don-code

My read is that a lot of content - specifically of the infinite-scroll variety - is very intentionally not designed to be printed.

There are a number of Slack and Discord conversations I'd have liked to export off to PDF, for instance (where a solid 90% of my "prints" go). I'm reminded that one of Slack's boot messages used to be, "Consider the environment before printing this Slack."

Indeed! If they'd reached out and said "I need 128GB of RAM to do this work" (which, for reasons I won't get into, was the case), we'd have figured out a more tenable solution.

Unfortunately, that conversation never happened.

There was a point where a coworker came to my team, complaining that our policies had erased an entire week's worth of high-priority work on him.

We asked where this happened. He had done his work on an EC2 instance named `temp-worker-5`.

We asked where the files on that EC2 instance were; maybe we could cleverly get them back off the disk. They were in `/tmp`. We had it mounted on tmpfs (in memory), and we were running tmpreaper to clean out `/tmp` every Sunday.

When presented with the sheer absurdity of doing work on an instance called `temp`, out of a directory called `/tmp`, he backed down...

I fear that we wouldn't be so lenient (or rational) on my team had an AI agent done this - we'd have to enable the AI, after all.

Obviously experiences will vary, but I have now had three organizations express concern (though not to the point of telling me I must stop) that I use third-party Firefox derivatives. This started with Pale Moon and then continued with LibreWolf.

The concern is that they "aren't authorized apps". As I've read it, this is a proxy for "we aren't certain these apps can be trusted, but nobody ever got fired for pointing the finger at Mozilla or Google when something goes wrong, but would be fired if they pointed it at this one random guy in the Netherlands".

As a NoScript user, I see a lot of similar content served via websites as well.

The ideal case is that a website gracefully degrades with JS disabled. That's largely a pipe dream with React usage being what it is, but some sites manage it. Github is one who still does this well.

The nonideal case is that I'm presented with a misrendered site, but I can still see some semantic information that may, or may not, be what I was looking for. Gitlab's an example of this.

The worst case comes where the site white-screens and doesn't load _anything_ of use. Reddit is an example where it loads nothing but a loading image. My worst pet peeve is the Stack Overflow-style failure mode, though, which explicit says: "Enable JavaScript and cookies to continue.", rather than making any attempt to show any content at all, which is most similar to what the author here is describing.

I was briefly involved with WITR during my undergrad. Loved what they were doing, but at that time the station definitely bought into the hipster attitude, music you've never heard of and will never hear again.

The only band I can remember hearing on WITR, then hearing anywhere else, is Neon Indian. They also sponsored a concert with Viva Brother, which briefly got some airplay on other stations, in spite of never (to my knowledge) having played them.

Well, sort of. My radio presets back in the 2000s were 101.7 WFNX, 104.1 WBCN, 107.3 WAAF, 100.7 WZLX, and 92.9 WBOS. Of those stations, only WZLX remains.

The greats - in my opinion, WBCN and WFNX - are long off the air. They were powerhouses of culture, both local and national. WFNX was the first to play Nirvana's Nevermind; Bono of U2 once said that "...without BCN, we'd all be (expletive)'d, let's face it."

Oedipus, a well-known Boston DJ since the 70s (who got started at 88.1 WMBR, MIT's station) and program director at WBCN, said of it: "The Telecommunications Act of 1996 destroyed radio... Corporations were now allowed to own multiple stations in the same market and across the country. Stations were no longer programmed locally."

Meanwhile, WAAF went online-only in 2020. WBOS got delegated to an HD2 subchannel in 2004. Likewise, The Boston Globe picked up many of the radio personalities for the online-only RadioBDC, now Indie617, but that seemed dead by the mid-2010s. I hear the former WFNX DJs Henry Santoro and Julie Kramer may still have some limited presence on 89.7 WGBH.

I was recently happy to find Adam 12 (formerly of WFNX (formerly of WBCN)) still DJing on 105.7 WROR.

88.9 WERS (Emerson College's station) is doing a great job of carrying the helm of Boston music culture. Occasionally I'll also tune into 90.3 WZBC (Boston College's station), which is a more "traditional" and less "commercial" college radio station.

