They recently uploaded a Youtube video, offering transparency on things they’ve been working on in the background or haven’t panned out. Framework Laptop 16 - Prototypes and scrapped ideas
I’ve noticed an ongoing pattern of Framework prematurely adopting inflexible designs that fit some original use-case, and then being constrained by backwards compatibility to that inflexible design, which leads to ugly workarounds or limitations in future products.
From memory, I believe this has happened 3 times, which sounds like a lot.
Ethernet Expansion Card. Framework had to release an expansion card which extends past the envelope of the laptop chassis as a compromise.
Trackpoint. According to the above video, Trackpoints (a mouse navigation stick located on a keyboard) can act as a pressure point which will result in the screen cracking if pressure is applied to the top of the laptop when closed, which implies either a lack of rigidity in the top cover, or not enough clearance between the screen and the keyboard, or the keyboard was made too thin to allow a trackpoint to recess properly
GPU Module. According to the Q&A, larger expansion models seem to be off the cards, limiting our current batch to the 8GB RTX 5070, which many people I’ve asked seem to feel “mid” on–not very competitive with gamer brands
Sad thing is we might be seeing a fundamental reason why companies don’t prioritize backwards compatibility much anymore. Every decision has an opportunity cost, and making everything swappable even multiple generations down the line limits what you can do, as you’re constrained by the initial design.
I’m super happy the 16 got some updates, but I’m wondering how the GPU challenge in particular can be overcome. Really seems like to get bigger cards in there, they’re going to have to offer a refreshed 16 design at some point.
You missed at least one: the full-sized SD-card reader couldn’t fit into the defined expansion-card slots with the available electronics, they had to create custom electronics for it.
I’m not a hardware engineer, but from four decades of software development, I can tell you that it’s devilishly tricky to account for all of the future uses you’ll want to put something to. I’ve had to modify or rewrite perfectly working code more times than I can count, just to allow it to handle use-cases that no one predicted when we originally designed the code.
To have managed to develop a hardware design that only had three/four items that weren’t perfect fits on the first try is actually very good. If they’d waited until the design was absolutely perfect before releasing it, we’d all still be waiting for the original FW13.
Darn right. I can spend a week (or a month!) brainstorming future needs for an API resource and two months after I’ve published V1 I’ll find cases unaccounted for. More brains might help with better foresight but we still can’t predict the future.
I actually want the haptic touchpad module. I don’t click mine at all, I tap for single click and two finger tap for right click. My keyboard is in the middle too, so I wouldn’t rest my palms on it at all.
