In this example, we’re setting up a job, which will run at 08:00 am daily and clean rails logs

Step 1 – create a file with what we want to run (delete_logs.sh):

#!/bin/(shell) 

cd ~/code/rails_project
/Users/username/.rbenv/shims/bundle exec rake log:clear

Setp 2 – set up the cron scheduler:

$ env EDITOR=vim crontab -e

# Add the line to run the script at 8am daily:

0 8 * * * sh ~/scripts/delete_logs.sh >/tmp/stdout.log 2>/tmp/stderr.log

• http://exercism.io/ – exercises in many languages, including Ruby
• http://paircolumbus.org
• https://www.railstutorial.org/
• https://www.codecademy.com/
• http://docs.railsbridge.org/
• https://online.pragmaticstudio.com/ – online course (not free, but pretty good)

Fira Code – a pretty cool font with a set of ligatures for common programming multi-character combinations.

To install – download from https://github.com/tonsky/FiraCode.

Install the fonts from ttf folder

Pick your fonts in VS (Tools – Options – Fonts and Colors)

Sublime is a pretty nice editor for working with Elixir, but I just can’t get the autocomplete plugin to work. So, giving Atom a try.

Install the following packages:

atom-elixir – https://atom.io/packages/atom-elixir
language-elixir – https://atom.io/packages/language-elixir
linter (apm install linter) – https://atom.io/packages/linter
linter-elixirc – https://atom.io/packages/linter-elixirc
script – https://atom.io/packages/script

I use script to build within Atom (shift-ctrl-b). The only issue is that it seems to use the top-level folder of your project as a base path. So, if you try to do something like Code.load_file(“file.exs”) in one of your nested directories, it’ll try to load it from the top level directory.

Addendum for Mac
brew install elixir
brew install erlang

Set elixir path for autocomplete-elixir (if using Atom):
/usr/local/bin/elixir

Client-side parsing of unubtrusive jQuery validation parameters. Assuming we are setting up an unobtrusive validator named “validatorname” and want to pass “parameterfromserver” from the the server:

$.validator.unobtrusive.adapters.add('validatorname', ['parameterfromserver'], function(options) {
	options.rules['validatorname'] = {
		parameterfromserver: options.params['parameterfromserver']
	};
	options.messages['validatorname'] = options.message;
});
$.validator.addMethod('validatorname', function(value, element, parameters) {
	var hereIsOurParameter = parameters.parameterfromserver;
});

Visual Studio 2015 has built-in functionality to locate a currently open file in the Solution Explorer using a shortcut. No need for any extensions.

Go to Tools – Options – Environment – Keyboard.
Find SolutionExplorer.SyncWithActiveDocument.
Assign a shortcut.

Done.

If you have to type with one hand, AutoHotkey is a free option to turn your regular keyboard into a half mirror keyboard, where each key on one side is mirrored to the other (activate it by pressing Space key first).

Here’s a script, copied from AutoHotkey forum with slight modifications:

#SingleInstance, force
; Half-QWERTY: One-handed Typing - version 3a
; http://www.autohotkey.com/forum/viewtopic.php?p=228783#228783
;
; HalfKeyboard invented by Matias Corporation between 1992 and 1996
; Originally coded in AutoHotkey by jonny in 2004
; Many thanks to Chris for helping him out with this script.
; Capslock hacks and `~ remap to '" by Watcher
; This implementation was done by mbirth in 2007
;
; version 3a script, mod by hugov:
; 2008-10-31:
; - mixed with "Capitalize letters after 1 second hold" at request of Calibran
; http://www.autohotkey.com/forum/post-228311.html#228311
; just tested very briefly so try at your own peril :-)
;

Exceptionless is an awesome open source real-time error, feature, and log reporting solution that can be self-hosted for free (they provide a very reasonable pricing plans as well). Here’re some tips on self-hosting in the Windows environment.

Use these detailed instructions to set it up. Instructions are quite good. However, there are some details not mentioned in the guide which you could find helpful for self-hosting.

Elastic Search

When setting up Elastic Search, it’s a good idea to change the default location of the data folder. It can help avoid an accidental loss of your data when updating to the next version of Elastic Search. The data path can be set in the elasticsearch.yml – “path.data” property.

Rename your cluster name – “cluster.name” property. This can help prevent someone from accidentally joining your cluster.

To directly query Elastic Search, you can install Sense chrome plugin and run queries inside Chrome.

Another useful ES GUI plugin: elasticsearch-gui
To install, run cmd from ES installation bin folder: [~/elasticsearch] $ bin/plugin –install jettro/elasticsearch-gui
To access it, go to http://yourServer:9200/_plugin/gui/index.html#/dashboard (assuming default ES port 9200)

If you want to set up backups for ES, first set the repo location in elasticsearch.yml
path.repo: [“backupLocation”]

From Chrome, Sense plugin, run the following query to create a repo for your backups:

PUT /_snapshot/exceptionless_backup
{
  "type": "fs",
   "settings": {
        "compress" : true,
        "location": "backupLocation"
    }
}

When you want to create a snapshot, run:

PUT /_snapshot/exceptionless_backup/yourShapshotName?wait_for_completion=true

To delete an old snapshot:

DELETE _snapshot/exceptionless_backup/yourShapshotName

Removing Old Data
Exceptionless provides a job which will automatically remove data over a given retention period. But if you self-host, the retention period is indefinite. If you prefer to have Exceptionless remove the old automatically, you can modify the retention period directly in the Elastic Search. Run the following ES query to set retention period to 45 days:

POST /organizations-v1/organization/YourOrganizationId/_update '{ "doc" : 
{
    "doc" : {        
 "_source": {
     "retention_days": 45
   }
    }
}

You can use Elasticsearch-gui plugin to figure out your organization Id.

Weirdness with IE 11.
For some reason IE 11 did not want to auto refresh itself. So, you could add these headers at IIS level:
Cache-Control:no-cache, no-store
Pragma:no-cache
This comes at the price of more hits to your server. But I couldn’t figure out the IE issue.

Redis
If you use Redis (recommended), this is how you specify the connection string:

<add name="RedisConnectionString" connectionString="serverName:6379" />

The important part is that there is no http in the server name