]> git.proxmox.com Git - mirror_novnc.git/log
mirror_novnc.git
4 years agoFix bug where inflate would read too much data
Niko Lehto [Tue, 4 Feb 2020 08:55:49 +0000 (09:55 +0100)]
Fix bug where inflate would read too much data

4 years agoSplit api of inflate
Niko Lehto [Mon, 3 Feb 2020 09:19:00 +0000 (10:19 +0100)]
Split api of inflate

Added ability to read data chunk wise.

4 years agoHandle errors from zlib/pako
Niko Lehto [Mon, 3 Feb 2020 09:04:20 +0000 (10:04 +0100)]
Handle errors from zlib/pako

4 years agoMove error handling to Inflate class
Niko Lehto [Mon, 3 Feb 2020 08:57:56 +0000 (09:57 +0100)]
Move error handling to Inflate class

Every call wants this check so this should be done inside the class.

4 years agoAdd missing copyright header for Inflator.js
Niko Lehto [Tue, 11 Feb 2020 13:20:56 +0000 (14:20 +0100)]
Add missing copyright header for Inflator.js

4 years agoRemove unused inflate argument
Niko Lehto [Mon, 3 Feb 2020 08:53:30 +0000 (09:53 +0100)]
Remove unused inflate argument

The value true was an invalid flush argument so it was in practice
unused.

4 years agoAdd util for unsigned and signed int. conversion
Niko Lehto [Tue, 28 Jan 2020 16:00:04 +0000 (17:00 +0100)]
Add util for unsigned and signed int. conversion

Will be used in later commit in extended clipboard handling.

4 years agoAdd deflator helper class for deflating data
Niko Lehto [Fri, 7 Feb 2020 12:23:21 +0000 (13:23 +0100)]
Add deflator helper class for deflating data

Wraps pako's deflate for easier usage.

4 years agoMake clipBoardPasteFrom() test more specific
Niko Lehto [Tue, 28 Jan 2020 09:01:54 +0000 (10:01 +0100)]
Make clipBoardPasteFrom() test more specific

Don't rely on clientCutText() to test clipboardPasteFrom().

4 years agoMake afterEach() hooks work when skipping tests
Pierre Ossman [Thu, 23 Jan 2020 13:27:37 +0000 (14:27 +0100)]
Make afterEach() hooks work when skipping tests

Mocha will now run afterEach() hooks when tests are skipped, so we need
to make them more robust against things being partially set up.

4 years agoRemove generated HTML by Cursor when it detaches
Juanjo Diaz [Thu, 23 Jan 2020 09:51:51 +0000 (11:51 +0200)]
Remove generated HTML by Cursor when it detaches

4 years agoMerge branch 'abstraction_for_detection' of https://github.com/samhed/noVNC
Pierre Ossman [Tue, 14 Jan 2020 08:45:28 +0000 (09:45 +0100)]
Merge branch 'abstraction_for_detection' of https://github.com/samhed/noVNC

4 years agoSimplify encodeUTF8/decodeUTF8 unittests
Samuel Mannehed [Fri, 3 Jan 2020 09:41:34 +0000 (10:41 +0100)]
Simplify encodeUTF8/decodeUTF8 unittests

4 years agoAdd unit tests for encodeUTF8 and decodeUTF8
Samuel Mannehed [Thu, 2 Jan 2020 16:29:41 +0000 (17:29 +0100)]
Add unit tests for encodeUTF8 and decodeUTF8

4 years agoRemove unused python scripts
Samuel Mannehed [Thu, 2 Jan 2020 12:56:07 +0000 (13:56 +0100)]
Remove unused python scripts

4 years agoFix chinese translation for "Disconnect"
Samuel Mannehed [Thu, 2 Jan 2020 12:24:42 +0000 (13:24 +0100)]
Fix chinese translation for "Disconnect"

Thanks for @wavezhang, @litongjava, and @bhzhu203 for helping out.

4 years agoUpdate chinese translation strings
Samuel Mannehed [Thu, 2 Jan 2020 12:18:24 +0000 (13:18 +0100)]
Update chinese translation strings

Thanks to @QQ2017 and @wavezhang for helping out with correcting these.

