julianlam

This is a living document, it will be updated from time to time as my test writing evolves.

It's easy to fall into the trap of writing your tests with less care than you would your regular code. After all, if tests pass, then ship it is de rigeur, so there's an incentive — especially if the clock is ticking closer to the end of workday — to get those tests passing ASAP.

Avoid globals/leaky states

One corner that is often cut is to introduce or reuse variables between tests. For example:

1. I am writing tests for updating a user's profile. I will create one user and run all tests against that single user.

julianlam

Sometimes you just need to prune a directory of all files older than a certain age. Think log files, or uploaded files... maybe ~/Downloads, perhaps?

find . -mindepth 1 -maxdepth 1 -mtime +90 -delete

• mindepth and maxdepth stop find from crashing if there are too many files
• mtime specifies instructs find to filter the files by the last modified time. +90 means 90 days
• delete is self-explanatory

julianlam

Those that know me know that I don't exactly have a great relationship with neat cabling practices. My desktop PC builds are rather haphazard, with power cables, SATA cables, and the like snaking across and through the case with no regard for aesthetics. After all, plug A needs to go to socket A, why not take the most direct route? Brace yourselves...

Accordingly, the same work ethic seemed to follow me into my home networking setup. Through a combination of organic growth (new devices added over time), and due to the added incentive of needing to get connected ASAP when we moved in, little to no thought was put to making everything look neat and tidy.

I'll spare you the imagery, but some lovely faults:

1. A DSL line that entered the house via a demarcation point at one end of the house, only to snake across the entire length of the house before actually entering

julianlam

I came across a scenario where I had a certificate (example.org, below) that supported multiple domains, but one domain's (broken.example.org, below) the A record no longer existed, causing certbot to abort renewals with an error:

Saving debug log to /var/log/letsencrypt/letsencrypt.log

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Processing /etc/letsencrypt/renewal/example.org.conf
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Cert is due for renewal, auto-renewing...
Plugins selected: Authenticator webroot, Installer None

julianlam

Working with <input type="datetime-local"> is a lot of fun. All major browsers have standardized their UIs, and offer full support, and so the days of using polyfills or jQuery fallbacks are numbered.

Should you need support for older or more esoteric browsers, it may be best to fall back to a bunch of <select> inputs instead.

The cool thing about datetime-local is that the UI itself is localised based on the user's browser/machine language settings.

For example, in Canada:

julianlam

I have a not-so-secret confession to make — I use the self-checkout line at grocery stores because they get me out of the store faster.

The choice used to be simple: at any given checkout queue, the queue for the self-checkout is usually non-existant, and I am adept enough at the interface so that it is not overly cumbersome to use.

My so-called "Speed Hacks"

1. If you take care to cart your items with all of the barcodes facing upwards, then at the scanner, you can use the hand-held scanner to scan everything without having to waste time turning the product across three dimensions trying to find the barcode¹. The time spent locating the barcode is spread out across my entire shopping trip, instead of at the till. If you use the "skip bagging" option, then you can even leave the items in the cart, instead of needlessly scanning/bagging/carting your items.

julianlam

Where I work, we run a bug bounty. We've done this for a number of years, and it's yielded great results in that bugs that could potentially affect user data are quickly and quietly dealt with at our own pace, and we retain control over the narrative (as opposed to having it disclosed publicly and dealing with the fallout.)

By far, however, the most annoying part of running the bug bounty program are the wannabe security researchers.

You know the kind;

• The ones who run automated scanners against every website they can find
• The ones who copy and paste technical jargon from websites in an attempt to feign legitimacy
• The ones who assign a "high" priority to their reports via subject—just to get your attention
• The ones who, when pressed for details, have no clue what they're talking about or even reporting

julianlam

I often have to toggle between a number of GMail accounts because even though they are managed by Zendesk, occasionally ham is caught by GMail's spam filter and never makes it to Zendesk. I've had enough important messages fall through that I became suspicious, and then discovered that I have replies and incoming leads being caught almost weekly!

Anyways, in short, GMail takes bloody forever to load anything that isn't Chrome, and the UI is not very snappy. Clicking "Use Basic HTML" at the bottom left is pointless because it doesn't actually trigger right away, and I've already waited many seconds before it sends me to that interstitial page asking me if I really wanted to use Basic HTML. Oh the horror.

In the future, just append ?ui=html or /h/, and voila!, watch your mailbox load in .5 seconds, flat.

So in the future, accessing your main mailbox would be https://mail.google.com/mail/u/0/h/. Note the trailing slash, that's needed.

I really miss when webs

julianlam

This is how to do it, because I have the memory of a goldfish and have to look it up every time.

The context

Let's say there's a server (123.123.123.123) running a service on port 1234.

You want to view it on your local machine on port 2345.

The command

julianlam

I came across an interesting problem without a good native browser solution, recently. Usually, browser APIs are so complete that this is a non-issue, but nonetheless...

The specific scenario involved a multiselect:

<select multiple id="mySelect">
  <option value="1">One</option>
  <option value="2">Two</option>
  <option value="3">Three</option>
 Four