Remove the help text under the sidebar subscribe button and style it inline

## What does this MR do?

Remove the help text under the sidebar subscribe button and style it inline

## Are there points in the code the reviewer needs to double check?


## Why was this MR needed?
Current help text is over obvious and pushes references too far down. Also the subscribe button is now consistently styled and not a complete block button anymore. approved by @awhildy 

## Screenshots (if relevant)

before:
![image](/uploads/0c9543d2bf184a73b65773caa5f9fc62/image.png)

after:
![image](/uploads/765338823376267427f4a16737e39998/image.png)


## Does this MR meet the acceptance criteria?

- [x] [Changelog entry](https://docs.gitlab.com/ce/development/changelog.html) added
- [ ] [Documentation created/updated](https://gitlab.com/gitlab-org/gitlab-ce/blob/master/doc/development/doc_styleguide.md)
- [ ] API support added
- Tests
  - [ ] Added for this feature/bug
  - [x] All builds are passing
- [x] Conform by the [merge request performance guides](http://docs.gitlab.com/ce/development/merge_request_performance_guidelines.html)
- [x] Conform by the [style guides](https://gitlab.com/gitlab-org/gitlab-ce/blob/master/CONTRIBUTING.md#style-guides)
- [x] Branch has no merge conflicts with `master` (if it does - rebase it please)
- [x] [Squashed related commits together](https://git-scm.com/book/en/Git-Tools-Rewriting-History#Squashing-Commits)

## What are the relevant issue numbers?


Closes #24281

See merge request !7389

Resolve "Improper use of bootstrap media query variables"

## What does this MR do?

Corrects media queries which are off by `1px` due to incorrect `$screen-{xs,sm,md,lg}-min` usage.

## Are there points in the code the reviewer needs to double check?

Ensure none of these errors were somehow intentional (I can't imagine a legitimate use case for `@media (max-width: $screen-sm-min)` though)

## Why was this MR needed?

When "min" breakpoints are used for "max-width" media queries, they create an overlap in rule breakpoints.  For instance in the following scenario:

```scss
.foo {
  @media (min-width: $screen-sm-min) {
    width: 100px;
    padding: 0;
  }
  @media (max-width: $screen-sm-min) {
    padding: 10px 0;
  }
}
```

If the viewport were set to exactly `$screen-sm-min` pixels, both of these media queries would apply.  It is important to use `max-width: $screen-xs-max` in this case which is equal to `$screen-sm-min - 1`.

## Does this MR meet the acceptance criteria?

- [ ] ~~[Changelog entry](https://docs.gitlab.com/ce/development/changelog.html) added~~
- [ ] ~~[Documentation created/updated](https://gitlab.com/gitlab-org/gitlab-ce/blob/master/doc/development/doc_styleguide.md)~~
- [ ] ~~API support added~~
- Tests
  - [ ] ~~Added for this feature/bug~~
  - [ ] All builds are passing
- [x] Conform by the [merge request performance guides](http://docs.gitlab.com/ce/development/merge_request_performance_guidelines.html)
- [x] Conform by the [style guides](https://gitlab.com/gitlab-org/gitlab-ce/blob/master/CONTRIBUTING.md#style-guides)
- [x] Branch has no merge conflicts with `master` (if it does - rebase it please)
- [x] [Squashed related commits together](https://git-scm.com/book/en/Git-Tools-Rewriting-History#Squashing-Commits)

## What are the relevant issue numbers?

Closes #25198

See merge request !7863

[ci skip]

Resolve "missing design for wiki page"

### To Do:

- [x] Move page navigation from sub-menu to sidebar menu
- [x] Update wiki page header
- [x] Update wiki edit page actions and header
- [x] Update `Git access` page design
- [ ] ~~Add search to sidebar navigation~~

### Screen Shots:

#### `@media (min-width: $screen-lg)`

![Screen_Shot_2016-11-16_at_12.25.47_PM](/uploads/2efc59639ac53d69808e78a8ba65fcbf/Screen_Shot_2016-11-16_at_12.25.47_PM.png)  
![Screen_Shot_2016-11-16_at_12.25.36_PM](/uploads/bb7e93558335edf4216082cc1ac3f8f5/Screen_Shot_2016-11-16_at_12.25.36_PM.png)  
![Screen_Shot_2016-11-16_at_12.26.43_PM](/uploads/c2dce7cff05680b9a03438ec7d9967b4/Screen_Shot_2016-11-16_at_12.26.43_PM.png)  
![Screen_Shot_2016-11-16_at_12.26.05_PM](/uploads/7dbae25bfc93fb2f01cd231db0cb47d3/Screen_Shot_2016-11-16_at_12.26.05_PM.png) 

#### `@media (min-width: $screen-md)`

![Screen_Shot_2016-11-16_at_12.28.18_PM](/uploads/4fdcc8c88c38093ea5a4fe0b22b79acf/Screen_Shot_2016-11-16_at_12.28.18_PM.png)  
![Screen_Shot_2016-11-16_at_12.28.21_PM](/uploads/ce84a51fdc0623b2b192d8334c2c2f6d/Screen_Shot_2016-11-16_at_12.28.21_PM.png)  

#### `@media (min-width: $screen-sm)`

![Screen_Shot_2016-11-16_at_12.28.05_PM](/uploads/cf612a36259c248ca597a0850546fe75/Screen_Shot_2016-11-16_at_12.28.05_PM.png)  
![Screen_Shot_2016-11-16_at_12.28.08_PM](/uploads/ae4374f9968c1cc064242fde3ad0e214/Screen_Shot_2016-11-16_at_12.28.08_PM.png)  

