Hacker News Reader

Comment by marcodiego | Hacker News Reader

What is really needed is a common chassis. A common carcass that allows me the put whatever I want inside it. If I want a pinebook[1], I want to be able to put pinebook guts inside it, if I want a mnt reform[2], I want to be able to put mnt reform guts inside it.

Too bad eoma68[3] is still sci-fi.

[1] https://www.pine64.org/pinebook-pro/

[2] https://mntre.com/media/reform_md/2020-05-08-the-much-more-p...

[3] https://www.crowdsupply.com/eoma68/micro-desktop

View on HN

> What is really needed is a common chassis.

I agree. I understand why that wasn't a thing 10 years ago while laptops continued to get thinner and bezels continued to get smaller and I/O was rapidly changing and evolving. Modern laptops are much more consistent, and standards like M.2 and USB-C have provide excellent support for low-profile expansion. Now seems like a great time to start rolling out standards for laptop motherboard connectivity, display housing, keyboard/trackpad housing, and I/O bays.

That's not where the money is though unfortunately. It's why manufacturers like the solder on components. Easier to manufacture and make as small as possible. I hate it, and think it's bad for consumers.

We managed to get standards like that on desktops though. What's specifically different about modern laptops?

Size, heat, weight, and noise are more challenging for laptop design than desktops. Desktops tend to have things spread out much more and you just don't care about some of these issues.

Size and weight would definitely need to be consider by component manufacturers, but it doesn't seem like that big of a problem when it comes to standardizing a laptop chassis.

Heat and noise are definitely bigger concerns since a compact laptop cannot rely on large radiators, fans, and liquid cooling loops for cooling. But if you standardize the screw holes and socket positions on motherboards and graphics cards, I see no reason why a chassis couldn't ship with their own case-specific cooling solutions which leverage heat pipes and low-profile fans to provide cooling. Processor locations are already pretty consistent on desktop motherboards and graphics cards, so this wouldn't be something particularly new.

And those standards are shifting away too. See recent iMacs and Mac Minis.

If the only barrier to building your own laptop was soldering the components together, I'd have built one already.

They like solder because it's more reliable.

There's a lot more to building laptop than putting different components together. Portability, mobility, heat dissipation, design to put as many things as safely (and profitably) possible etc.

Building a desktop is relatively easy. Desktop is designed to be stationary and it's a lot more forgiving when connecting parts. There's a lot more room to manoeuvre and for heat dissipation.

I think to achieve the purpose, where end-user will be able easily customise a laptop will require a larger footprint and won't be appealing to many users.

In some sense Thinkpad X200 would fit what you're saying, its modding community is quite active and there are a lot of people selling old parts or parts designed to upgrade this laptop.

Last time I checked the only missing piece was a way to upgrade the display, which didn't age very well (IIRC original resolution was 1280x800 and no HiDPI)

The form factor of EOMA is simply unsuitable. Compared to any laptop motherboard or SBC, the volume available in the slot is tiny.

I would expect this is similar to the tesla issue of building a common carcass (skateboard) to build different types of cars on top of.

Reply on news.ycombinator.com