLWithChatKey
* feat: fix and test CreateCommunityURLWithData
* feat: impl and test parseCommunityURLWithData (not working)
* feat: UrlDataResponse as response share urls api
* feat: impl& tested ShareCommunityChannelURLWithChatKey
* feat: impl & tested ParseCommunityChannelURLWithChatKey
* fix: bring urls to new format
* feat: add regexp for community channel urls
* feat: impl & test contact urls with chatKey, Ens and data
* fix: encodeDataURL/encodeDataURL patch from Samyoul
* fix: fix unmarshalling protobufs
* fix: fix minor issues, temporary comment TestParseUserURLWithENS
* fix: allow url to contain extra `#` in the signature
* fix: check signatures with SigToPub
* chore: lint fixes
* fix: encode the signature
* feat: Check provided channelID is Uuid
* fix(share-community-url): Remove if community encrypted scope
* fix: review fixes
* fix: use proto.Unmarshal instead of json.Marshal
* feat(share-urls): Adds TagsIndices to community data
* feat: support tag indices to community url data
---------
Co-authored-by: Boris Melnik <borismelnik@status.im>
When deleting a message for me, the image url wasn't preserved,
resulting in the image disappearing on the client side.
This commit adds the processing of returned messages so that the image
is preserved.
- Add ERC20 contract
- Add decimals field to community_tokens db table
- Adjusting API to handle assets deployment
- Add decimals field to CommunityTokenMetadata
Issue #10987
Nameserver is passed by the OS on creation/restore, this commit adds the
ability to pass it at login time.
We don't want to store it on disk since that's bound to change, and
currently there's a bug on golang that prevents getting the DNS from the
system on android.
There's only one scenario in which a `RevealedAccount` will have an
empty `ChainIDs` list attached to it:
When the community in question requires users to satisfy certain
criteria to join, and the user's wallet does not own the necessary funds
on any of the supported chains.
If there are **no** permissions to join on the community, then we want
to reveal all (selected) accounts with all supported chainIDs.
This is necessary so that, once the community *does* become
permissioned, it'll have address + chain information from all joined
members.
Closes: https://github.com/status-im/status-desktop/issues/11255
This commit adds new tables to the database and APIs in `Messenger` and
communities `Manager` to store `CheckChannelPermissionsResponse`s.
The responses are stored whenever channel permissions have been checked.
The reason we're doing this is so that clients can retrieve the last
known channel permission state before waiting for onchain checks to
finish.
Sometimes confirmation for raw messages are received before the record
is actually saved in the database.
In this case, the code will preserve the Sent status.
Improve `RequestToJoinCommunity` to accept `Addresses` in the request. If `Addresses` is not empty, we then only pass to the owner the selected addresses. The others are ignored.
Does not validate that the addresses in the slice are part of the user's wallet. Those not part of the wallet are just ignored.
This API is used to get a permission status of all channels of a given
community.
Clients can use this API to get the provided information for all
community channels with a single RPC call instead of doing one call
for each channel separately.
Similar to `CheckPermissionToJoin()` we now get
a `CheckChannelPermissions()` API.
It will rely on the same `PermissionResponse` types, but gives
information about both `ViewOnlyPermissions` and
`ViewAndPostPermissions`.
This seems to be a bug that was introduced when two features, admin
permissions and "always reveal wallet accounts" where merged.
We need to make sure we **first** check the revealed accounts and only
**then** do we perform permission checks on them. Otherwise we can run
into scenarios where fake addresses are used and users will be accepted
to the community.
found
Turns out that, when we return with an error, instead of
a non-statisfied check permissions response, we can run into cases where
members that should be kicked are not kicked.
Change smart contract with new API.
Update gas amount for deployment.
Add Burn() and EstimateBurn() functions.
Add RemainingSupply() functions.
Issue #10816
It happens that an envelope is sent before it's tracked, resulting in
long delays before the envelope is marked as sent.
This commit changes the behavior of the code so that order is now
irrelevant.
When we reply to our own message with an image, we didn't set the URL of
the image of the message, which resulted in the image not being
displayed correctly.
