]> git.proxmox.com Git - mirror_novnc.git/log
mirror_novnc.git
2 years agofeat: add French localization strings master
Jose (Ito) Matsuda [Fri, 7 May 2021 02:35:18 +0000 (22:35 -0400)]
feat: add French localization strings

2 years agoMerge pull request #1545 from ascillato/patch-2
Samuel Mannehed [Sat, 24 Apr 2021 23:20:02 +0000 (01:20 +0200)]
Merge pull request #1545 from ascillato/patch-2

Update Spanish Translation

3 years agoUpdate Spanish Translation
Adrian Scillato [Fri, 23 Apr 2021 15:10:53 +0000 (12:10 -0300)]
Update Spanish Translation

Update Spanish Translation.

Added missing translations. Fixed some typos.
Also, some changes were made to use a more formal and international spanish.

3 years agoInitiate connection from RFB constructor
Pierre Ossman [Sun, 18 Apr 2021 09:26:51 +0000 (11:26 +0200)]
Initiate connection from RFB constructor

We need to do this in order to safely attach to existing WebSocket
objects. There may be message events already pending so we must set up
our event handlers before returning.

This means we will now throw errors instead of generating "disconnect"
events on problems as the caller no longer has the opportunity to set up
event handlers.

This might have been the correct approach from the start as it mimics
how e.g. the WebSocket constructor works.

3 years agoDon't fake open events in Websock
Pierre Ossman [Sun, 18 Apr 2021 11:55:11 +0000 (13:55 +0200)]
Don't fake open events in Websock

We don't know if the caller is prepared to receive those events right
now as normally they would get them on a fresh new stack later. We also
can't delay delivery since then we might deliver the event after any
pending "message" events.

Better to push the problem one layer up to the caller, which now needs
to be more aware of the state of the WebSocket object it is trying to
use.

3 years agoRefuse to use already closed WebSocket objects
Pierre Ossman [Sun, 18 Apr 2021 11:53:43 +0000 (13:53 +0200)]
Refuse to use already closed WebSocket objects

We can't do anything useful with them anyway.

3 years agoProvide readyState for Websock objects
Pierre Ossman [Sun, 18 Apr 2021 10:52:16 +0000 (12:52 +0200)]
Provide readyState for Websock objects

It mainly reports the state of the underlying object in consistent
manner.

3 years agoMove Websock event handlers to own methods
Pierre Ossman [Sun, 18 Apr 2021 09:39:03 +0000 (11:39 +0200)]
Move Websock event handlers to own methods

Avoid cluttering up the RFB constructor.

3 years agoAdd unit tests for connect/attach errors
Pierre Ossman [Sun, 18 Apr 2021 09:07:43 +0000 (11:07 +0200)]
Add unit tests for connect/attach errors

3 years agoAdd unit tests for passing WebSocket objects
Pierre Ossman [Fri, 16 Apr 2021 14:47:56 +0000 (16:47 +0200)]
Add unit tests for passing WebSocket objects

3 years agoStop explicitly testing connection states
Pierre Ossman [Fri, 16 Apr 2021 13:40:45 +0000 (15:40 +0200)]
Stop explicitly testing connection states

These are internal and we should be testing the externally observable
behaviour.

3 years agoRevert "Fixed a race condition when attaching to an existing socket"
Pierre Ossman [Fri, 16 Apr 2021 11:28:47 +0000 (13:28 +0200)]
Revert "Fixed a race condition when attaching to an existing socket"

This reverts commit ef27628c6dff6120b0ed0d4728cc6e8a32b7be53. By
bypassing setTimeout() it creates other race conditions so this is not
the proper fix for the issue.

3 years agoMerge pull request #1537 from TimSBSquare/bugfix/existing-channel-race-condition
Samuel Mannehed [Tue, 30 Mar 2021 13:59:37 +0000 (15:59 +0200)]
Merge pull request #1537 from TimSBSquare/bugfix/existing-channel-race-condition

Fixed a race condition when attaching to an existing socket

3 years agoFixed a race condition when attaching to an existing socket
Tim Stableford [Tue, 23 Mar 2021 12:58:49 +0000 (12:58 +0000)]
Fixed a race condition when attaching to an existing socket

This is an error that presents itself with RTCDataChannel's, I suspect this could not
happen with a pre-existing WebSocket.

