MNT Reform open source laptop with trackball(mntre.com)
mntre.com
MNT Reform open source laptop with trackball
https://mntre.com/media/reform_md/2020-05-08-the-much-more-personal-computer.html
31 comments
OMG that "review" is pure gold! <3
I am really hoping the proliferation of 4GB RAM - limited devices like the Pinebook Pro (and this) will encourage people to develop things that are far more memory efficient, so eventually 4GB will be enough for anyone, even if they want to run Slack, for example. There is no technical reason whatsoever that a glorified IRC client should be taking up gigabytes of RAM. The same goes for many web services. I seem to recall reading my email on a device with 8MB RAM total in my childhood; why does Google need 2GB for my Gmail tabs?
Plain html mode is still there
I heard (on a podcast with @mntmn ?) that someone already took the keyboard module and modified it to use an ortholinear layout. I'd be super interested in schematics for that, if anyone knows where to find them. Getting a bit into mechanical keyboards these days and it's pretty confusing for my poor little brain to use ortholinear on the workstation but staggered in the laptop. So it would be great to get this done before my unit (hopefully) arrives in december.
I wish there were more laptops that did batteries this way, rather than each device having it's own separate type of battery pack/cell/whatever. I know that will reduce battery performance, since they can't be tailored to each device as much. However as it stands, most device have a built-in live span after which replacement batteries will no longer be available to purchase, even if it was possible to install them (which for many devices, it isn't without special tools and skills).
I have G4 iBook that is in great condition and still runs well, but the battery is pretty much dead. Even though it's user swap-able, good luck buying a new battery that performs well (you can get knockoffs from sketchy eBay vendors, but they don't work very well).
If all the laptops used a set of standardized cells, you would sacrifice thin-ness and battery life to a certain extent, but now you can keep using your very expensive investment for many years, rather than having to chuck it (or at least only use it one place) after the battery stops holding a charge.
I have G4 iBook that is in great condition and still runs well, but the battery is pretty much dead. Even though it's user swap-able, good luck buying a new battery that performs well (you can get knockoffs from sketchy eBay vendors, but they don't work very well).
If all the laptops used a set of standardized cells, you would sacrifice thin-ness and battery life to a certain extent, but now you can keep using your very expensive investment for many years, rather than having to chuck it (or at least only use it one place) after the battery stops holding a charge.
Charles,
Duracell sells new batteries for laptops. That is how I got a brand new battery for my Thinkpad x220. I just checked and they got batteries for iBook G4:
https://www.duracelldirect.co.uk/laptop-notebook/apple/ibook...
Double check if it is the correct part number, you might need to go through their laptop selector page.
Duracell sells new batteries for laptops. That is how I got a brand new battery for my Thinkpad x220. I just checked and they got batteries for iBook G4:
https://www.duracelldirect.co.uk/laptop-notebook/apple/ibook...
Double check if it is the correct part number, you might need to go through their laptop selector page.
TIL, thank you for sharing!
Oh, I gave you the UK link, sorry, I don't know where in the world you are, but I bet you can ship a brand new battery there either from the UK site or the US site.
I checked the US version of that site just now - those batteries are really expensive compared to the cells inside.
If you can assemble your own then it will of course be cheaper but I'd rather buy that more expensive product than some weird knockoff. I'm quite happy with the one I'm using in my x220. Previously I tried buying used batteries in computer fairs here in London and all of them were way below capacity and usually operating below 40% efficiency.
I use the knockoff chinese batteries for my thinkpads that cost ~ $30. Haven't had a problem.
But that's not what I was talking about. I was referring to the GP comment, about using individual LiOn 18650 batteries that can be individually replaced. Those cost a few bucks each, and your X220 battery has about 6 of them inside. That's what the MNT Reform does, as well, which is what's so cool about their batteries.
But that's not what I was talking about. I was referring to the GP comment, about using individual LiOn 18650 batteries that can be individually replaced. Those cost a few bucks each, and your X220 battery has about 6 of them inside. That's what the MNT Reform does, as well, which is what's so cool about their batteries.
https://www.crowdsupply.com/mnt/reform
I found the description here more informative. I would prefer a 17 inch version with more memory and batteries.
I found the description here more informative. I would prefer a 17 inch version with more memory and batteries.
Previous discussion:
https://news.ycombinator.com/item?id=23141667
https://news.ycombinator.com/item?id=23141667
And three days before that: https://news.ycombinator.com/item?id=23113430
The whole project sounds fantastic, I hadn't heard about it yet. The casing looks boxy but that makes it hackable of course.
Oh wow, this is almost exactly the sort of computer I keep wishing would be made. Design wise it seems like a "spiritual successor" to this SPARC laptop I have [0], with a mechanical keyboard and trackball, except significantly better in every way. I'm gonna have to find a way to fit this in my budget ;)
Only thing that concerns me, is I've had pretty bad luck with an older i.MX6 SoC and its Vivante GPU. The etnaviv project is incredible but I've never gotten it to work reliably. Apparently this specific quad-core i.MX6 is known for having weird MMU/memory problems with etnaviv, but I'm curious -- how much custom software does this need over "stock" Debian to get a reasonably smooth and reliable desktop?