This commit does a few things:
- Adds a migration that adds chainids to communities_request_to_join_revealed_addresses
- Removes RevealedAddress in favor of RevealedAccount which is now a struct that contains the revealed address, as well as the signature and a list of chain IDs on which to check for user funds
- Changes the logic of sending requests to join a community, such that after creating address signatures, the user node will also check which of the addresses has funds on which networks for the community's token permissions, and add the chainds to the RevealedAccount
- Updates checkPermissionToJoin() such that only relevant chainids are used when checking user's funds. Chain IDs are retrieved from RevealedAccounts and matched against token permission criteria chain IDs
There's two scenarios in which we're leaving a community:
We either get kicked or we leave ourselves.
In case of leaving ourselves it's fine to unsubscribe from further
community updates be cause we deliberately chose to leave.
In case of being kicked however, this is different.
Say I'm kicked from a community because its token permissions have
changed, in this case we don't want clients to manually re-subscribe to
the community to get informed when there were further changes.
status-go should rather not unsubscribe if we know for sure we've been
kicked by someone else.
fix flaky test: TestRetrieveBlockedContact
resolve conflict when rebase origin/develop
Feat/sync activity center notification (#3581)
* feat: sync activity center notification
* add test
* fix lint issue
* fix failed test
* addressed feedback from sale
* fix failed test
* addressed feedback from ilmotta
go generate ./protocol/migrations/sqlite/...
feat: add updated_at for syncing activity center notification
* feat: add mutual state update system message
* feat: send mutual state update on accepting CR
* feat: send mutual state update when removing a contact
* fix: don't send MutualStateUpdateMessage over wire
* fix: mutual state update message text fixed
* fix: new clock to ensure system message after CR and add chat to the response
* feat: add AC notification for contact removal
* feat: replace "sent" mutual state system message with "added"
Extended the migration process with a generic way of applying custom
migration code on top of the SQL files. The implementation provides
a safer way to run GO code along with the SQL migrations and possibility
of rolling back the changes in case of failure to keep the database
consistent.
This custom GO migration is needed to extract the status from
the JSON blob receipt and store it in transfers table.
Other changes:
- Add NULL DB value tracking to JSONBlob helper
- Index status column on transfers table
- Remove unnecessary panic calls
- Move log_parser to wallet's common package and use to extract token
identity from the logs
Notes:
- there is already an index on transfers table, sqlite creates one for
each unique constraint therefore add only status to a new index
- the planned refactoring and improvements to the database have been
postponed due to time constraints. Got the time to migrate the data
though, extracting it can be done later for a more efficient
implementation
Update status-desktop #10746
* chore(upgradeSQLCipher): Upgrading SQLCipher to version 5.4.5
Changes:
### github.com/mutecomm/go-sqlcipher
1. The improved crypto argorighms from go-sqlcipher v3 are merged in v4
Tags:
v4.4.2-status.1 - merge `burn_stack` improvement
v4.4.2-status.2 - merge `SHA1` improvement
v4.4.2-status.4- merge 'AES' improvement
2. Fixed `go-sqlcipher` to support v3 database in compatibility mode (`sqlcipher` already supports this) (Tag: v4.4.2-status.3)
3. Upgrade `sqlcipher` to v5.4.5 (Tag: v4.5.4-status.1)
### github.com/status-im/migrate/v4
1. Upgrade `go-sqlcipher` version in `github.com/status-im/migrate/v4`
### status-go
1. Upgrade `go-sqlcipher` and `migrate` modules in status-go
2. Configure the DB connections to open the DB in v3 compatibility mode
* chore(upgradeSQLCipher): Use sqlcipher v3 configuration to encrypt a plain text database
* chore(upgradeSQLCipher): Scanning NULL BLOB value should return nil
Fixing failing tests: TestSyncDeviceSuite/TestPairingSyncDeviceClientAsReceiver; TestSyncDeviceSuite/TestPairingSyncDeviceClientAsSender
Considering the following configuration:
1. Table with BLOB column has 1 NULL value
2. Query the value
3. Rows.Scan(&dest sql.NullString)
Expected: dest.Valid == false; dest.String == nil
Actual: dest.Valid == true; dest.String == ""
* chore: Bump go-sqlcipher version to include NULL BLOB fix
Add support for unfurling a wider range of websites. Most code changes are
related to the implementation of a new Unfurler, an OEmbedUnfurler, which is
necessary to get metadata for Reddit URLs using oEmbed, since Reddit does not
support OpenGraph meta tags. The new unfurler will also be useful for other
websites, like Twitter. Also the user agent was changed, and now more websites
consider status-go reasonably human.