If the remote connection creates a data channel then the local (VNC) side gets a channel
created callback. It may also be the case that in that very same tick the socket is also
opened and buffered data received. This meant that (in my tests) about 1/3 of the time
noVNC would fail to respond to the initial message from the server because it was received
and subsequently not handled during that initial tick.

Also made the documentation reflect this new behaviour and document the existing behaviour.

3 years agoAdd Portuguese (Brazil) translation
Liddack [Tue, 16 Mar 2021 01:19:02 +0000 (22:19 -0300)]
Add Portuguese (Brazil) translation

3 years agoAdd VeNCrypt Plain authentication tests
lhchavez [Sat, 13 Mar 2021 13:40:38 +0000 (05:40 -0800)]
Add VeNCrypt Plain authentication tests

This change adds tests for the VeNCrypt Plain authentication. In doing
that, this also fixes a typo that was introduced in a recent change.

3 years agoMerge branch 'homogenize-credentials-testing' of https://github.com/lhchavez/noVNC
Pierre Ossman [Thu, 11 Mar 2021 15:47:35 +0000 (16:47 +0100)]
Merge branch 'homogenize-credentials-testing' of https://github.com/lhchavez/noVNC

3 years agoMerge branch 'feature/support-existing-rtcdatachannel-or-websocket-squashed' of https...
Pierre Ossman [Thu, 11 Mar 2021 15:30:22 +0000 (16:30 +0100)]
Merge branch 'feature/support-existing-rtcdatachannel-or-websocket-squashed' of https://github.com/TimSBSquare/noVNC

3 years agoNormalize the credentials presence check
lhchavez [Thu, 11 Mar 2021 13:38:13 +0000 (05:38 -0800)]
Normalize the credentials presence check

Most places that check for the presence / absence of credentials compare
them against `undefined`, except the one for Plain authentication.

This change makes the very last place to use the same pattern (instead
of checking for falsiness) for consistency. Additionally, there are ways
to configure PAM to accept empty passwords, so it's possible for a user
to legitimately send an empty string as password.

3 years agoAdded support for RTCDataChannel
Tim Stableford [Thu, 18 Feb 2021 08:42:34 +0000 (08:42 +0000)]
Added support for RTCDataChannel

This work is originally by Ryan Castner <castner.rr@gmail.com> and
submitted as a PR here https://github.com/novnc/noVNC/pull/1362

Architecturally it is much the same except it doesn't rename a lot
of variables to make this more reviewable. It also avoids unrelated
changes such as replacing .onclose with an event listener, which
caused numerous test failures.

It also adds in ppoffice's fix to initialise the buffers.

Like the original author I don't have enough time available to
refactor this project to the new style event listeners.

Review cleanup for RTCDataChannel support (see below)

* More descriptive error when url or channel not set.
* Moved websocket property check to WebSock.
  This had unintended consequences in the tests that required some
  fixup. Mostly due to some tests not always passing FakeWebsocket.
  FakeWebsocket also needs to set the listeners to null to be compatible
  with what is in thw browser and expected by the property check code.
  The property check code now also takes into account class prototypes
  for test compatibility.
* Removed unreachable code.
* Reverted comment.
* Cleanup raw channel reference in rfb on websock close.
* Use readyState to check whether a socket is open rather than assuming.
* Updated RFB constructor documentation

Removed an unused boolean passed to attach

3 years agoAllow longer passwords in Plain authentication
lhchavez [Thu, 4 Mar 2021 01:34:02 +0000 (17:34 -0800)]
Allow longer passwords in Plain authentication

Some people have longer passwords than 256 characters (hooray for
password managers!). Server implementations also allow longer passwords:
TigerVNC allows up to 1024 characters.