4 years agoAdd comment for browser and platform detection
Samuel Mannehed [Thu, 24 Oct 2019 14:43:07 +0000 (16:43 +0200)]
Add comment for browser and platform detection

4 years agoAdd short description at the top of browser.js
Samuel Mannehed [Wed, 23 Oct 2019 15:08:56 +0000 (17:08 +0200)]
Add short description at the top of browser.js

4 years agoProperly detect scrollbar gutter
Samuel Mannehed [Wed, 23 Oct 2019 15:09:41 +0000 (17:09 +0200)]
Properly detect scrollbar gutter

As a rule, instead of hard-coding a behavior on specific platforms we
should do dynamic detection.

This commit moves away from always hiding scrollbars on Android and iOS
and instead detects the rendered width of scrollbars in the browser.

4 years agoRemove Google Chrome Frame plugin compatability
Samuel Mannehed [Tue, 31 Dec 2019 01:15:07 +0000 (02:15 +0100)]
Remove Google Chrome Frame plugin compatability

Chrome Frame has been retired we so we can remove it from this
compatability tag.

4 years agoRestore X-UA-Compatible meta tag to vnc_lite
Samuel Mannehed [Tue, 31 Dec 2019 01:07:31 +0000 (02:07 +0100)]
Restore X-UA-Compatible meta tag to vnc_lite

It's still required since without this IE's default setting for "Display intranet web pages in compatibility mode" will cause errors.

4 years agoRestore X-UA-Compatible meta tag to vnc.html
Samuel Mannehed [Tue, 31 Dec 2019 01:06:02 +0000 (02:06 +0100)]
Restore X-UA-Compatible meta tag to vnc.html

It's still required since without this IE's default setting for "Display intranet web pages in compatibility mode" will cause errors.

4 years agoAllow cursor to be updated while connecting
Pierre Ossman [Mon, 30 Dec 2019 08:30:00 +0000 (09:30 +0100)]
Allow cursor to be updated while connecting

We haven't got a server provided cursor at this point, but we might
have done something local, e.g. enabled the dot cursor.

4 years agoMake sure "undefined" can be a default parameter value
Pierre Ossman [Wed, 25 Dec 2019 11:10:21 +0000 (12:10 +0100)]
Make sure "undefined" can be a default parameter value

Lower layers can consider null to be a valid value, when we'd rather
they treat the value as not set.

4 years agoHandle slow loading of images
Pierre Ossman [Mon, 23 Dec 2019 14:40:17 +0000 (15:40 +0100)]
Handle slow loading of images

Internet Explorer seems to flag images as loaded prematurely, which
can result in rendering bugs. We can detect this by looking at the
dimensions though.

4 years agoValidate decoded image dimensions
Pierre Ossman [Mon, 23 Dec 2019 14:37:48 +0000 (15:37 +0100)]
Validate decoded image dimensions

They are expected to be a certain size, so verify this so no server
tries to do something broken.

4 years agoAvoid using Array.includes()
Pierre Ossman [Mon, 23 Dec 2019 14:11:32 +0000 (15:11 +0100)]
Avoid using Array.includes()

Internet Explorer doesn't have this method. Use the safer indexOf()
instead.

4 years agoBuild in the behavior to ignore decodeUTF8 errors
Samuel Mannehed [Mon, 23 Dec 2019 09:25:02 +0000 (10:25 +0100)]
Build in the behavior to ignore decodeUTF8 errors

Makes the code clearer and more explicit in intent.

4 years agoUse undefined as the default value for password
Pierre Ossman [Mon, 23 Dec 2019 08:35:51 +0000 (09:35 +0100)]
Use undefined as the default value for password

An empty password is techincally legal, and now supported by the
RFB core, so we cannot use that as a placeholder for "no password".

