I would like to see a few things Framework’d (you can use that btw)
Mobile Phone (More like the Fair Phone, but better?)
Tablet
SBC (Not for development, just a micro PC.)
Smartwatch (would be nice, but I can understand it not being in the wheelhouse)
Monitor (Totally doable, swapable panels would be nice for tuning your setup exactly how you want it, with the outputs you want with it too using expansion modules, if possible powered via USB-C)
eGPU (Yes, this needs to happen)
Docking station (Essentially its just a platform you slot modules into, which would cross work with the expansion modules)
4/5G Modem Module
GPS Module
This is all stuff I would buy, and would attempt to convince others to buy too.
I’ve recently become very interested in getting one to replace my loud and hot desktop and hopefully, to replace my Raspberry Pi 4 as an always-on server/ HTPC; I’m particularly interested in one of:
Beelink GTR5/ GTR6
Minisforum UM690
Minisforum HX90
However, they are seem to share the following flaws:
Poor cooling (and so, noisy fans - not good for an always-on server!); usually afflicted by one of the following:
Fan whirring incessantly, even when idle (and cool)
Fan going from 0 - JET ENGINE BLAST when the temperature goes up by even 1C above idle
Fan flipping back and forth irritatingly between two different speeds, even when cool
This wouldn’t be so bad except that none of the above expose fan control to the operating system(!) so you’re completely at the mercy of the EC firmware.
Additionally, Technical Support is pretty bizarre: just finding out what version of the BIOS is current is non-trivial, and you there’s no site to download it from - instead, you have to email them and request it!
If you were to ask for, say, the EC source code they would probably think you mad
A representative thread from the last few days can be found here:
Needless to say, I think Framework could do better on all fronts!
Everyone is getting carried away here. Framework would do well to concentrate on just the laptop in its current form, as they’re already falling behind in supporting it and keeping it updated properly.
12th Gen BIOS 3.06 has been in beta since December 2022, after vulnerabilities were announced in September 2022. That means lag time from public announcement of a vulnerability in their BIOS to final release of fixes for the Framework laptop BIOS is currently around >= 5 MONTHS. And yet we are seeing forum posts about developing entirely new product lines.
Intel 13th gen has been out for a bit now. No sign of AMD motherboards. ARM-based laptop-grade solutions are on the horizon. Other manufacturers rapidly standardizing to high refresh rate screens. Vastly superior touchpads. DDR5 moving to CAMM, with the likelihood that we’ll be seeing quite a few more laptops with upgradeable RAM and storage, removing one of the main advantages of the Framework platform. Other laptops are becoming more repairable than you might realize as the legal framework (sorry) has changed in certain jurisdictions (e.g. France).
Not to mention no diagnosis or lessons learned regarding the Framework’s horrible battery life and how they plan to at least fix this on future iterations.
I own a 12th gen Framework and do like it, since I don’t need it to last for even a reasonable period of time without being plugged in, but…
Framework really needs to get it together with their primary product. The rest of the industry is just iterating so much faster and they seem to just be content with their first device and have called it a day.
No. Unlike any big company, Framework does not “intentionally leak” new products on the horizon to build up hype for it.
The next product could be announced next year or tomorrow. We simply dont know.
What we do know is that Framework is exploring other product groups beyond laptops.
While I do agree that Framework needs to iron out the remaining issues and that rather fast, I have to say that my 12th gen is pretty solid and there is no need to throw it all out for no real increase (13th gen) or even worse a new architecture (arm, risc).
On one hand you complain about long bugfix time, but then suggest redesigning the board for alternatives… you cant have it both ways!
I agree! My point was that Framework will need to continue to iterate on their primary product in the next 1-3 years, and that MAY (or may not) include adopting new architectures which would be a large undertaking. What it certainly DOES include is matching the level of basic functionality found in competing laptops (screen, touchpad, etc.) since these devices are likely to adopt some degree of repairability as well which removes some degree of differentiation for Framework.