3 years agoChange phrasing for discussion group link
Pierre Ossman [Fri, 15 Jan 2021 15:16:38 +0000 (16:16 +0100)]
Change phrasing for discussion group link

It's listed with the issue templates, so make sure it follows the same
style in phrasing.

3 years agoAdd link to discussion group from issues
Pierre Ossman [Fri, 15 Jan 2021 15:14:23 +0000 (16:14 +0100)]
Add link to discussion group from issues

So that people can easily find it and not file bug reports for
things that are just questions.

3 years agoHide link to create blank issues
Pierre Ossman [Fri, 15 Jan 2021 15:08:13 +0000 (16:08 +0100)]
Hide link to create blank issues

We want users to use the templates so we don't miss any relevant
information.

3 years agoUpdate Japanese translation
NNN1590 [Fri, 15 Jan 2021 05:04:49 +0000 (14:04 +0900)]
Update Japanese translation

3 years agoMerge branch 'more_noie' of https://github.com/CendioOssman/noVNC
Pierre Ossman [Wed, 30 Dec 2020 15:13:52 +0000 (16:13 +0100)]
Merge branch 'more_noie' of https://github.com/CendioOssman/noVNC

3 years agoUpdate conversion documentation for Node.js
Pierre Ossman [Wed, 30 Dec 2020 14:57:02 +0000 (15:57 +0100)]
Update conversion documentation for Node.js

We now only support conversion to CommonJS, in order to support Node.js
older than version 15.

3 years agoRemove documentation about converting the app
Pierre Ossman [Wed, 30 Dec 2020 14:56:42 +0000 (15:56 +0100)]
Remove documentation about converting the app

This is no longer possible as we now require browser support for
modules.

3 years agoUse snap actions instead of the broken container
Pierre Ossman [Wed, 30 Dec 2020 13:18:29 +0000 (14:18 +0100)]
Use snap actions instead of the broken container

The container didn't work properly for our base snap anyway.

3 years agoPackage files directly in snapcraft.yaml
Pierre Ossman [Wed, 30 Dec 2020 12:56:47 +0000 (13:56 +0100)]
Package files directly in snapcraft.yaml

We don't need to convert things anymore, so reference files directly in
the snap yaml file.

3 years agoStore result from NPM and snap builds
Pierre Ossman [Wed, 30 Dec 2020 12:40:55 +0000 (13:40 +0100)]
Store result from NPM and snap builds

To ease debugging.

3 years agoRun NPM and snap builds on every push
Pierre Ossman [Wed, 30 Dec 2020 12:40:03 +0000 (13:40 +0100)]
Run NPM and snap builds on every push

To make sure these things still build. That means we need to make the
actual deploy parts optional.

3 years agochore: sync with W3C documentation
Quentin Dreyer [Tue, 22 Dec 2020 09:50:23 +0000 (10:50 +0100)]
chore: sync with W3C documentation

https://www.w3.org/TR/uievents-key/

3 years agoMerge branch 'jp' of https://github.com/CendioOssman/noVNC
Pierre Ossman [Thu, 10 Dec 2020 09:23:10 +0000 (10:23 +0100)]
Merge branch 'jp' of https://github.com/CendioOssman/noVNC

3 years agoRemove some unnecessary use of done argument
Pierre Ossman [Thu, 10 Dec 2020 09:02:31 +0000 (10:02 +0100)]
Remove some unnecessary use of done argument

3 years agoFake key releases for some Japanese IM keys
Pierre Ossman [Thu, 10 Dec 2020 09:01:04 +0000 (10:01 +0100)]
Fake key releases for some Japanese IM keys

Windows behaves very oddly for some Japanese IM keys in that it won't
send a key release event when the key is released. In some keys it never
sends the event, and in some cases it sends the release as the key is
pressed the subsequent time.

3 years agoSend combination keysyms for some Japanese keys
Pierre Ossman [Thu, 10 Dec 2020 08:43:49 +0000 (09:43 +0100)]
Send combination keysyms for some Japanese keys

Windows doesn't give us stable symbols for a bunch of Japanese IM keys,
instead alternating between two symbols. This state is not synchronised
with the IM running on the remote server so to have stable behaviour we
have to collapse these multiple symbols in to a single keysym.

