Framework Raises $18M Series A
frame.work
frame.work
I really hope they stay true to their vision and that this means expanding it into other consumer product lines.
I might be doing something wrong, but using integer scaling at 2x results in a UI that's a bit too big for my tastes, but if I use 1x I feel it's too small, so fractional scaling seems to be necessary for me to get the right balance. It drives an external monitor at 1920x1080, and was just looking at buying another HDMI adapter, but I haven't yet been able to determine whether or not it's capable of driving two 1920x1080 displays in addition to its internal panel. That would be my ideal setup for WFH, I think. In case it's not obvious, my vision isn't good enough for me to feel strongly about HiDPI. =)
But, to echo what others have said: this laptop is a clear winner for my use cases. I'm very pleased that they managed to nail modularity without sacrificing build quality and aesthetics.
- "s2idle" = S0ix and "deep" = S3
- S0ix is similar to how smartphones go to sleep by just idling everything and not actually sleeping. On the Framework it drains around 1 or 2% per hour (I haven't measured exactly.)
- S3 sleep is the more familiar suspend-to-RAM functionality. The power drain is significantly less than S0ix.
Both sleep modes work fine in Ubuntu 21.10 on my Framework. I actually have mine set to S0ix because while S3 sleep has less battery drain, it takes something like 10 seconds longer than S0ix to wake back up after opening the lid. And since I'm mostly working from home, the battery drain isn't an issue for me.
Still, the difference between the two is significant enough that I kind of wish there were separate menu options for both. Though I realize that having "sleep" and "deeper sleep" menu options is not the most user-friendly design.
Of course you can also get hibernate (i.e. suspend to disk and turn off power completely) running, but my understanding is that hibernate is currently incompatible with Secure Boot, since it's non-trivial to verify that the system is resuming with the same kernel as the original boot. So you have to choose which one you want, hibernate or SB.
My current Linux laptop (released in 2015) can sleep for about week, resume in 3-4 seconds. Am I reading your comment correctly that buying a Framework would degrade my user experience?
To be fair, I don't think this is Framework's fault. My understanding is that this is a limitation of the current generation of Intel CPUs, so pretty much any current-gen Intel-based computer will have the same issue.
The way Windows laptops do this is to go from light sleep to deep sleep after being idle for a long period. You should be able to pull off the same trick in Linux.
Decided to actually time it.
It's 10 seconds from power button to grub bootloader, then 12 seconds from bootloader to login screen.
I decided to run the same test coming out of S3, and it takes 14 seconds on my macine to go from power button to login screen.
So waking from hibernation is roughly 8 seconds slower, or approximately the time it takes to get to grub.
https://wiki.archlinux.org/title/Power_management/Suspend_an...
https://wiki.archlinux.org/title/Power_management#Suspend_an...
I have my machine set to suspend on lid close and hibernate on power button.
Thank you for confirming.
I guess I will have to wait until an improved version of Framework that supports proper sleep.
mem_sleep_default=deep
https://www.youtube.com/watch?v=Jibqw5C3gk8&t=15m45s - timestamp to where he discusses fixing sleep issues
It sounds like the setting is there, but it does not work. Framework needs to fix that.
1. https://community.frame.work/t/issues-with-power-usage/12261
https://community.frame.work/t/issues-with-power-usage/12261...
> Issue was that I was running linux-zen, strangely. With linux-zen, the CPU wouldn’t get past C3 and idle power usage was 7-13W. With regular linux, the CPU reaches C8 and idle power usage is ~5W. Strange that linux-zen did that, but ok ig
The other main culprit is whether or not you can successfully enable PSR for the display. IIRC, the laptop cannot hit C8 or C10 sleep states without PSR enabled, even if you have the ASPM power saving mode enabled.
Speaking for myself, running Debian unstable, after lowering the display brightness, with the right settings my laptop easily hits 2.5-3W idle.
To be fair, it's not Framework's fault that fractional scaling in Linux is not great.
Except, I’m realizing it just now, it’s maybe why LibreOffice looks so bad on my Mac.
Seems like just telling apps to draw to a larger window is the way to go, which is why I don't understand why Apple and Wayland do it by scaling down an integer. Gotta be a good reason.