Related to issue https://github.com/status-im/status-mobile/issues/15918
Example hostnames that are now unfurleable: reddit.com, open.spotify.com,
music.youtube.com
Other improvements:
- Better error handling, especially because I wasn't wrapping errors correctly.
I also removed the unnecessary custom error UnfurlErr.
- I made tests truly deterministic by parameterizing the http.Client instance
and by customizing its Transport field (except for some failing conditions
where it's even good to hit the real servers).
This commit adds LoginAccount endpoint.
This makes it consistent with CreateAccount and RestoreAccount as they
use similar config.
The notable difference with the previous endpoint is the API, which is
the same as CreateAccount/RestoreAccount, and the fact that it will
override your networks configuration.
Storing them in the config is now not needed anymore, as that's always
driven from the backend, and we won't allow custom networks in the new
wallet.
Fixes an issue where if a group chat was first received from a non-contact, and later received from a contact, it still wouldn't save it as active.
That's because we checked if we were **newly** added instead of just if we were added. That meant that in the case I described above, the chat would then never have the chance to be set active.
* fix(community): stop re-joining comm when receiving a sync community msg
Fixes an issue with chats being reset. Since joining a community resaves the chats with the synced default value, it resets the sate of the chats, losing the unread messages, the muted state and more.
The solution is to block the re-joining of the community. In the case of the sync, we catch that error and just continue on.
* fix(import): fix HandleImport not saving the chat
Doesn't change much, but it could have caused issues in the future, so since we might have modified the chat, we make sure to save them
Also adds a test
* fix tests
- old `accounts` table is moved/mapped to `keypairs` and `keypairs_accounts`
- `keycards` table has foreign key which refers to `keypairs.key_uid`
- `Keypair` introduced as a new type
- api endpoints updated according to this change
* fix: create a CR on contact sync with received CR state
* fix: create a CR on contact sync with sent CR state
* Review fixes
* Fix: ignore own contact installation or syncing
This commit does a few things:
1) Extend create/import account endpoint to get wallet config, some of
which has been moved to the backend
2) Set up a loop for retrieving balances every 10 minutes, caching the
balances
3) Return information about which checks are not passing when trying to
join a token gated community
4) Add tests to the token gated communities
5) Fixes an issue with addresses not matching when checking for
permissions
The move to the wallet as a background task is not yet complete, I need
to publish a signal, and most likely I will disable it before merging
for now, as it's currently not used by desktop/mobile, but the PR was
getting to big
This is the initial implementation for the new URL unfurling requirements. The
most important one is that only the message sender will pay the privacy cost for
unfurling and extracting metadata from websites. Once the message is sent, the
unfurled data will be stored at the protocol level and receivers will just
profit and happily decode the metadata to render it.
Further development of this URL unfurling capability will be mostly guided by
issues created on clients. For the moment in status-mobile:
https://github.com/status-im/status-mobile/labels/url-preview
- https://github.com/status-im/status-mobile/issues/15918
- https://github.com/status-im/status-mobile/issues/15917
- https://github.com/status-im/status-mobile/issues/15910
- https://github.com/status-im/status-mobile/issues/15909
- https://github.com/status-im/status-mobile/issues/15908
- https://github.com/status-im/status-mobile/issues/15906
- https://github.com/status-im/status-mobile/issues/15905
### Terminology
In the code, I've tried to stick to the word "unfurl URL" to really mean the
process of extracting metadata from a website, sort of lower level. I use "link
preview" to mean a higher level structure which is enriched by unfurled data.
"link preview" is also how designers refer to it.
### User flows
1. Carol needs to see link previews while typing in the chat input field. Notice
from the diagram nothing is persisted and that status-go endpoints are
essentially stateless.