3 years agoFix typo for ZenkakuHankaku key
Pierre Ossman [Thu, 10 Dec 2020 08:43:08 +0000 (09:43 +0100)]
Fix typo for ZenkakuHankaku key

3 years agoUse toggle keysym for Eisu key
Pierre Ossman [Thu, 10 Dec 2020 08:42:19 +0000 (09:42 +0100)]
Use toggle keysym for Eisu key

This matches how the key behaves on a Linux system.

3 years agoMerge branch 'noie' of https://github.com/CendioOssman/noVNC
Pierre Ossman [Mon, 7 Dec 2020 13:30:05 +0000 (14:30 +0100)]
Merge branch 'noie' of https://github.com/CendioOssman/noVNC

3 years agoRemove createEvent() fallbacks
Pierre Ossman [Fri, 4 Dec 2020 16:03:57 +0000 (17:03 +0100)]
Remove createEvent() fallbacks

We can now rely on proper constructors for our events.

3 years agoRemove createImageData() fallback
Pierre Ossman [Fri, 4 Dec 2020 15:51:22 +0000 (16:51 +0100)]
Remove createImageData() fallback

All our browsers should be new enough now that we can rely on the
ImageData constructor.

3 years agoRemove many small, obsolete, old browser hacks
Pierre Ossman [Fri, 4 Dec 2020 15:44:54 +0000 (16:44 +0100)]
Remove many small, obsolete, old browser hacks

These are for browsers no longer supported anyway.

3 years agoUse Fetch API for getting JSON data
Pierre Ossman [Fri, 4 Dec 2020 15:43:44 +0000 (16:43 +0100)]
Use Fetch API for getting JSON data

We no longer need to support Internet Explorer so we can use a more
proper API here.

3 years agoRemove unused injectParamIfMissing()
Pierre Ossman [Fri, 4 Dec 2020 15:43:04 +0000 (16:43 +0100)]
Remove unused injectParamIfMissing()

It should have been removed in 58fc267b2bac1b89450b76a21e91ffd25e7e4358
with the caller.

3 years agoUpdate keycode mappings to latest version
Pierre Ossman [Mon, 16 Nov 2020 12:44:27 +0000 (13:44 +0100)]
Update keycode mappings to latest version

This update fixes Korean layouts.

3 years agoMerge branches 'logs' and 'timeout' of https://github.com/jcpunk/noVNC
Pierre Ossman [Tue, 10 Nov 2020 12:21:15 +0000 (13:21 +0100)]
Merge branches 'logs' and 'timeout' of https://github.com/jcpunk/noVNC

3 years agoMake timeouts/heartbeats easy to setup
Pat Riehecky [Mon, 9 Nov 2020 19:25:29 +0000 (13:25 -0600)]
Make timeouts/heartbeats easy to setup

3 years agoPermit setting syslog on websockify
Pat Riehecky [Thu, 29 Oct 2020 18:47:07 +0000 (13:47 -0500)]
Permit setting syslog on websockify

3 years agoRemove keypress handling
Pierre Ossman [Thu, 15 Oct 2020 15:19:20 +0000 (17:19 +0200)]
Remove keypress handling

We no longer support any browser that requires this legacy handling.

3 years agoDrop support for legacy Edge
Pierre Ossman [Thu, 15 Oct 2020 15:16:00 +0000 (17:16 +0200)]
Drop support for legacy Edge

3 years agoRemove legacy conversion of modules
Pierre Ossman [Thu, 15 Oct 2020 14:30:44 +0000 (16:30 +0200)]
Remove legacy conversion of modules

We no longer support Internet Explorer so we can now require that
browsers support modules.

Some conversion to commonjs still remains for nodejs.

3 years agoDrop support for Internet Explorer
Pierre Ossman [Thu, 15 Oct 2020 14:48:08 +0000 (16:48 +0200)]
Drop support for Internet Explorer

3 years agoInclude current websockify in noVNC snap
Pierre Ossman [Thu, 8 Oct 2020 14:55:12 +0000 (16:55 +0200)]
Include current websockify in noVNC snap

The one from the system is too old to support current versions of noVNC,
so we need to bundle our own.

