HN user

sathomasga

158 karma
Posts3
Comments77
View on HN

Nor do most folks using it. Which isn't totally novel: I have almost no idea how an internal combustion engine works. But I do know enough to realize that it's not appropriate to drive my car on a bike path.

Survey from a company that's in the business of AI coding and thus has a monetary interest in promoting the technology. No details on who conducted the survey (the company itself?) or how the 609 respondents were selected. If limited to the company's own customers, massive selection bias. The results may or may not reflect reality, but this "report" is just marketing bullshit.

I think Cory Doctorow described said eng manager as a "human crumple zone" that serves to absorb the blame for failures.

If the ads that Apple shows me in the App Store, Apple News, etc. are supposed to be personalized, then Apple is really, _really,_ *really* bad at personalization. They're clearly not extracting any meaningul private information (or public information, for that matter) about me at the moment.

It doesn’t. After a response, you can ask it for its sources, but it will just make stuff up.

Out of curiosity I asked it a microbiology question. It gave a reasonable but rather anodyne response. When I asked for the sources, it provided two plausibly sounding papers, with titles, authors, etc. neither of which actually exists.

Definitely!

In my home state of Georgia, the slate of electors pledged to Donald Trump and Mike Pence on the 2020 presidential election was: Joseph Brennan, Ken Carroll, Vikki Consiglio, Carolyn Fisher, Patrick Gartland, Gloria Godwin, David Hanna, Mark Hennessy, Susan Holmes, John Isakson, Cathleen Latham, Daryl Moody, C. J. Pearson, David Shafer, Shawn Still, and C. B. Yadav.

The individuals who signed the fake electoral certificate did not include John Isakson, Patrick Gartland. C. J. Pearson, or Susan Holmes. Instead, the signers included Burt Jones, John Downey, Brad Carver, and Mark Amick.

Those four individuals, at least, were fake.

As an EE undergrad 40 years ago, the title made me think of the programming languages taught in my courses at the time.

- FORTRAN (no surprise here, I was an EE major not CS)

- Pascal (this was the new kid on campus at the time, only the CS folks used it)

- MIX (Knuth's generic assembly language)

As someone who's more likely to cycle somewhere than to drive, I wish there was as much attention paid every time a human-piloted car did something like this. It happens much more frequently than most folks realize. I used to follow "bicycle" news stories in a news aggregation app, but had to stop because my timeline was filled with "bicyclist killed by driver" stories. There was almost always at least one every day.

I'm by no means a fan of self-driving cars, but despite tragedies like this one, I wouldn't be surprised if overall they're much safer than human pilots even for cyclists.

I merely cited this as one example of a confusing UI, but this response makes my point even stronger. For most applications, the normal user flow is from general (what the app calls Repository Manager, I guess) to specific (a particular repository). Well-designed user interfaces anticipate the typical user flow and give the user a chance to "go back." For example, a user may accidentally click on the wrong repository. That user likely wants to return the app to its previous state so as to make the correct choice.

Neither of these steps really does that. The first opens a command palette which does not have anything resembling a "go back" command. It does list "Recent Repositories," but there is no reason for a user to assume that "Recent" is the same as All the repositories listed on the first screen. The command palette also omits the folder hierarchy present on the first screen, and that hierarchy may be key to finding the desired repo. Opening a new tab is not at all intuitive for a user that wants to go back. And the most obvious result of the command is a new blank screen.

There a lots of things about Fork that look to be very well-done, and it appears to have a great feature set. But if I'm going to use a GUI instead of the command line for git, then I expect the app to be intuitive and easy to use. For me, Fork is not that. I appreciate that other users may have different opinions, and I do wish the developers great success with the project.

Same reaction. My Tower subscription is up for renewal next month, so I'm primed to give this a try. A cursory look, however, and I find the user interface to be baffling. As one example, on loading the app presents a list of repositories to manage. Select one and you get to interact with that particular repo. But, as far as I can tell, there's no simple way to go back to the repository list to select a different repo to manage. At least, no way other than closing the app and relaunching it. Maybe I'm just used to the way Tower presents things and Fork is simply different, but I felt totally lost in the app, unsure of what to do or how to accomplish non-trivial tasks. With Tower, by contrast, all the interactions seem intuitive and straightforward. Looks like Tower will be getting my money in June as scheduled.

But that’s easier said than done for a lot of people.

It's also very easy for some people, myself as an example. I took the standard career path, eventually serving as VP of Engineering and CTO for several startups. (None successful enough to give me FU money, alas.) Now, at age 60, I've gone back to being just a regular developer with a manager 30 years my junior. I love it. No time wasted on worthless meeting, corporate BS, etc. All I do all day is write code. That's what gives me joy, and I'm very fortunate to be able to do it.