Just outside Boston are likewise some public and community stations with good programming - 88.3 WRPS (Rockland's public radio station) and 101.3 WFNX (Scituate's community station - love that they picked up the WFNX callsign!) are pretty good, if you're in range.

Let's hypothesize this is true.

I worked at a startup that had _money_ - a lot of it - flowing in. Stakeholders wanted to see growth, which meant features. They did not want to see us making platform improvements; those didn't show well to clients. Best throw money at the problem, rather than sacrifice the all-important features.

We had a number of architectural quirks that reflected this. One was that we ran our databases in AWS on ephemeral instance store. Yes, the kind that doesn't persist when you stop or lose the instance. The kind that you should never run a database on. We did it so that we could join them eight of them together in RAID 0 on the biggest instance types, and pull 400,000 IOPS to the disk that way. We paired this with hot spares of each shard, as well as hourly backups.

We did this to support a Rails / ActiveRecord data model, for data that should never have been part of a Rails data model. We had exactly one person working part-time on query optimization.

The maintenance burden was immense. The AWS bill, even more so. But again, we are hypothesizing that we could host all of this data on a laptop.

To an extent, maybe up until our thousandth customer or so, I actually think we could have. But we wouldn't have spent most of our time working on _features_ - we'd have instead spent most of our time trying to performance optimize. That would've meant the investors wouldn't have seen growth, which in turn would've highly impacted our success. We had a well defined domain, and optimizing the queries for performance was not it.

I agree with you in regard to paying attention while driving, but the perception issue was actually around "you can't be bothered to be at your desk for meetings, so we perceive you as lazy", not "you're a risk to those around you, so we perceive you as irresponsible".

I find driving to be one of the most useless ways of spending my time, and if it's for more than half an hour, I do try to figure out some way to increase the value of that time.

I have a weekly commitment that leaves me driving home (~40min) at 9pm, and I usually eat dinner (just a sandwich) while I drive. That also has the advantage of making it so that I'm not eating an hour before bed.

If I know that I need to call someone, I'll usually try to schedule that call while I'm driving. I used to take meetings while driving as well, though I stopped because it was perceived poorly by others.

What's sort of sad is that I can take public transit to all of my regular commitments, and that lets me keep doing something (reading, working, whatever). The schedules are poor, though, and they blow my commute times completely out of the water. For example, I've got a 5-7pm commitment that is a 15-minute drive one way, but if I wanted to go by bus, I'd have to leave at 3:30pm (latest it comes before I need to be there), and get back on it at 8pm (the earliest it comes after I'm done).

I agree with the author's premise - that one feedback loop optimizes for speed, and the other for scale - but I don't think the market is bearing the conclusion - that AI should be utilized to enable more rapid experimentation, where we better scale what works.

Many vendors seem to be learning (or not learning, but just throwing their weight against it anyway) that adding hastily-generated AI features are causing customer dissatisfaction, as more people brand the features "slop".

In the best case, the users give the company more chances. Infinitely more chances.

In a worse case, the users assume the new feature will always be bad, given their first impression. It's hard for a vendor to make people reconsider a first impression.

The absolute worst case is that AI enables a new market, but the first attempts are so poor that the first movers make people write that market off as a dead end, leading to a lost opportunity.

There was a point where two friends and I each lived alone in an apartment, and I was the only one who had a (2-door) car. We still occasionally did Costco runs.

We'd go in and walk the store - the whole store - aisle by aisle.

If I saw something like a 2-pound bag of tortellini, but thought two pounds was too big a quantity for me, I'd ask, "does anybody want to split two pounds or tortellini?" One might say yes, so we'd throw the tortellini in the shopping cart.

At the end, one person (the membership holder) would pay, and we'd divvy up the result of our haul into reusable containers, in the parking lot. One of us would then take point on itemizing the receipt, and we'd pay back the person with the membership.

In hindsight, I think we did this more to socialize than to save money, but we definitely did save money. Even as a single apartment-dweller, I bought my fair share of 24-packs of yogurt and 5-pound bags of frozen vegetables.

My experience has been that it's easy to say, "oh, it's just me", but much harder to subject someone you care about to the same standard that you would yourself. I'm in a similar position with the thermostat, even though something we initially bonded over was that we both kept our thermostats at a low temperature that was outside the window of being socially acceptable.