add x86_64 check in downloadScript

[deansheather]

Adds a check to the download script to make sure the output of uname -m is x86_64 as code-server doesn't have releases for other platforms. If the architecture isn't supported you get output similar to the following:

go run . --skipsync fuyumi
2019-05-27 16:15:49 INFO	ensuring code-server is updated...
++ uname -m
+ '[' i686 != x86_64 ']'
++ uname -m
Unsupported server architecture i686. code-server only has releases for x86_64 systems.
+ echo 'Unsupported server architecture i686. code-server only has releases for x86_64 systems.'
+ exit 1
2019-05-27 16:15:50 FATAL	error: failed to update code-server:
---ssh cmd---
ssh  fuyumi /bin/bash
---download script---
set -euxo pipefail || exit 1

[ "$(uname -m)" != "x86_64" ] && echo "Unsupported server architecture $(uname -m). code-server only has releases for x86_64 systems." && exit 1
pkill -f ~/.cache/sshcode/sshcode-server || true
mkdir -p ~/.local/share/code-server ~/.cache/sshcode
cd ~/.cache/sshcode
wget -N https://codesrv-ci.cdr.sh/latest-linux
[ -f ~/.cache/sshcode/sshcode-server ] && rm ~/.cache/sshcode/sshcode-server
ln latest-linux ~/.cache/sshcode/sshcode-server
chmod +x ~/.cache/sshcode/sshcode-server: exit status 1
exit status 1

Closes #92.

[frank-dspeed]

This is a valid Solution LGTM

[sreya]

Can't the same objective be achieved by adding a statement to the README?

[deansheather]

Well, we could add a warning to the README too. I think having the sanity check as well will prevent users from being confused with possibly cryptic error messages if they didn't entirely read the README though.

[sreya]

This is fine since we're only supporting one arch but that conditional statement will gradually get uglier as more architectures are supported (if that ever happens), seems cleaner to me to have the user refer to the README

[deansheather]

Added a remote server OS support paragraph in the README.