My point was that the inability to push a BIOS update for 5 months, when they stated specifically in a Sept 2022 post that they had already received all the needed upstream packages (from Insyde), does not inspire confidence that Framework is currently able to iterate competitively WRT their primary product, much less branch out into some entirely new product category i.e. the topic of this thread “What should we build next?”
Yep that sucks, but we dont have the reasons for this though.
Honestly a laptop is one of the most complicated products in the tech industry to start with. Everyone expects a macbook quality notebook nowadays, even from new players. I wouldnt worry about this. The next product could be much simpler and still just as useful. We simply dont know.
I agree however that I wish some things, especially crucial bug fixes were faster. But probably even the Framework devs agree with this.
Would love to see the next iteration support lapdock features in a module. The GPD Pocket 3 has it. I love the Framework laptop… Let’s make this the only device I need to carry!
They should try to occupy small niches without current players. For example, 17 inch lapots with low power cpus with low power discrete graphics (like a p series cpu with a 1650). Alternatively a powerful AMD Apu.
There is not a lot of market, but there no rivales if framework enter there. They can monopolise the segment.
It will also synergize well with their current line up.
Aluminum or recycled plastic mouse with replaceable or repairable switches/pads/wheel/etc.
I imagine a design similar to the laptop and with wired UCB-C (though I’m sure wireless would work too, if less “repairable” with a battery).
Expands the idea of remaking consumer electronics without biting off too much too fast.
Also a phone styled like the Essential PH-1 but repairable like a Fairphone but later, of course.
Not sure if this has been mentioned already but I’d love to see a screen for the laptop that has a glass cover rather than plastic. My screen is pretty badly scratched already, I can also see artifacts from the plastic not being perfectly flat.
Fully repairable noise-cancelling earbuds.
The ones that exist are really really brittle… and no real way to change their battery (for “true wireless” or wired types) or to fix the wires (for those earbud where right and left are wired together – extremely short and tiny wires mixed with nylon fibres, impossible to properly solder, especially e.g. when they are cut just near the opening in the casing).
Anyway, if any company makes repairable ones, that would be one less PITA in the consumers electronic world.
Laptop chassis with 15" class screen. Start here, where existing components mostly cover what’s needed.
I’ve taken a look at the feasibility of doing something like this after getting my FW, and trying to think about what would be both useful and likely to be physically possible.
My direction of thought was for a 15.4" 3:2 screen, and this is what I came up with based on the physical size that implies:
The larger screen size, and assuming the same bezel width, gives an additional 1-9/16" width. The right side of the mainboard looks taken up with the ports and the WiFi module. The left side, however, looks pretty flat.
a: This space could in part be used for a larger module. If the choice is to keep the modules symmetrical around the USB-C plug, which is sensible for use in either slot, by pushing the mainboard a little away from the hinge, into the newly opened space, this should leave the existing module dimensions, plus a 1-9/16"x1-9/16" square extension for a large module.
b: This larger width should be easily sufficient for 2 USB-A ports or 3 USB-C ports on one module. In addition, the existing module space will be available for primary circuitry, leaving the entire new space for connectors/jacks/memory card slots/etc.
c: I’d hope the thicker case would be just enough for the expansion card to properly hold an RJ-45 ethernet port, probably with either a USB-A or USB-C (two of the latter, if possible)
d: It should be simple to have a blank larger expansion card that will just take an existing standard expansion card. Unfortunately a large blank that just takes up all the new space can’t take 3.
A thicker design, which would allow for a larger battery - commercial airline limit (99Wh, IIUC). In keeping with the Framework spirit, I would design it thicker, but otherwise have the same form factor, so a 13" battery could be used instead if the user wants to for any reason.
A place for a second M.2 header in the chassis that future mainboards could have a connector laid out to make use of for a second NVME drive. Even if, as some note, this may reduce performance, I think having the option fits the spirit of FW better.