4 years agoPeter is no longer part of the noVNC team
Samuel Mannehed [Fri, 20 Dec 2019 23:26:57 +0000 (00:26 +0100)]
Peter is no longer part of the noVNC team

4 years agoMerge pull request #1327 from vanym/decodeUTF8_try_catch
Samuel Mannehed [Sat, 7 Dec 2019 19:12:27 +0000 (20:12 +0100)]
Merge pull request #1327 from vanym/decodeUTF8_try_catch

Fixes exception when desktop name contains non-utf8 character

4 years agoAdd try catch in every place that uses decodeUTF8
VanyM [Thu, 5 Dec 2019 19:36:51 +0000 (22:36 +0300)]
Add try catch in every place that uses decodeUTF8

4 years agoAdd support for Unix Tight auth
Chris "Koying" Browet [Sat, 3 Aug 2019 07:11:25 +0000 (09:11 +0200)]
Add support for Unix Tight auth

4 years agoExpand password dialog to work for usernames too
Chris Koying Browet [Thu, 5 Dec 2019 14:34:18 +0000 (15:34 +0100)]
Expand password dialog to work for usernames too

Some VNC authentication schemes use usernames, our UI should support
these.

4 years agoMerge pull request #1318 from chrisjdev/emptyPassword
Samuel Mannehed [Thu, 5 Dec 2019 11:15:43 +0000 (12:15 +0100)]
Merge pull request #1318 from chrisjdev/emptyPassword

Allow connecting with empty string credentials

4 years agoAllow connecting with empty string credentials
Chris J [Thu, 7 Nov 2019 20:15:15 +0000 (15:15 -0500)]
Allow connecting with empty string credentials

Checking for undefined instead of falsy. That way an empty strings are allowed.

4 years agoAdd encodeUTF8 function to core/util/strings.js
Samuel Mannehed [Wed, 4 Dec 2019 09:13:25 +0000 (10:13 +0100)]
Add encodeUTF8 function to core/util/strings.js

4 years agoMake Cursor.detach() safe to call when not attached
Pierre Ossman [Fri, 29 Nov 2019 09:08:15 +0000 (10:08 +0100)]
Make Cursor.detach() safe to call when not attached

Avoids having checks in higher layers.

4 years agoFix chinese translation errors
Samuel Mannehed [Thu, 28 Nov 2019 11:51:32 +0000 (12:51 +0100)]
Fix chinese translation errors

Thanks to @QQ2017 for providing the translation strings

4 years agoSimplify logic for status hierarchy
Samuel Mannehed [Mon, 25 Nov 2019 16:03:55 +0000 (17:03 +0100)]
Simplify logic for status hierarchy

Removes unnecessary variable

4 years agoEnsure warning status timeouts are honored
Samuel Mannehed [Mon, 25 Nov 2019 15:38:03 +0000 (16:38 +0100)]
Ensure warning status timeouts are honored

When showing a new status popup we want to set a timer for how long to
show it. In cases where we show many statuses in a fast succession we
need to remove any running timeouts when showing a new one.

There are exceptions when new statuses won't be shown, and thats if a
more severe status is already showing, i.e and error or a warning.

Warnings can still have timeouts. There was a bug that occured when we
tried to show a normal status while a warning was showing. The bug
caused the warning status timeout to be removed even if the normal
status was never shown. We should only remove running timeouts if we're
actually going to show a new status.

4 years agoMerge branch 'babel' of https://github.com/CendioOssman/noVNC
Pierre Ossman [Tue, 12 Nov 2019 13:12:18 +0000 (14:12 +0100)]
Merge branch 'babel' of https://github.com/CendioOssman/noVNC

4 years agoDon't use arrow functions in legacy loader
Pierre Ossman [Tue, 12 Nov 2019 09:25:18 +0000 (10:25 +0100)]
Don't use arrow functions in legacy loader

The browsers that need the legacy code do not support such fancy
modern things.

4 years agoLoad support scripts first
Pierre Ossman [Tue, 12 Nov 2019 09:24:40 +0000 (10:24 +0100)]
Load support scripts first

E.g. SystemJS requires the Promise polyfill, so make sure all our
support files are loaded first.

4 years agoUpgrade to latest babel
Pierre Ossman [Mon, 11 Nov 2019 14:46:12 +0000 (15:46 +0100)]
Upgrade to latest babel

There has been a lot of renaming and restructuring in babel, so we need
to modify our code to handle the latest version. We also need to adjust
the way we build our babel worker as babel itself no longer runs in older
browsers such as Internet Explorer.

