Allen Pike

> they're just acknowledging the fact that 99% of users don't need to see 99% of the URLs they visit on a daily basis.

While we're talking about things we don't need, let's include this change.

The fact is, for most of the last two decades we've already had a UI where users who don't care to attend to the URL don't have to, and users who care to notice can. What does this add? Nothing. But it does take away some legibility for people who care, and discoverability for people who might learn to.

99% of people using web browsers really get no cues from the path? Cite please. URLs aren't high tech any more than the address to your house is, and my observation is that even non-developers who are simply experienced browsers pick up cues -- even from barely legible URLs mostly meant to be parsed by machines. You don't have to be a programmer to observe that typing a string in takes you to a page, or that the string changes when the browser loads a change, and put together the address correlation until you start to understand what a URL is without even really thinking about it.

Or at least, you wouldn't have to be a programmer to learn to make that connection based off of simple observation skills if we kept the current model. If we move to this new hide-the-URL UI, probably you would (self-fulfilling prophesy!).

And sure, the web has lots of URLs that don't provide a lot of easily parsed cues. In the interest of being a little less selective, though, let's look at a few…

Just remember to automate the manual process at some point.

from https://allenpike.com/2023/dialup-world-isp:

"1. When a customer submitted their signup form to Dialup World, they would add a row for that new customer to their 20,000-row Excel spreadsheet.

2.Then, they would put that signup form in a pile with all the other signup forms of customers who had signed up on that day of the month.

[billing steps removed]

5. When a customer called to cancel their service, they would stop that customer from being billed further by simply – I shit you not – finding that customer’s signup form and trashing it. No signup form in the pile, no more bills going out.

You gotta hand it to these people. The seminal startup essay “Do Things that Don’t Scale” didn’t come out until 2013, and these legends were doing things that didn’t scale way back in the late 90s. They got 20,000 signups billed with the process equivalent of a shoelace and a discarded sauce packet."

Assuming they distributed the signup forms evenly amongst the 20 working days of a month, that would mean the billing group would have 1,000 bills to handle each day. Slogging day in, day out, forever. Might as well have named the group/department Sisyphus.

This smaller notification size would look pretty terrible on 4" devices like the iPhone SE. In fact, it already looks kind of terrible in the mockup provided.

IMO Apple is long overdue admitting that putting navigation buttons on the top bar was a mistake. Fine in the era of the original iPhone, but as phones get progressively bigger and taller, less so. The swipe back gesture works in most situations, but I was a fan of Allen Pike's mockup of bottom navigation on the iPhone X (or, "Pro", at the time):

https://www.allenpike.com/2017/developing-for-iphone-pro/

Having navigation at the bottom is one of the few things Android got absolutely right at launch when compared to Apple. But back to notifications, Android isn't great either: if a new notification arrives while I'm swiping old ones away there's no consideration that I might not want to swipe away something I haven't even read yet, so it disappears immediately and with no way to get it back. Frustrating.

It is amazing how much the Apple focused bloggers resemble political spin doctors these days. Create a talking point full of pro-Apple assumptions such as iOS 7 is an resource intensive OS that only Apple could support (another currently popular talking point is "iOS 7 isn't flat, it has depth" as if other OSs don't). Of course Android mostly sells cheap phones, that is given. Doesn't matter what the actual sales figures are. Forget the old argument that Apple places battery life over silly resource intensive things like widgets. Hammer it home with multiple articles (this is essentially a rewrite of the popular article from yesterday: http://www.allenpike.com/2013/ios7-catch-me-if-you-can/) and back patting links and retweets. Charge people thousands of dollars to get into your inner circle of spin via sponsorships of your blog or podcast, rinse, repeat.

I was thinking about this when reading an article on HN a couple weeks ago, "LLMs Aren’t Just “Trained On the Internet” Anymore" [1] (HN discussion [2]).

The thesis was that it won't matter if LLMs exhaust the available corpus of text that humanity has generated thus far, because it will always be easy to pay humans to create an inexhaustible supply of more text.

The article suggested that it was all going to be PhDs, lawyers, and poets creating this training data, but that all sounded like hand-wavy BS to me. I assume that, if more text is needed to train LLMs in the future, the vast majority of it is going to be pennies-on-the-page, mechanical turk-type workers. And the LLMs are going to reflect that kind of input.

1. https://allenpike.com/2024/llms-trained-on-internet

2. https://news.ycombinator.com/item?id=40549021

I've been in a similar situation but can see the light at the end of the tunnel. Giving yourself time to feel shitty about a shitty situation is pretty normal, but if you like doing good work then your interest in the work will be motivating when you dig back into it.

Lots of new areas and ideas floating around right now, and it'd be worth looking around before setting your sights on an old career trajectory.

How common is it for recruiters to work with commercial real estate agents? It's weird how little I know about your industry.

From HN earlier, thought it was on topic: https://allenpike.com/2022/giving-a-shit