```
#+begin_src plantuml :results verbatim
Client->>Server: Call wakuext_getTextURLs
Server-->>Client: Normalized URLs
Client->>Client: Render cached unfurled URLs
Client->>Server: Unfurl non-cached URLs.\nCall wakuext_unfurlURLs
Server->>Website: Fetch metadata
Website-->>Server: Metadata (thumbnail URL, title, etc)
Server->>Website: Fetch thumbnail
Server->>Website: Fetch favicon
Website-->>Server: Favicon bytes
Website-->>Server: Thumbnail bytes
Server->>Server: Decode & process images
Server-->>Client: Unfurled data (thumbnail data URI, etc)
#+end_src
```
```
,------. ,------. ,-------.
|Client| |Server| |Website|
`--+---' `--+---' `---+---'
| Call wakuext_getTextURLs | |
| ---------------------------------------> |
| | |
| Normalized URLs | |
| <- - - - - - - - - - - - - - - - - - - - |
| | |
|----. | |
| | Render cached unfurled URLs | |
|<---' | |
| | |
| Unfurl non-cached URLs. | |
| Call wakuext_unfurlURLs | |
| ---------------------------------------> |
| | |
| | Fetch metadata |
| | ------------------------------------>
| | |
| | Metadata (thumbnail URL, title, etc)|
| | <- - - - - - - - - - - - - - - - - -
| | |
| | Fetch thumbnail |
| | ------------------------------------>
| | |
| | Fetch favicon |
| | ------------------------------------>
| | |
| | Favicon bytes |
| | <- - - - - - - - - - - - - - - - - -
| | |
| | Thumbnail bytes |
| | <- - - - - - - - - - - - - - - - - -
| | |
| |----. |
| | | Decode & process images |
| |<---' |
| | |
| Unfurled data (thumbnail data URI, etc)| |
| <- - - - - - - - - - - - - - - - - - - - |
,--+---. ,--+---. ,---+---.
|Client| |Server| |Website|
`------' `------' `-------'
```
2. Carol sends the text message with link previews in the RPC request
wakuext_sendChatMessages. status-go assumes the link previews are good
because it can't and shouldn't attempt to re-unfurl them.
```
#+begin_src plantuml :results verbatim
Client->>Server: Call wakuext_sendChatMessages
Server->>Server: Transform link previews to\nbe proto-marshalled
Server->DB: Write link previews serialized as JSON
Server-->>Client: Updated message response
#+end_src
```
```
,------. ,------. ,--.
|Client| |Server| |DB|
`--+---' `--+---' `+-'
| Call wakuext_sendChatMessages| |
| -----------------------------> |
| | |
| |----. |
| | | Transform link previews to |
| |<---' be proto-marshalled |
| | |
| | |
| | Write link previews serialized as JSON|
| | -------------------------------------->
| | |
| Updated message response | |
| <- - - - - - - - - - - - - - - |
,--+---. ,--+---. ,+-.
|Client| |Server| |DB|
`------' `------' `--'
```
3. The message was sent over waku and persisted locally in Carol's device. She
should now see the link previews in the chat history. There can be many link
previews shared by other chat members, therefore it is important to serve the
assets via the media server to avoid overloading the ReactNative bridge with
lots of big JSON payloads containing base64 encoded data URIs (maybe this
concern is meaningless for desktop). When a client is rendering messages with
link previews, they will have the field linkPreviews, and the thumbnail URL
will point to the local media server.
```
#+begin_src plantuml :results verbatim
Client->>Server: GET /link-preview/thumbnail (media server)
Server->>DB: Read from user_messages.unfurled_links
Server->Server: Unmarshal JSON
Server-->>Client: HTTP Content-Type: image/jpeg/etc
#+end_src
```
```
,------. ,------. ,--.
|Client| |Server| |DB|
`--+---' `--+---' `+-'
| GET /link-preview/thumbnail (media server)| |
| ------------------------------------------> |
| | |
| | Read from user_messages.unfurled_links|
| | -------------------------------------->
| | |
| |----. |
| | | Unmarshal JSON |
| |<---' |
| | |
| HTTP Content-Type: image/jpeg/etc | |
| <- - - - - - - - - - - - - - - - - - - - - |
,--+---. ,--+---. ,+-.
|Client| |Server| |DB|
`------' `------' `--'
```
### Some limitations of the current implementation
The following points will become separate issues in status-go that I'll work on
over the next couple weeks. In no order of importance:
- Improve how multiple links are fetched; retries on failure and testing how
unfurling behaves around the timeout limits (deterministically, not by making
real HTTP calls as I did). https://github.com/status-im/status-go/issues/3498
- Unfurl favicons and store them in the protobuf too.