A place for future mainboards to have a graphics card added, space permitting. Again, put it on the edge of the board and use additional chassis space for it. Sounds like this basically is #3 put to different use.
le touch screen, especially foldable or double touch screen (ie: top and bottom and no keyboard) then you can use it in a sideways orientation with 2 screens side by side
Although to be fair we kind of already have the fairphone. But I think that’s where “excessive modularity” is. You don’t need a phone literally made out of plastic blocks snapped together. You can use PCB interconnects to connect multiple different circuit boards together, ribbon cables, etc. This way you can still achieve the density of a typical un-repairable phone, except it is repairable.
I don’t use my phone a lot (or almost at all), so I most certainly don’t care. A 15 inch laptop (or a 16 inch) however I will definitely seriously consider.
Graphing calculator! Time to end the monopoly (or the “monopoly”) of Texas Instrument’s Ti Nspire CX CAS II! Open source? Maybe. However I don’t think testmakers will like that idea.
HDD/SSD enclosure for your old SSDs.
More affordable version of Framework Charger (?) I loved my charger until it got stolen .
DC to USB PD board (actually decent quality!)
Dell made one, but they stopped making it like 4 years ago because low demand.
It’s effectively a DC-DC switch (a.k.a. Buck-Boost converter) tied to a USB PD negotiator.
Or this thing (which I called “FireStarter 5000”) https://www.amazon.com/dp/B08P2LCY5H
I bought it. It worked until some random event decided to kill the PD negotiator pins. Then it stopped being a PD adapter and became a 5.2V adapter.
It might be difficult to make it an “official product” because how it deals with multiple different proprietary charger protocol (Dell’s 1-wire IC, HP’s 1-wire IC, Lenovo’s 1-wire IC), but I will buy one in a minute.
Can also probably just ignore the protocol part and let the user jumper-pin the maximum power draw.
Calculators need to be certified before they can be used in testing. This is to prevent possible cheating. This is why Texas Instruments have such a monopoly. Although, I would approve of another player in that arena. TI has almost a 90% margin on those calculators, it’s ridiculous. Open-Source doesn’t necessarily pose a problem. Hacking calculators has been around for a long time. And the hacking scene would frequently notify TI of exploits so they could be patched to prevent cheating. Allowing hacking does not equate to condoning cheating in my view.
This is sadly true for certification and false for the opensource part. Take the example of Numworks, a french company who created at the start an open-source calculator. Some dude showed how to install an OS allowing you to cheat and reported it to Poland’s Education ministry. Result? Numworks was banned in Poland, Spain (losing a big part of Europe’s market) and almost the USA until they moved the code from open to closed source and disabled the ability to use custom OS without warning the community.
Calculator space is an horrible place, in which open-source can sadly do more harm than beneficts to the company.
I would love to see a repairable bluetooth speaker! You could probably buy all parts off the shelf (except for the case maybe) and only have to figure out a way to connect them in a repairable way. Speakers like the UE Megaboom are very modular already, its just impossible to get spare parts and the repair is not easy (no screws etc.).
I’m looking at Mini PCs, and wondering what would give them more longevity.
Being a gamer, discreet graphics is almost a necessity for me. However, would MXM, or 2280 M.2 be better for GPU upgrades in the future?
Most of the Mini PCs focus on 4k and 8k office use. Integrated graphics seems to be able to handle 4k displays well enough these days, but for quality 4k video I am not sure the common 2.5gbps bandwidth is sufficient.
Port modularity on a Mini PC is probably of keen interest.
Home servers are typically a subset of the general consumer market. Most anything can be repurposed into a home server. This provides little to no draw in the consumer or “prosumer” market.
External GPUs are similarly part of the a subset of the general consumer market. What would be a repairable and upgradeable external GPU, but simply a housing?
What bandwidth are you referring to? I don’t know of any port or part that has that low of a bandwidth. Well, except an ethernet jack I guess.
Neither, eGPU handle this task fine and once TB5 rolls out with 80Gbps, bandwidth won’t be a concern either with eGPUs capable of actually utilizing a 3080 to its full capability.
.