Adam Panzica

I make cool things with Robots

[Adam Panzica @aruisdante]

World Photography Day Giveaway 2023 https://t.co/1h9xRf8HA5

[Adam Panzica @aruisdante]

@kevinrschultz @GergelyOrosz https://t.co/tOEzPZ3AXJ

[Adam Panzica @aruisdante]

@leobicknell @GergelyOrosz That’s not actually true; customer engagement drops dramatically if app performance is reduced. When I worked at Uber, the metrics on falloff for a 1 second increase in home load were pretty staggering. Customers also tend to delete apps that hog battery, which was also tracked.

[Adam Panzica @aruisdante]

@GergelyOrosz @ChasanisNickos This doesn’t mean I think it’s a good model. It’s just the model many of these legacy companies are used to, and it’s very, very difficult to break them out of it, even if they want to change to become more “Silicon Valley like” because they look at Tesla and get very afraid.

[Adam Panzica @aruisdante]

@GergelyOrosz @ChasanisNickos Sometimes this segmentation is purely so that “chargeback” can happen. That is, the subsidiary licenses the technology to the parent, and the parent pays them (or internally license to each other). That accounting model doesn’t work if all the IP is accessible to everyone.

[Adam Panzica @aruisdante]

@GergelyOrosz Basically, I agree completely: AI “code review” is actually linting, not code review. It’s certainly extremely valuable, but serves a very different function than what could review _should_ be serving in a healthy org with good devex.

[Adam Panzica @aruisdante]

@GergelyOrosz In addition, when onboarding new team members we would explicitly include them on all reviews in the team for a period of time, both to bring them up to speed, and give “fresh eye” challenges where they would see code smells we had become nose blind to.

[Adam Panzica @aruisdante]

@BriannaWu I think the GT4 is better looking then the GT3, and I own a GT3. Bros gonna bro.

[Adam Panzica @aruisdante]

@DropTheDie At the end of the day design resources are limited, so you spend them on the swath of gameplay most people spend the most time in.

With 10 levels, you’re dropping the tutorial+endgame, so you’re overwhelming newbies/boring vets. Or nothing changes except the numbers skipped.

[Adam Panzica @aruisdante]

@DropTheDie This is true, but you see the same trend in video game RPGs as well. The tutorial pre title-card roll is rarely inspired (and often skippable!), and the “end game content” is usually just grindy big numbers content for “challenge.”

[Dwarven Forge @dwarvenforge]

The Kickstarter closes tomorrow at 9 PM ET.

If you are still unsure of what you want to order but know you want to back the campaign, pledge before the KS closes.

You can adjust your order later in the Pledge Manager to suit your taste!

[Adam Panzica @aruisdante]

@LaurenWalshArt And thankfully we’re in the US, with a robust derivative work protection. This isn’t the case everywhere. In Japan for example the author retains significantly more control, so fan art would in fact be legally enforceable as copyright infringement.

[Adam Panzica @aruisdante]

@FaynaSorcerer @DnDBeyond @dimension20show @quiddie Yeah. I remember when NewEgg was doing that random drawing each day to combat scalpers. People kept saying “person X as one three times and I haven’t won any! That’s not random at all!” And it’s like… actually that’s exactly what random means.

[Adam Panzica @aruisdante]

@MackmyW @DnDBeyond @dimension20show @quiddie So put those two properties together, and you have strong evidence that the system is in fact a uniform random distribution.

[Adam Panzica @aruisdante]

@MackmyW @DnDBeyond @dimension20show @quiddie In a truly random system, a perfect uniform distribution can’t exist unless your sample size is infinity. But the data shows that on any given day the distribution is _close_ to uniform, and across days there are no correlations in its non-uniformity (it’s randomly non-uniform).