- For this PR, I added unfurling support only for websites with OpenGraph
https://ogp.me/ meta tags. Other unfurlers will be implemented on demand. The
next one will probably be for oEmbed https://oembed.com/, the protocol
supported by YouTube, for example.
- Resize and/or compress thumbnails (and favicons). Often times, thumbnails are
huge for the purposes of link previews. There is already support for
compressing JPEGs in status-go, but I prefer to work with compression in a
separate PR because I'd like to also solve the problem for PNGs (probably
convert them to JPEGs, plus compress them). This would be a safe choice for
thumbnails, favicons not so much because transparency is desirable.
- Editing messages is not yet supported.
- I haven't coded any artificial limit on the number of previews or on the size
of the thumbnail payload. This will be done in a separate issue. I have heard
the ideal solution may be to split messages into smaller chunks of ~125 KiB
because of libp2p, but that might be too complicated at this stage of the
product (?).
- Link preview deletion.
- For the moment, OpenGraph metadata is extracted by requesting data for the
English language (and fallback to whatever is available). In the future, we'll
want to unfurl by respecting the user's local device language. Some websites,
like GoDaddy, are already localized based on the device's IP, but many aren't.
- The website's description text should be limited by a certain number of
characters, especially because it's outside our control. Exactly how much has
not been decided yet, so it'll be done separately.
- URL normalization can be tricky, so I implemented only the basics to help with
caching. For example, the url https://status.im and HTTPS://status.im are
considered identical. Also, a URL is considered valid for unfurling if its TLD
exists according to publicsuffix.EffectiveTLDPlusOne. This was essential,
otherwise the default Go url.Parse approach would consider many invalid URLs
valid, and thus the server would waste resources trying to unfurl the
unfurleable.
### Other requirements
- If the message is edited, the link previews should reflect the edited text,
not the original one. This has been aligned with the design team as well.
- If the website's thumbnail or the favicon can't be fetched, just ignore them.
The only mandatory piece of metadata is the website's title and URL.
- Link previews in clients should be generated in near real-time, that is, as
the user types, previews are updated. In mobile this performs very well, and
it's what other clients like WhatsApp, Telegram, and Facebook do.
### Decisions
- While the user typing in the input field, the client is constantly (debounced)
asking status-go to parse the text and extract normalized URLs and then the
client checks if they're already in its in-memory cache. If they are, no RPC
call is made. I chose this approach to achieve the best possible performance
in mobile and avoid the whole RPC overhead, since the chat experience is
already not smooth enough. The mobile client uses URLs as cache keys in a
hashmap, i.e. if the key is present, it means the preview is readily available
(naive, but good enough for now). This decision also gave me more flexibility
to find the best UX at this stage of the feature.
- Due to the requirement that users should be able to see independent loading
indicators for each link preview, when status-go can't unfurl a URL, it
doesn't return it in the response.
- As an initial implementation, I added the BLOB column unfurled_links to the
user_messages table. The preview data is then serialized as JSON before being
stored in this column. I felt that creating a separate table and the related
code for this initial PR would be inconvenient. Is that reasonable to you?
Once things stabilize I can create a proper table if we want to avoid this
kind of solution with serialized columns.
* fix: unable to reset password for newly created account using CreateAccountAndLogin
* remove unnecessary print
* add TestCreateAccountAndLogin
* update TestCreateAccountAndLogin
* bump version
* sync local deleted messages
* rebase
* add REPLACE
* fix lint
* defer rows.Close() / rename function
* add local pair test
* replace unused clock with _
Michal Iskierko
Felicio Mununga
Andrea Maria Piana
Mikhail Rogachev
flexsurfer
frank
Patryk Osmaczko
Alexander
Igor Sirotin
Jonathan Rainville
Sale Djenic
Pascal Precht
Mykhailo Prakhov
Volodymyr Kozieiev
Anthony Laibe
Dario Gabriel Lipicar
Boris Melnik
Samuel Hawksby-Robinson
John Ngei
Ibrahem Khalil
Roman Volosovskyi
Khushboo-dev-cpp
MishkaRogachev
mprakhov
Stefan
Alex Jbanca
Omar Basem
Icaro Motta