If you had asked me how much it cost to drive a round-trip from Charlottetown, PEI to Halifax, Nova Scotia, I would have guessed about $80. Google Maps estimates the drive is about 3hr 30m.

If I thought about it a bit more, I probably could have come up with a more accurate estimate. A recent short road trip on this route gave me the exact numbers, and I thought I’d share.

This table shows the basic costs of driving from Charlottetown to Halifax (and back) on January 27, 2020 in Canadian dollars, including tax. The car was a Mazda CX-3 with all-wheel drive rented from Budget.

I didn’t include the cost of the car rental from Budget (two days for $109.26) or the taxi ride to the car-rental lot ($17). Though I suppose it’s technically possible to make the trip without Junior Mints, they are very refreshing.

Idea for the nerdiest TV show ever: A high-budget HBO-style series about an alternate history where the only difference is that Apple bought BeOS instead of NeXT.

We tried being kind — and it helped.

I loved the CBC radio show Wiretap. Sadly, the CBC website has gone through enough redesigns in the years since its broadcast that there doesn’t seem to be a canonical website for Wiretap anymore, hence the Wikipedia link.

Jonathan Goldstein, host of Wiretap, is a Canadian national treasure, even if he now works for a US company in New York since acquired by a Swedish streaming conglomerate legally based in Luxembourg.

Goldstein now hosts the Heavyweight podcast, and it is a worthy spiritual successor to Wiretap.

I usually dislike the feeling of being aware of the writer when watching or listening to a narrative like a TV show, movie, or podcast. In fiction, I find this happens when the writer is clearly smarter than the character (or worse, wants you to know how smart they are*). When the cleverness of the author bleeds through, it breaks the four wall and takes you out of the moment. That said, in episode #31 of Heavyweight, Goldstein had a line that did shine a light on the cleverness of the writer, and I was 100% on-board:

There are no mobius-strip-shaped roads in a euclidean space, like Canada.

Jonathan Goldstein, in episode #31 of Heavyweight

If you liked Wiretap, you’ll like Heavyweight. If you hated Wiretap, you’ll hate Heavyweight.

* (I am smart)

When I was a kid I wanted to be an aerospace engineer (and later, a rock star). At the time, there was no Web for which to be a Web Designer.

I sometimes wonder what jobs will come in to being by the time my own kids have careers.

• Job that didn’t exist 30 years ago: Web Designer
• Job that didn’t exist 10 years ago: Host of live-stream of launches for private rocket company
• Job that doesn’t exist today (right?): Host of self-flying helicopter fails clip show.

Speaking of CSS, I’ve run into a CSS rendering issue and I’d like to know if I’m missing something, or if it’s a genuine bug.

In Safari version 13.0.2 (14608.2.40.1.2) as of October 31, 2019 (👻), if the line-height in a <button> tag is set to 1, then the descender of tech characters like j, g, or p get cut off. For example:

button { line-height: 1; }

Here’s a Codepen example you can try.

I can easily work around this by setting the line-height: normal; or to a larger value like 1.2, but I’d rather keep the tight line-height in this case.

We had an election here in Canada yesterday. The Elections Canada website has a page with the results. There’s a little in-line bar chart to help visualize the relative values.

What is generally considered a good practice in CSS (the code used the style and layout pages on the Web) is making 33.1% look bigger than 34.4%:

This issue is fairly simple. The width of the bar-charts are defined with the em unit. Both the Conservative 34.4% bar and the 33.1% Liberal bar are to be 3em wide (they’re rounding off for the small comparison).

The trouble comes in the way the winning party row is highlighted. The winning row is set to have a larger font size (120% the regular size). Since the em unit is relative to font size, the winning row chart gets distorted to look 20% larger.

The fix is easy – use rem units (which are relative to the base document font-size rather than the local font-size) rather than em units for the bar charts.

I’ve passed this on to Elections Canada, but wanted to post here because it’s in interesting little issue that could be easily misconstrued. I also wanted you to know how smart I am.