3 years agoAdd workaround for Firefox PNG rounding bug
Pierre Ossman [Mon, 28 Sep 2020 10:24:56 +0000 (12:24 +0200)]
Add workaround for Firefox PNG rounding bug

3 years agoRemove Firefox Alt workaround
Pierre Ossman [Mon, 28 Sep 2020 09:12:16 +0000 (11:12 +0200)]
Remove Firefox Alt workaround

The bug got fixed way back in Firefox 63, and it is also misbehaving
with modern Firefox as they no longer consider AltGr an Alt-key.

3 years agoHandle empty rects in RAW decoder as well
Pierre Ossman [Mon, 7 Sep 2020 10:58:52 +0000 (12:58 +0200)]
Handle empty rects in RAW decoder as well

It was overlooked in the previous commit because we couldn't feed
empty data messages through the test framework.

3 years agoIgnore recording variables in our linter
Samuel Mannehed [Sat, 5 Sep 2020 09:04:11 +0000 (11:04 +0200)]
Ignore recording variables in our linter

3 years agoApproximate comparison of JPEG result
Pierre Ossman [Fri, 4 Sep 2020 14:48:44 +0000 (16:48 +0200)]
Approximate comparison of JPEG result

The browsers have various rounding errors so we need to compare the
output data only approximately and not exactly.

3 years agoHandle empty rects from the server
Pierre Ossman [Fri, 4 Sep 2020 14:16:44 +0000 (16:16 +0200)]
Handle empty rects from the server

These are very pointless for the server to send, but not a violation of
the protocol so we need to be able to handle them. We've seen this
happen in real world scenarios a few times.

3 years agoMerge branch 'rgbx' of https://github.com/CendioOssman/noVNC
Pierre Ossman [Fri, 4 Sep 2020 11:40:23 +0000 (13:40 +0200)]
Merge branch 'rgbx' of https://github.com/CendioOssman/noVNC

3 years agoFix call to _recvMessage()
Samuel Mannehed [Fri, 4 Sep 2020 08:38:23 +0000 (10:38 +0200)]
Fix call to _recvMessage()

Missed to rename this one in commit ea858bfa275d85541ac26d035967959714e8e255

3 years agoFix names of recording variables
Samuel Mannehed [Fri, 4 Sep 2020 08:33:35 +0000 (10:33 +0200)]
Fix names of recording variables

The name of these variables must match how they were set when the
recording was created.

Reverts part of 95632e413d75030c49577c0e5cf1003208a5ba09

3 years agonoVNC 1.2.0 v1.2.0
Pierre Ossman [Tue, 14 Jul 2020 08:13:20 +0000 (10:13 +0200)]
noVNC 1.2.0

3 years agoUpdate generated JS files for translations
Pierre Ossman [Tue, 14 Jul 2020 08:16:40 +0000 (10:16 +0200)]
Update generated JS files for translations

3 years agoUpdate Swedish translation
Samuel Mannehed [Wed, 8 Jul 2020 21:18:00 +0000 (23:18 +0200)]
Update Swedish translation

3 years agoMerge pull request #1435 from nnn1590/fix-japanese-translation
Samuel Mannehed [Wed, 8 Jul 2020 21:25:28 +0000 (23:25 +0200)]
Merge pull request #1435 from nnn1590/fix-japanese-translation

Fix Japanese translation

3 years agoExplicitly set NPM registry URL
Pierre Ossman [Tue, 7 Jul 2020 10:49:50 +0000 (12:49 +0200)]
Explicitly set NPM registry URL

Otherwise the action doesn't configure npm properly to authenticate and
our publish actions will fail with 404 errors.

3 years agonoVNC 1.2.0 beta
Pierre Ossman [Tue, 7 Jul 2020 08:48:24 +0000 (10:48 +0200)]
noVNC 1.2.0 beta

3 years agoRemove note about clipboard ISO 8859-1 restriction
Pierre Ossman [Tue, 7 Jul 2020 08:38:06 +0000 (10:38 +0200)]
Remove note about clipboard ISO 8859-1 restriction

We now support full Unicode, provided the server also supports the
proper extension.