FWIW I've been using elementaryOS 6.1 for most of the time and it's been really nice too. Full disclosure though it was pretty DIY; wifi did not work right away (I had to upgrade the kernal), I had to install some software to enable the fingerprint reader, and I needed to manually enable fractional scaling.
I definitely recommend it to run Linux, the community is fantastic and another strong selling point.
I also love the expansion ports.
I am really glad they have raised a series A, because in 3-5 years I want the freedom to upgrade the motherboard without creating a ton of e-waste.
Never take outside money unless there is simply no other way.
It'll be really interesting to see where they go next. Given that they're already in computer and custom board space, I'd make a guess that it's going to be routers - similarly embedded devices and they'd make a great option for replaceable parts (i.e. WiFi, connection options etc).
As opposed to just taking what they have and putting it in a larger form factor, which, yeah, doesn't seem terribly ambitious.
Pop OS pretty much works without fuss on Framework, but taking care of the final 5-10% of tuning would make all the difference for many people to make the switch.
These are plagued by planed obsolescence. I would totally pay for repairable ones even with a premium. Imagine a dishwasher that you can hack and program yourself :D
Imo the main issue, is that these things are so cheap that replacement is cheaper than repair labor + cost of parts.
My fridge also had some kind of fault with the drainage tube (kept freezing up) but there's simply no way to access it. The back panel is not removable.
> PCBs that have been manufactured for decades ... same designs.
Absolutely not the case. Go and look up any washing machine control board. They're different for almost every model.
Washing machines & fridges may be better; I've never had to take one apart.
The only good solution to this problem I've come up with is requiring manufacturers to dispose of unwanted appliances for free. That way they have at least some incentive to make them last for a long time and be easily recyclable / repairable.
To be honest, even if you can't find a PCB replacement, a faulty PCB can be fixed by inspecting the components and/or replacing them. Usually it costs only a couple of dollars and an hour of soldering time, less if you are good at it.
Microcontrollers obviously are not as easy, but as luck would have it, my old (early 2000s) machine doesn't even have one - it is driven by those rotary encoder dial thingies that move in a circle and open/close various traces to begin/end program steps.
That's a pretty steep DIY requirement. Wonder how robust the market is for these kinds of repair skills. Hard to know when you have a problem like this, who to call about it, and how much it'll cost to fix.
Yeah right. Maybe if you're lucky or it's all low power stuff but circuits aren't generally designed to isolate failures. If one thing fails it will often cause other components to fail invisibly.
And yes I have tried it. The main transistor on an amplifier I had blew. Replaced it. Still doesn't work.
I was able to source parts for my 2004 Whirlpool french-door fridge until the sealed system finally died two years ago. That required a new fridge, sadly.
We replaced our washing machine and dryer a couple years ago, but that was only to upgrade to HE models. I could have repaired the circa 2000 washing machine. Not sure if the new HE models will be as easy to repair.
These guys have built a nice little business fixing appliance pcb's. Most household appliances are fairly repairable given the time/patience. The issue becomes the cost/benefit. I have a kitchenaid dishwasher that is about 7 years old now. About 3 years ago the heater control element broke and it cost me about ~$250 to fix. It would have cost a repairman ~$350. The dishwasher itself only cost $500. So if it breaks again I'm not sure I'll so readily try and fix it...
One example I came across recently after hearing how hard Dyson vacuums are to repair is Lupe https://lupetechnology.com/
"Someone" should make a list of these companies, or maybe there is already some site or place which focuses on quality-repairability-longetivity?
I was considering the i5 1135g7 Framework with 16gb of RAM. A friend of mine bought it, and it's idle power draw is of 8-9W with Gnome 3 and Firefox open.
I decided to get a laptop with a 5800H and an RTX 3070 instead. My power consumption as I'm typing this comment right now is of 7W with a dozen tabs and an IDE, messaging, etc... open. That is despite a 46% larger screen, and a much faster, unlocked TDP, 8 core CPU.
That's a real shame, IMO. Combined with a 55Wh battery, it's quite problematic, and was a deal-breaker for me.
I would argue that there is value in a brand that consumers trust to last for a long time. For example, if Framework launched a printer, you might be tempted to pay a premium for a product you knew was high quality and repairable.
[1] https://www.annuityadvantage.com/resources/life-expectancy-t...