## Does this MR meet the acceptance criteria?

- [x] [Changelog entry](https://docs.gitlab.com/ce/development/changelog.html) added
- [x] [Documentation created/updated](https://gitlab.com/gitlab-org/gitlab-ce/blob/master/doc/development/doc_styleguide.md)
- Tests
  - [x] All builds are passing
- [x] Conform by the [merge request performance guides](http://docs.gitlab.com/ce/development/merge_request_performance_guidelines.html)
- [x] Conform by the [style guides](https://gitlab.com/gitlab-org/gitlab-ce/blob/master/CONTRIBUTING.md#style-guides)
- [x] Branch has no merge conflicts with `master` (if it does - rebase it please)
- [x] [Squashed related commits together](https://git-scm.com/book/en/Git-Tools-Rewriting-History#Squashing-Commits)

## What are the relevant issue numbers?

Closes #18546

See merge request !7429

Fix GitHub importer to import PR where source repo/fork was renamed/deleted

Closes #24594

See merge request !7865

Fix URL rewritting in the Help section

Closes #25199

See merge request !7875

Replace static fixture for shortcuts_issuable_spec

## What does this MR do?

Replace `issuable.html.haml` for `shortcuts_issuable_spec.js` by already existing dynamically generated fixture.

## What are the relevant issue numbers?

#24753

See merge request !7685

Make deleting with optimistic locking respect NULL

Make deleting with optimistic locking respect NULL

For now deleting with optimistic locking is broken when
lock_version is still NULL, because Rails would try to
delete with `lock_version = 0` while in the database
the column is still `NULL`.

The monkey patches would force Rails just pass whatever
in the column, and stop Rails from casting `NULL` into `0`
when the value is read from database.

Closes #24766

See merge request !7867

Resolve "Merge request dashboard page takes over a minute to load"

See merge request !7760

Gracefully recover from Redis connection failures in Sidekiq initializer

Precompiling assets and other Rake tasks would fail if Redis were not available in certain cases.

* Closes gitlab-org/gitlab-ce#25143
* Closes gitlab-org/omnibus-gitlab#1743

See merge request !7835

Signed-off-by: Rémy Coutable <remy@rymai.me>

Replace references to MergeRequestDiff#commits with st_commits when we care only about the number of commits

See merge request !7668

* Closes gitlab-org/gitlab-ce#25143
* Closes gitlab-org/omnibus-gitlab#1743

Document the need to use a PAT with Registry when 2FA is on

Fixes https://gitlab.com/gitlab-org/gitlab-ce/issues/22465

See merge request !7874

GitLab 8.12 introduced a new permissions model which tightened the
security of Container Registry. It is now required to use a personal
token if 2FA is enabled.

[ci skip]

Instead of doing n queries for n states, do one query to get all the
counts grouped by state, and figure out what the count is for each state
is from that. We can still cache the individual counts (it can't hurt),
but this will help with initial load.

Note that the `opened` scope on `Issuable` includes the `opened` and
`reopened` states, which is why there's a special case.

`any?` on an AR relation performs a `SELECT COUNT`, which we don't need.

1. We are very likely to have issues or MRs, so the `SELECT COUNT` is
   often unnecessary.
2. Even where there are no items returned, the overhead of the
   `SELECT *` instead of `SELECT COUNT` is relatively small.

Calling `to_a` on the relation lets us use `Enumerable#any?`, which will
return immediately if there are objects returned.

Small refactor of webhooks docs

See merge request !7872

[ci skip]

Clarify that a commit SHA doesn't work for CI triggers

Related https://gitlab.com/gitlab-org/gitlab-ce/issues/22029

See merge request !7871

[ci skip]

Fix wrong link in builds artifacts admin docs

Fixes https://gitlab.com/gitlab-org/gitlab-ce/issues/23168

See merge request !7870

when we care only about the number of commits

We do not have to instantiate all objects in this case.

[ci skip]

Only include EmailHelpers in mailer specs and specs using them

See merge request !7836

Remove an order dependency in LabelsFinder specs

Makes the LabelsFinder spec expectations order-independent

There is an explicit ordering to `LabelsFinder` - title ascending - but two of the labels (a group label and a project label) have the same title, so the sort ordering isn't stable for those in all circumstances. It looks like this is intentional.

See merge request !7788

Clean up JiraService

Fixes https://gitlab.com/gitlab-org/gitlab-ce/issues/24967

Fixes https://gitlab.com/gitlab-org/gitlab-ce/issues/24777

See merge request !7756

Improve the `Gitlab::OAuth::User` error message

## What does this MR do?

Fixes a logging issue submitted by a customer at https://gitlab.zendesk.com/agent/tickets/50728. 

The error saving the user is logged to application.log. Previously,
the entry had no context and was unusable - `Error saving user:
[Email address already taken]`. Adding the auth hash UID and email
makes the error more helpful.

For such a small logging change, do we need a changelog entry?

See merge request !7851

The implicit interface of project services states that the "execute"
method is meant to be called when project hooks are executed.
Currently JiraService does not support any project events even though
JiraService#supported_events says that "commit" and "merge_request"
are supported. They are only used to render correct options in
JIRA configuration screen, but they are not supported.
Because of that, this commit makes "execute" method a no-op.

For now deleting with optimistic locking is broken when
lock_version is still NULL, because Rails would try to
delete with `lock_version = 0` while in the database
the column is still `NULL`.

The monkey patches would force Rails just pass whatever
in the column, and stop Rails from casting `NULL` into `0`
when the value is read from database.

Closes #24766

Refactor branch chooser in issuable form

Continuation of #23864.

Sister of gitlab-org/gitlab-ee!930.

See merge request !7852