4 years agoMerge pull request #1319 from wavezhang/patch-2
Samuel Mannehed [Mon, 11 Nov 2019 15:46:36 +0000 (16:46 +0100)]
Merge pull request #1319 from wavezhang/patch-2

Update zh_CN.po

4 years agoAlways include Promise polyfill for legacy browsers
Pierre Ossman [Mon, 11 Nov 2019 12:36:30 +0000 (13:36 +0100)]
Always include Promise polyfill for legacy browsers

It is now used by our general code and not just by the conversion
routines, so we need to make sure it is always included for the
old browsers.

4 years agoClean up handling of untransformed files
Pierre Ossman [Mon, 11 Nov 2019 12:33:47 +0000 (13:33 +0100)]
Clean up handling of untransformed files

This control flow is difficult enough as it is to follow. Move the
handling of the untransformed files to a separate block to make it
slightly easier to understand.

4 years agoRemove redundant "no copy" check
Pierre Ossman [Mon, 11 Nov 2019 12:32:19 +0000 (13:32 +0100)]
Remove redundant "no copy" check

We also check this list as a filter to walkDir(), so no need for the
extra check here.

4 years agoUse proper backticks for generating legacy script tags
Pierre Ossman [Mon, 11 Nov 2019 09:01:43 +0000 (10:01 +0100)]
Use proper backticks for generating legacy script tags

4 years agoUpdate zh_CN.po
wavezhang [Fri, 8 Nov 2019 06:08:23 +0000 (14:08 +0800)]
Update zh_CN.po

fix spell errors, improve translations

4 years agoFix missing caps lock events on iOS
Alex Tanskanen [Thu, 7 Nov 2019 15:19:08 +0000 (16:19 +0100)]
Fix missing caps lock events on iOS

Caps Lock on iOS only trigged key release or key press events.
When it's clicked it would only send keydown, and next time
it would only send keyup and so on. It should send both a key press
and a key release.

Also added the unit tests for macOS since those were missing.

Co-Authored-By: Alex Tanskanen <aleta@cendio.se>
4 years agoAdd README to app/locale warning not to modify
Samuel Mannehed [Thu, 7 Nov 2019 11:49:41 +0000 (12:49 +0100)]
Add README to app/locale warning not to modify

4 years agoAdd "macOS shuffle" to iOS as well
Alex Tanskanen [Mon, 4 Nov 2019 13:22:46 +0000 (14:22 +0100)]
Add "macOS shuffle" to iOS as well

Since iOS functions like macOS with regards to Alt behaving like AltGr,
we need the same workaround on iOS as well.

4 years agoHandle missing Shift events on Windows
Pierre Ossman [Mon, 4 Nov 2019 09:17:45 +0000 (10:17 +0100)]
Handle missing Shift events on Windows

This is a bug in the OS that leaks through to the browsers. We need
to fake a Shift release here to avoid Shift getting stuck in the remote
session.

4 years agoSend NumLock on macOS, even though the key is Clear
Pierre Ossman [Fri, 1 Nov 2019 12:13:35 +0000 (13:13 +0100)]
Send NumLock on macOS, even though the key is Clear

There is no obvious choice what works best here, but this is what
TigerVNC has been doing for years without complaints. Let's follow
them until we get reports that this doesn't work well.

4 years agoFix typo for MailSend key
Pierre Ossman [Fri, 1 Nov 2019 09:49:19 +0000 (10:49 +0100)]
Fix typo for MailSend key

4 years agoUpdate to latest UI Events key specification
Pierre Ossman [Fri, 1 Nov 2019 09:25:23 +0000 (10:25 +0100)]
Update to latest UI Events key specification

4 years agoFix AltGr for a few more keys in IE and Edge
Pierre Ossman [Fri, 1 Nov 2019 08:59:02 +0000 (09:59 +0100)]
Fix AltGr for a few more keys in IE and Edge

Some keys apparently send 'Unidentified' rather than an unshifted value.
Make sure those are also handled. Examples are \ and | on a Swedish
keyboard.