If they buy from Dell, and I buy from Framework, Dell sells 5x as many laptops.
I wanted a life expectancy that was evenly divisible by ten. A 42 year old woman and a 38 year old man both have a life expectancy of 40 years based on the actuarial table I found. I played the odds and chose a 38 year old man for the example.
But from the blog post looks like they are doing extremely fine.
https://en.wikipedia.org/wiki/PC_Card
You could get standardised addons for laptops, but I guess they disappeared due to lack of interest/better alternatives.
It's weird how the world goes around in circles.
I don't know how much licensing would cost, but Framework could possibly offer a driver for the internal card as an addon. Then the slot could be used for a small headphone amp which would be more third party friendly I think.
It'd be particularly cool to get balanced outputs.
I feel there has been a growing market for a company like this over the last few years with the RTR movement and upgrade cycles for hardware delivering less than they previously did.
Pros: - price. I have spare RAM and NVME ssd so this brings the price further down - modular port...
Cons: - display/hinge is not very stable - touch pad is not as good as macbook's. I can't reliably click-and-drag using a singe hand (thumb finger to click, then index finger to drag). On macbook this gesture works perfectly fine.
Wish there is a bigger version, and/or a swappable battery!
It's just a bit too lose, to the point that the screen will slowly fold flat over the course of an hour on my lap. It's also just lose enough that it transfers a lot of vibrations on my desk (from typing) into wobble on the camera during screen-shares.
If that hinge were just a bit tighter, I'd have no complaints about it. It's a great laptop.
If I want to switch to Dutch, it won't let me. My basic understanding of French is that this product is not yet available in my country. Still horrible experience. Why not show me the English page?
This website makes the wrong assumption someone's language is tied to the location.
This is a bit of a premature "remote debugging" done, though.
From what I see, they use Solidus and Rails as backend. This does not rule out your conclusion. But it is not a typical thing: in Rails (and Spree, from which Solidus is forked) it is, for example, far easier to read out the `Accept-Language` header than to do geo-ip language detection. It is easier to the right thing than the bad. Again: not saying this makes it guaranteed the cause.
More likely causes: location-based caching (you got a cached version which was generated using French), or same thing with a CDN.
More likely, because with Rails (and Solidus, and Spree) it is easy to make too-simple-caches, easier to ignore a parameter when building a cache, than to do right thing.
I'm merely pointing out that your analysis of why it happens, is lacking.
It reminds me of customers who submit a feature-request or issue in the form of a concrete solution, rather than a good description of the problem they encounter.
"I want a dropdown with flags to switch to another language" vs "I cannot switch languages".
The latter presumes nothing, leaves the implementation to product-manager, -owner, UX designers or developers. The former is not only bad (flags for languages are stupid) it has a big chance of being dismissed entirely. Big chance the dev sees it and thinks "hmm, out of date, we already have that dropdown", but that due to a bug, setup, caching, incompatibility, browser-plugin or whatnot, it doesn't show up for the client.
My second statement wasn't a technical analysis but a principle one. One should not assume someone speaks the language of the country he or she is in. Plus, sometimes a country speaks more than 1 language.
https://news.ycombinator.com/item?id=30166510
As mentioned there, for countries like Switzerland with more than two languages, whichever language chosen would be wrong for a majority.
I'd even claim that for the market they're targeting, having English and nothing else for all regions would likely be way better than the current situation.
Any guesses?
I'd love to buy a Framework phone — if they can guarantee Android updates.
Making it repairable, upgradable, customizable and using morally sourced raw materials seems like too much for a start-up to handle. If they focused on 1 or 2 goals at a time, I reckon they'd be able to manufacture these faster and cheaper.
... they went from concept to manufacture of a completely new and novel laptop design in a year. The result, in that year, is a laptop that has a premium feel but is as cheap or cheaper than other competitors in the space, while also hitting all the marks on their mission of repairability and consumer focus.
This is, as far as I can tell, a complete, out-of-the-park success. I can't imagine how they could've executed any more flawlessly.
Dell recently showed a repairable concept https://www.delltechnologies.com/en-us/blog/pushing-the-boun... that gives pretty strong "don't buy Framework, just wait for us to catch up" vibes. Discussion: https://news.ycombinator.com/item?id=29569470
Small companies are a great tool to squeeze out market inefficiencies. Just look at EVs, which have been useful for decades and only recently seen companies spending large amounts of money on them as a (formerly) small company was laying the marketing groundwork.