3 years agoUpdate translation template file
Pierre Ossman [Fri, 3 Jul 2020 14:11:53 +0000 (16:11 +0200)]
Update translation template file

3 years agoMerge branch 'deploy' of https://github.com/CendioOssman/noVNC
Pierre Ossman [Fri, 3 Jul 2020 14:06:08 +0000 (16:06 +0200)]
Merge branch 'deploy' of https://github.com/CendioOssman/noVNC

3 years agoPublish pre-releases to beta channels
Pierre Ossman [Fri, 3 Jul 2020 14:00:04 +0000 (16:00 +0200)]
Publish pre-releases to beta channels

Extra important for NPM which doesn't allow replacing a release once it
is published.

3 years agoMention gestures in README
Samuel Mannehed [Sun, 28 Jun 2020 21:01:23 +0000 (23:01 +0200)]
Mention gestures in README

3 years agoDeploy new snap on release
Pierre Ossman [Fri, 26 Jun 2020 12:12:37 +0000 (14:12 +0200)]
Deploy new snap on release

3 years agoMove snap specific script to snap directory
Pierre Ossman [Fri, 26 Jun 2020 12:11:55 +0000 (14:11 +0200)]
Move snap specific script to snap directory

3 years agoRemove redundant VERSION file
Pierre Ossman [Fri, 26 Jun 2020 11:25:25 +0000 (13:25 +0200)]
Remove redundant VERSION file

We want to avoid having this in multiple places.

3 years agoDon't detach handler if it doesn't exist
Pierre Ossman [Thu, 25 Jun 2020 12:45:58 +0000 (14:45 +0200)]
Don't detach handler if it doesn't exist

If the beforeEach() step has been skipped then there won't be anything
here to detach.

3 years agoHandle quick Cursor detach after mouse up
Pierre Ossman [Thu, 25 Jun 2020 12:37:21 +0000 (14:37 +0200)]
Handle quick Cursor detach after mouse up

This timer might fire after the Cursor object has detached from a DOM
element, causing crashes. This will likely not happen in real scenarios,
but the tests are quick enough to trigger this.

3 years agoUpdate build badges for GitHub actions
Pierre Ossman [Wed, 24 Jun 2020 11:02:03 +0000 (13:02 +0200)]
Update build badges for GitHub actions

3 years agoMerge branch 'actions' of https://github.com/CendioOssman/noVNC
Pierre Ossman [Wed, 24 Jun 2020 10:50:31 +0000 (12:50 +0200)]
Merge branch 'actions' of https://github.com/CendioOssman/noVNC

3 years agoUse GitHub actions instead of Travis/Sauce Labs
Pierre Ossman [Tue, 23 Jun 2020 11:56:33 +0000 (13:56 +0200)]
Use GitHub actions instead of Travis/Sauce Labs

The ability to use Sauce Labs for pull requests has now stopped working,
as Travis warned about several years ago. Instead run our tests directly
on GitHub on their various virtual machines.

3 years agoRemove default Karma options
Pierre Ossman [Tue, 23 Jun 2020 11:54:46 +0000 (13:54 +0200)]
Remove default Karma options

No need to mention things where we already use the default value.

3 years agoAlway focus on touchstart
Pierre Ossman [Tue, 16 Jun 2020 13:32:38 +0000 (15:32 +0200)]
Alway focus on touchstart

The new gesture detection code will always prevent the default behaviour
of touchstart, so this check no longer works properly. We might want to
add something similar to GestureHandler in the future, but let's wait
and see what use cases are requested.