4 years agoKeep the virtual keyboard after using extra keys
Samuel Mannehed [Thu, 31 Oct 2019 15:24:06 +0000 (16:24 +0100)]
Keep the virtual keyboard after using extra keys

If using the extra keys always gives focus to the screen then an
on-screen keyboard would be closed. When using on-screen keyboards we
instead want to give focus to our virtual keyboard input element.

4 years agoRevert iOS keyup workaround
Pierre Ossman [Thu, 31 Oct 2019 14:36:40 +0000 (15:36 +0100)]
Revert iOS keyup workaround

It seems Apple has fixed their buggy keyup events, so remove the
workaround and allow keys to be kept pressed again.

This is a revert of 9e99ce126ca8f6f350fa015c0e58d35103c62f7e.

4 years agoHandle broken numpad delete key in Chrome
Pierre Ossman [Thu, 31 Oct 2019 13:51:36 +0000 (14:51 +0100)]
Handle broken numpad delete key in Chrome

4 years agoExplain why Clear maps to KP_Begin
Pierre Ossman [Thu, 31 Oct 2019 13:08:36 +0000 (14:08 +0100)]
Explain why Clear maps to KP_Begin

4 years agoTry to handle Meta key properly
Pierre Ossman [Thu, 31 Oct 2019 13:07:30 +0000 (14:07 +0100)]
Try to handle Meta key properly

The standards have unfortunatly caused some confusion between the Windows
key and the original Meta key. Try to handle the common case sanely at least.

4 years agoMerge pull request #1312 from samhed/master
Samuel Mannehed [Mon, 28 Oct 2019 09:45:08 +0000 (10:45 +0100)]
Merge pull request #1312 from samhed/master

Fix so that you can use the keyboard after using the extra keys

4 years agoMove focus to the screen when using extra keys
Samuel Mannehed [Fri, 25 Oct 2019 15:44:20 +0000 (17:44 +0200)]
Move focus to the screen when using extra keys

A regression from 2afda54 and friends was that you couldn't use the
extra keys and then directly use the keyboard, you would have to click
in the session first.

This commit restores the correct behavior and also adds a visual queue
to the fact that the screen got the focus by fading the controlbar.

4 years agoAdd missing scancode for sendTab
Samuel Mannehed [Mon, 28 Oct 2019 08:59:58 +0000 (09:59 +0100)]
Add missing scancode for sendTab

4 years agoRename document.capturedElem to captureElement
Samuel Mannehed [Wed, 23 Oct 2019 13:59:43 +0000 (15:59 +0200)]
Rename document.capturedElem to captureElement

To better fit most naming.

4 years agoImprove Windows key image
Pierre Ossman [Wed, 23 Oct 2019 13:51:41 +0000 (15:51 +0200)]
Improve Windows key image

The previous one didn't have sharp lines, or follow pixel boundaries
properly.

4 years agoMerge pull request #1309 from samhed/disappearing_cursor
Samuel Mannehed [Mon, 21 Oct 2019 12:16:48 +0000 (14:16 +0200)]
Merge pull request #1309 from samhed/disappearing_cursor

Fix disappearing cursor

4 years agoHide the emulated cursor when target is null
Samuel Mannehed [Mon, 21 Oct 2019 09:27:43 +0000 (11:27 +0200)]
Hide the emulated cursor when target is null

Makes it easier to understand what happens when a real element isn't
passed as a target to updateVisibility(). Also makes the code more
robust to future changes.

Co-authored-by: Alex Tanskanen <aleta@cendio.se>
Co-authored-by: Niko Lehto <nikle@cendio.se>
4 years agoFix disappearing cursor after click
Samuel Mannehed [Mon, 21 Oct 2019 09:26:40 +0000 (11:26 +0200)]
Fix disappearing cursor after click

In the cursor emulation when deciding if the cursor should be hidden -
Instead of checking what's under the cursor, we check the element that
has capture.

This introduced another bug in the cursor emulation. The cursor did not
always disappear properly when using our cursor emulation together with
our setCapture polyfill. More specifically, we saw a problem when a
capture ended on an element without cursor emulation.