The real genius of Framework is that it doesn't even have to be the Framework company manufacturing your favorite keyboard, it's just a standard form factor.
I personally would buy a replacement ortholinear keyboard, if one was available.
Not sure how that's useful as the form factor doesn't really allow for a 7-row keyboard, or full-sized arrow keys. Which is a real missed opportunity in my opinion: a laptop (any computer, really) is primarily a device with inputs (keyboard, touchpad) and outputs (display) so naturally you want those to be as good as possible (and if there isn't a 99.9+% consensus on what that means, modular/configurable) as that's what you'll end up using for hours every day. Why compromise on a single "pretty" layout of physical keys?
https://www.notebookcheck.net/Framework-Laptop-13-5-Review-I...
1)Default Ubuntu install was very laggy. Not sure what was up with this.
2)Default Arch install is pretty much perfect.
3)Both Ubuntu and Arch have ok battery life. Pretty much as I expected for an Intel laptop. (i.e. not great)
4)By default, battery drain with the lid closed is pretty bad though. Don't expect to leave it laying around and it to still have battery when you go back to it. There is probably an easy solution to this but I haven't had time to look into it.
There is some community discussion on ways to optimize suspend power consumption as well: https://community.frame.work/t/ubuntu-21-10-on-the-framework...
I'm talking about schematics, PCB layouts and 3D models of the chassis.
Nothing proprietary or patented?
A Framework car with open source ECM would be quite an ambitious project!
Protectli is very nice, but the area could use a push forward.
I very much am looking forward to their next product, and I hope they stick to their goal of a sustainable, consumer-friendly laptop ecosystem.
When I shopped for a development laptop recently, some gaming laptops were really good matches for my needs. But the keyboards / key caps were squishy rubbery things that I couldn't get used to.
It would be awesome each buyer could tweak the laptop design based on personal preferences, where large manufacturers could never justify selling that exact config due to low demand.
With a friend, we had an idea of making custom keyboards for all laptops. With a 3d printer you may get away dealing with most of the covers. We desisted from the idea noticing that to change the keyboards you have to practically disassembly the complete laptop, testing would be very expensive too.
Anyway, congrats; eagerly awaiting my UK pre-order!
Since the EU is not an immigration union, as a permanent resident of Germany I am only allowed to work in Germany, not in any other EU country. Granted, I could probably get a work visa for another EU country, but if I wanted to remain a German resident things would get tricky very fast.
For fresh or prospective immigrants it's generally less favorable; and every EU country has different regulations as well as attitudes to immigrants from different regions.
Thus I recommend, when you get to that point, that you do some research about which EU countries are most advantageous to you and your potential workforce, and open subsidiaries in the countries that make sense. Probably more than one, out of the gate. Because missing out on the pool of immigrant talent in Europe is probably not your goal.
And congratulations on the funding!
(No affiliation with Deel, but about to start a new remote job with them as EoR)
Can you please blog a bit on origins of framework? I read in press releases elsewhere that you 3D printed the prototype laptop. It'd be very interesting to see how such startups get off the ground. Given that multiple tech-shops have tried and binned plans for modular hardware, framework's early success is remarkable... yet, the story of its nascent stages remain elusive.
I would like to write more about Framework as well, but will give it more time to be more sure of what worked and what didn't.
Some feedback from a global visitor-who-tried-but-gave-up-to-preorder:
Please decouple preferences for language/location/currency. You may be surprised about the number of people in your target market who aren't native speakers of the local language in their place of residence, and using credit cards denominated in a foreign currency (sometimes with ridiculous conversion rates).
Having them default to the same might be reasonable but never make hard assumptions about your visitors like that.
Look at iherb.com and aliexpress for two sites that IMO get this right.
It drives me crazy when these are coupled because they have to make an opinionated choice that, for the Swiss use case, has to be wrong for the majority of users.
Very excited for your raise!
Can miscellaneous third parties can create and sell their own Framework-compatible expansion modules?
Can other manufacturers make laptops with Framework-compatible expansion slots?
Example: https://github.com/commaai/openpilot