3 years agoCompensate for visual viewport when moving cursor
Pierre Ossman [Tue, 16 Jun 2020 12:24:00 +0000 (14:24 +0200)]
Compensate for visual viewport when moving cursor

3 years agoFix Japanese translation
NNN1590 [Mon, 15 Jun 2020 08:44:33 +0000 (17:44 +0900)]
Fix Japanese translation

3 years agoMerge pull request #1414 from CendioOssman/gesture
Samuel Mannehed [Fri, 12 Jun 2020 14:11:30 +0000 (16:11 +0200)]
Merge pull request #1414 from CendioOssman/gesture

Add gesture handling

3 years agoFake cursor position when using touch
Samuel Mannehed [Thu, 11 Jun 2020 13:52:34 +0000 (15:52 +0200)]
Fake cursor position when using touch

With the new gestures we will simulate the cursor being in a different
location than any of the touch points. This is a bit too complex for the
Cursor class, so let's just explicitly tell it where we want the cursor
rendered.

3 years agoMove mouse event handling to RFB class
Pierre Ossman [Wed, 10 Jun 2020 14:13:03 +0000 (16:13 +0200)]
Move mouse event handling to RFB class

Move the last remaining bits to the RFB class to keep things simple, as
the Mouse class no longer provides any real value.

3 years agoIncrease wheel step threshold
Pierre Ossman [Wed, 10 Jun 2020 12:12:22 +0000 (14:12 +0200)]
Increase wheel step threshold

The previous value made the detection too sensitive and it was very
difficult to scroll precisely. A value of 50 pixels should give similar
behaviour to systems that don't do fine grained scrolling.

3 years agoMove wheel event handling to RFB class
Pierre Ossman [Wed, 10 Jun 2020 11:59:10 +0000 (13:59 +0200)]
Move wheel event handling to RFB class

The Mouse class does very little now so it mostly just obfuscate things.
Move everything directly in to the RFB class instead.

3 years agoRemove very legacy mouse event handling
Pierre Ossman [Wed, 10 Jun 2020 10:56:50 +0000 (12:56 +0200)]
Remove very legacy mouse event handling

This is only needed on such ancient versions of Internet Explorer that
it wouldn't satisfy our other requirements anyway.

3 years agoRemove delayed wheel timer
Pierre Ossman [Wed, 10 Jun 2020 10:55:27 +0000 (12:55 +0200)]
Remove delayed wheel timer

This isn't really expected behaviour from a user, i.e. that an extremely
small wheel movement still gives a large scroll event in the remote application.

3 years agoAdd lint rule for function declaration indentation
Samuel Mannehed [Wed, 10 Jun 2020 08:23:50 +0000 (10:23 +0200)]
Add lint rule for function declaration indentation

3 years agoAdd touch gestures for mouse emulation
Pierre Ossman [Mon, 18 May 2020 13:17:06 +0000 (15:17 +0200)]
Add touch gestures for mouse emulation

Add several single and multitouch gestures to simulate various mouse
actions that would otherwise be impossible to perform.

This replaces the old system where you could select which mouse button
a single touch would generate.

3 years agoStart fake test clock at real clock time
Pierre Ossman [Thu, 11 Jun 2020 14:43:38 +0000 (16:43 +0200)]
Start fake test clock at real clock time

Some code relies on the clock having a somewhat sane value, so let's not
start at 0.

3 years agoAvoid fractional pixel sizes from Display
Pierre Ossman [Thu, 11 Jun 2020 14:31:09 +0000 (16:31 +0200)]
Avoid fractional pixel sizes from Display

3 years agoAvoid printing the Websock buffer in tests
Pierre Ossman [Mon, 8 Jun 2020 13:35:28 +0000 (15:35 +0200)]
Avoid printing the Websock buffer in tests

It takes forever and just messes up the output.

3 years agoSwitch to RGBx pixel format
Pierre Ossman [Sun, 7 Jun 2020 12:22:07 +0000 (14:22 +0200)]
Switch to RGBx pixel format

This is what the browser wants so it avoids having to spend time
converting everything. Unfortunately it usually means the server instead
needs to convert it for us, but we assume it has more power than we do.

3 years agoStandardise on a single blit function
Pierre Ossman [Sun, 7 Jun 2020 11:44:20 +0000 (13:44 +0200)]
Standardise on a single blit function

Keep everything simpler by always blitting in the same pixel format.
It's up to the decoders to convert if they need to.