There’s a proportional representation joke in here somewhere…

I got to join Matt Rainnie this week on the CBC Radio Mainstreet program for the SpinTime DJ segment. I got to choose three songs, which is harder than it sounds. First, I had to get over myself and realize I didn’t have to take it as seriously as those who assembled the Voyager Golden Record. I then looked for three great songs that someone listening on the way home from work on Friday might not have otherwise heard, but might enjoy.

First, I selected Coax Me by Sloan from their 1994 album Twice Removed. I was 16 years old when this came out and this song and video are burned into my brain. It’s up there with Everything You’ve Done Wrong as my favourite Sloan songs.

Next, I selected A Lot to Be Loved by Mardeen from the 2008 album Read Less Minds. Some CBC listeners might know from Mo Kenney’s cover of their song Telephones from the same album. This is a great, simple, pop/rock song.

Finally, I chose Wordwide by Adam Again from the 1992 album Dig. This is one of the best songs I’ve ever heard. If you consider the ratio of how great a song is compared to how likely people are to have actually heard it, this becomes an even more extraordinary gem.

The segment airs this afternoon some time between 4-6PM on Mainstreet on CBC Radio One, Charlottetown. It aired on Friday, September 20th, and is now available online: Mainstreet PEI with Matt Rainnie – Sep. 20, 2019: Spin Time – Steven Garrity

Thanks to Matt for having me in – he’s truly gifted at the art of conversation for the radio.

Many of us in the technology world are beginning to grapple with the unintended consequences of the platforms and products we’ve been created. I’ve recently noticed a pattern with potential unintended consequences where software products leak information about your presence or actions to others. It may not be fair to use the term “leak”, as it’s intentional on the part of the products, but those using the software may be less aware.

In Google Docs (and Sheets, Slides, Draw, etc.), when you are viewing or editing a document, you can see the names and faces of those who are also viewing in real time.

The notes/wiki system Notion shows a similar name / face of others who are viewing pages.

The Slack group chat app shows others when you’re typing in a channel.

The Figma design tool shows the names (and cursors) of others viewing a shared document.

These indications of presence or activity are often helpful. Knowing where a person is focused when you’re working together can be a helpful proxy in lieu of body language. Knowing that someone is typing a response to your question helps make the asynchronous flow of a text-base conversation feel a bit more natural.

These features also have unintended side-effects. In the case of Google Docs, I may want to refer to a shared document, but not broadcast my reading. If you have a difficult situation with a co-worker and privately look up an internal company policy to see how best to address it, you probably don’t want to ‘run into’ that same co-worker in the “How to deal with jerks at work” Google Doc.

In the case of Slack’s indication that you’re typing, you may start a typing out a comment and decide the conversation is better off without your input. Only now, everyone knows you’re there – reading, and is wondering what you are about to say. By not saying anything explicit, you’ve still communicated something to the group (I was going to speak, but decided not to, I’m about to speak, My cat is walking across my keyboard).

It’s always easier to point out the problems with complex systems like these than it is to recognize what they do well. Slack made an improvement to allow you to preview a channel (a discussion room) without ‘joining’ (joining and leaving Slack channels is announced to those in the channel).

Before the change to allow a ‘silent’ preview of a channel, it was like you had to jump into a room full of people, announce your presence, see that you didn’t want to be there, and announce your departure.

Even now, there’s no way to silently leave a Slack channel. It is helpful to know when others leave a channel, but it also makes it socially difficult to leave channels. “Steven has left the #people-who-do-not-hate-dolphins channel” is a not a good look.

I’d like to be able to provide some concentrated rules-of-thumb for these situations, but they’re complex. These are the best conclusions I’ve been able to drawn so far:

• Let people know if, how, and when your software exposes a person’s activity or presence to others.
• Allow people to opt-out of features that broadcast their presence or activity.

This clever little widget called Good Fast Cheap by Kunuk Nykjær is a delightful illustration of how the classic project adage:

Good, fast, cheap–choose two

can be recast as:

Expensive, slow, shitty–choose one

Discovered via the often delightful CodePen Spark newsletter.