* lingohub:
  Created and pushed by LingoHub. Project: 'Rocket.Chat' by User: 'gabriel.engel@gmail.com'.

Conflicts:
	packages/rocketchat-livechat/app/i18n/en.i18n.json

This is a good start. LIke noted by @litewhatever there are limitations that related to the way meteor works. I can see some workarounds we could do, but that would need to be discussed and maybe submitted back to MDG. 

Setting allows admin to modify Meteor.loginTokenExpires from default 90 days to something else.
Server restart is required for this setting to take effect.
Default value is still 90 days if not modified by admin.

Fixed indentation and added unicode support

Fix 'create new' in private group list opening 'create channel' flex

Show toastr messages on the left side

Option to turn off livechat

Rocket.Cat Shouldn't be Admin, allowing first users to be admins again

The avatar image move need to only be done if a previous username exists. 
Also we need `encodeURIComponent` to support names in unicode.

fix broken image-link when og:image contains "&" (e.g. Google Maps)

Avatars for Unicode usernames

When a new avatar is saved the user name gets encoded using `encodeURIComponent`, this does not affect names in latin characters, however it does change non-latin/unicode names to an encoded string.

However, when reading the avatar image file, the function uses the username directly without enocinding it, so latin names work since the encoding doesn't change anything, however, Unicode names don't since the usernames don't match the encoded names.

This fix adds the encoding function when reading the avatars.
Also adds the encoding to the function that writes avatars from URL.

Livechat survey

Set code blocks to always be LTR globally

Fix for mentioning RTL names

Livechat popout

Conflicts:
	packages/rocketchat-livechat/app/client/views/room.coffee

The current @mention (for user names and channel names) uses the regex `\b` to detect the name boundaries, however that token doesn't support unicode characters. This solution replaces the word boundary with regex that match end of the word with any of the following white space, colon, comma and dot.

However, ending a name with a dot still has a problem as it doesn't get a mention link and this fix does not fix it. Part of the problem is that dots are allowed in the name validation regex. So this problem need to be solved later.

Another thing that this fix adds is a description to the UTF8 Name Validation settings, to inform chat admins about the characters that they need to avoid using in their regexes to not break other chat functionalities that depends on special characters to denote special meanings.

Livechat manager