We solved this by introducing another visibility check on a timer in
the cursor emulation. However this led to yet another problem where
this timer conflicted with the timer in the setCapture polyfill.

We removed the timeout in the setCapture polyfill and created a
variable to make sure that all the events remaining in the queue can be
completed.

Co-authored-by: Alex Tanskanen <aleta@cendio.se>
Co-authored-by: Niko Lehto <nikle@cendio.se>
4 years agoCheck next elem at mouseleave in cursor emulation
Samuel Mannehed [Mon, 21 Oct 2019 09:25:01 +0000 (11:25 +0200)]
Check next elem at mouseleave in cursor emulation

It's not obvious that we want to hide the cursor when we get a leave,
it depends on the element that we're leaving to. This makes the code
more robust.

Co-authored-by: Alex Tanskanen <aleta@cendio.se>
Co-authored-by: Niko Lehto <nikle@cendio.se>
4 years agoRename variables in setCapture proxy
Samuel Mannehed [Mon, 21 Oct 2019 09:23:21 +0000 (11:23 +0200)]
Rename variables in setCapture proxy

The names of many variables were too similar. To make the code easier
to follow we renamed:

* _captureElem to _capturedElem
* _captureElemChanged() to _capturedElemChanged()
* captureElem to proxyElem
* elem to target

Co-authored-by: Alex Tanskanen <aleta@cendio.se>
Co-authored-by: Niko Lehto <nikle@cendio.se>
4 years agoFix alt text for drag button
Pierre Ossman [Mon, 14 Oct 2019 08:17:44 +0000 (10:17 +0200)]
Fix alt text for drag button

We had left an old placeholder text on this button.

4 years agoNever show drag icon if clipping is disabled
Pierre Ossman [Mon, 14 Oct 2019 08:15:19 +0000 (10:15 +0200)]
Never show drag icon if clipping is disabled

Toggling the enabled state is a remnant from an earlier version
of the code where we could determine if the the session is actually
clipped, and not just that the setting is enabled.

Right now we only change things based on the setting, so let's
completely hide the button when clipping is disabled.

4 years agoUpdate copyright to 2019 for modified files
Samuel Mannehed [Mon, 30 Sep 2019 13:35:33 +0000 (15:35 +0200)]
Update copyright to 2019 for modified files

4 years agoCode comments for how the receieve queue works
Samuel Mannehed [Wed, 25 Sep 2019 19:04:46 +0000 (21:04 +0200)]
Code comments for how the receieve queue works

4 years agoClarify why we ENABLE_COPYWITHIN is false
Samuel Mannehed [Wed, 25 Sep 2019 19:03:37 +0000 (21:03 +0200)]
Clarify why we ENABLE_COPYWITHIN is false

4 years agoMerge pull request #1298 from CendioNiko/edgeCursorURI
Samuel Mannehed [Wed, 25 Sep 2019 12:46:58 +0000 (14:46 +0200)]
Merge pull request #1298 from CendioNiko/edgeCursorURI

Fix url cursor detection on Edge

4 years agoMerge pull request #1299 from CendioNiko/vmwarecursor
Samuel Mannehed [Wed, 25 Sep 2019 12:45:29 +0000 (14:45 +0200)]
Merge pull request #1299 from CendioNiko/vmwarecursor

Add support for VMware cursor encoding

4 years agoMerge pull request #1289 from CendioNiko/master
Samuel Mannehed [Wed, 25 Sep 2019 12:43:55 +0000 (14:43 +0200)]
Merge pull request #1289 from CendioNiko/master

Desktop name improvements

4 years agoMerge pull request #1281 from jalfd/optimize-receive-buffer
Samuel Mannehed [Wed, 25 Sep 2019 11:26:02 +0000 (13:26 +0200)]
Merge pull request #1281 from jalfd/optimize-receive-buffer

Optimize receive buffer

4 years agoClarify comments for broken alt in FF on Windows
Samuel Mannehed [Tue, 24 Sep 2019 14:00:05 +0000 (16:00 +0200)]
Clarify comments for broken alt in FF on Windows

4 years agoMerge pull request #1280 from jalfd/alt-key-recursion
Samuel Mannehed [Tue, 24 Sep 2019 13:56:13 +0000 (15:56 +0200)]
Merge pull request #1280 from jalfd/alt-key-recursion

Avoid recursion in Alt check on Firefox

4 years agoMerge pull request #1279 from CendioOssman/nomodule
Samuel Mannehed [Tue, 24 Sep 2019 13:40:05 +0000 (15:40 +0200)]
Merge pull request #1279 from CendioOssman/nomodule

Use "nomodule" instead of manual check

4 years agoUse "nomodule" instead of manual check
Pierre Ossman [Fri, 16 Aug 2019 09:18:45 +0000 (11:18 +0200)]
Use "nomodule" instead of manual check

Very few browsers are left in the wild that supports modules but not
"nomodule", so let's simplify our handling a bit.

Safari 10 supports modules but not 'nomodule', this means that this
particular version of Safari will be broken. Due to this we have to
bump up the required Safari version to 11.

4 years agoAdd support for VMware cursor encoding
Niko Lehto [Tue, 17 Sep 2019 07:28:30 +0000 (09:28 +0200)]
Add support for VMware cursor encoding

Supports both classic cursor type and alpha cursor type. In classic
mode the server can send 'inverted' pixels for the cursor, our code
does not support this but handles these pixels as opaque black.

Co-authored-by: Samuel Mannehed <samuel@cendio.se>
4 years agoFix url cursor detection on Edge
Niko Lehto [Mon, 23 Sep 2019 08:50:45 +0000 (10:50 +0200)]
Fix url cursor detection on Edge

_supportCursorURIs was set to true even when Edge didn't support
URIs because the fallback value "default" was used.

4 years agoRestore page name after disconnect
Niko Lehto [Wed, 4 Sep 2019 13:05:14 +0000 (15:05 +0200)]
Restore page name after disconnect

4 years agoTest unicode desktop names
Niko Lehto [Wed, 4 Sep 2019 12:47:40 +0000 (14:47 +0200)]
Test unicode desktop names

4 years agoAdd support for DesktopName extension
Niko Lehto [Wed, 4 Sep 2019 12:23:37 +0000 (14:23 +0200)]
Add support for DesktopName extension

This extension allows session name to be changed during runtime.

4 years agoSet viewport size for autoscale tests
Pierre Ossman [Fri, 23 Aug 2019 13:48:30 +0000 (15:48 +0200)]
Set viewport size for autoscale tests

We were incorrectly relying on the viewport being indirectly set
for us. Make sure we are explicit in what we want for these tests.

4 years agoMerge branch 'upgrade' of https://github.com/CendioOssman/noVNC
Pierre Ossman [Fri, 23 Aug 2019 13:08:56 +0000 (15:08 +0200)]
Merge branch 'upgrade' of https://github.com/CendioOssman/noVNC

4 years agoRemove pointless Display.clear()
Pierre Ossman [Fri, 23 Aug 2019 13:05:58 +0000 (15:05 +0200)]
Remove pointless Display.clear()

It served no meaningful purpose and it had bugs. So let's remove it
rather than try to fix it.

4 years agoRemove unused Display.logo attribute
Pierre Ossman [Fri, 23 Aug 2019 13:03:13 +0000 (15:03 +0200)]
Remove unused Display.logo attribute

4 years agoDeprecate showDotCursor option for RFB constructor
Pierre Ossman [Fri, 23 Aug 2019 11:58:40 +0000 (13:58 +0200)]
Deprecate showDotCursor option for RFB constructor

It is not relevant for the connection stage so it should not have
been a constructor argument to begin with. Ship with a warning for
a release before we remove it.

4 years agoMake sure showDotCursor can be modified before connecting
Pierre Ossman [Fri, 23 Aug 2019 11:57:30 +0000 (13:57 +0200)]
Make sure showDotCursor can be modified before connecting

The cursor object is only attached to our canvas whilst connecting,
so we need to make sure we don't try to update anything when were
not connected or we'll get a crash.

4 years agoRegenerate module loader after tool upgrades
Pierre Ossman [Thu, 15 Aug 2019 13:39:35 +0000 (15:39 +0200)]
Regenerate module loader after tool upgrades