[0] https://src.paperfox.es/rdi.jpg
Only thing that concerns me, is I've had pretty bad luck with an older i.MX6 SoC and its Vivante GPU. The etnaviv project is incredible but I've never gotten it to work reliably. Apparently this specific quad-core i.MX6 is known for having weird MMU/memory problems with etnaviv, but I'm curious -- how much custom software does this need over "stock" Debian to get a reasonably smooth and reliable desktop?
[0] https://src.paperfox.es/rdi.jpg
This looks great! I think I may pre-order, but I’m only a bit hesitant. I know Apple brought some of the bigger ARM-on-desktop news... is the future already here (or in a few short months)? How’s life on ARM’d Debian?
Slow. The Nitrogen8M SOM used in the MNT Reform uses the i.MX 8M which is a 4x Cortex-A53 @1.5 GHz and a single Cortex-M4F @ 266 MHz. The performance should be comparable to a Raspberry Pi 3 B+. Still, I love the idea and I hope they support future SOMs from Boundary Devices.
From my dogfooding experience, it really depends on what you need to run and what compromises you are willing to make.
The biggest slowdowns I see with javascript-heavy web applications. Native applications like LibreOffice, GIMP, Inkscape, Ardour are fine, and so is working in terminals or with standard GTK3 apps. I also managed to patch KiCAD and etnaviv into working together to allow for GPU accelerated canvas.
"Normal" websites and PDFs are fine and scrolling is smooth in chromium. With VLC I can watch full HD H.264 movies, software decoded, and YouTube is fine too, if you can tolerate a bit longer loading times than on a faster Intel/AMD computer. There's open source driver support for H.264 video decoding, but the userland infrastructure for it is not yet in common applications.
I have to say I'm using the sway compositor, which is a clone of i3 for wayland. It's more resource efficient compared to a full blown DE like GNOME 3.
For Emacs, I use the new gccemacs branch with precompiled elisp which gives a nice speed boost.
What doesn't work: - Applications that need OpenCL - Applications that need Desktop OpenGL 3+ (currently, etnaviv does desktop OpenGL 2.1 and GLES 2.0). That means you're stuck with Blender 2.79b, for example.
Regardings the SOMs: I don't wanna tell too much yet, but we are working on several upgrades/sidegrades.
The biggest slowdowns I see with javascript-heavy web applications. Native applications like LibreOffice, GIMP, Inkscape, Ardour are fine, and so is working in terminals or with standard GTK3 apps. I also managed to patch KiCAD and etnaviv into working together to allow for GPU accelerated canvas.
"Normal" websites and PDFs are fine and scrolling is smooth in chromium. With VLC I can watch full HD H.264 movies, software decoded, and YouTube is fine too, if you can tolerate a bit longer loading times than on a faster Intel/AMD computer. There's open source driver support for H.264 video decoding, but the userland infrastructure for it is not yet in common applications.
I have to say I'm using the sway compositor, which is a clone of i3 for wayland. It's more resource efficient compared to a full blown DE like GNOME 3.
For Emacs, I use the new gccemacs branch with precompiled elisp which gives a nice speed boost.
What doesn't work: - Applications that need OpenCL - Applications that need Desktop OpenGL 3+ (currently, etnaviv does desktop OpenGL 2.1 and GLES 2.0). That means you're stuck with Blender 2.79b, for example.
Regardings the SOMs: I don't wanna tell too much yet, but we are working on several upgrades/sidegrades.
> "Normal" websites and PDFs are fine and scrolling is smooth in chromium.
Any difference in your dogfooding between Chromium and FF performance? I've got a Pinebook Pro and Chromium seems to load heavy pages considerably faster than FF.
Also-- what's the heat like when, say, playing video on one of the heavier websites? Like Reddit's infinite scroll where various cat videos are just flying past and autoplaying. Not that I've ever gotten caught in that hot loop... :)
Any difference in your dogfooding between Chromium and FF performance? I've got a Pinebook Pro and Chromium seems to load heavy pages considerably faster than FF.
Also-- what's the heat like when, say, playing video on one of the heavier websites? Like Reddit's infinite scroll where various cat videos are just flying past and autoplaying. Not that I've ever gotten caught in that hot loop... :)
Chromium tends to be a bit faster rendering wise, but there’s a new fix for etnaviv which allows smooth layer compositing in FF now as well. Javascript heavy apps and WebGL are faster in Chromium for me.
I don’t have any reddit experience to speak of.
But for example, mastodon is much faster than twitter, even if they have very similar functionality.
I don’t have any reddit experience to speak of.
But for example, mastodon is much faster than twitter, even if they have very similar functionality.
> I don’t have any reddit experience to speak of.
Sorry, I just meant a sort-of "stress test" where both CPU and GPU are being used heavily. I'm curious if part of the laptop can get a bit hot as the PBP can under these conditions.
Sorry, I just meant a sort-of "stress test" where both CPU and GPU are being used heavily. I'm curious if part of the laptop can get a bit hot as the PBP can under these conditions.
I'm no expert and can't examine all the bits in detail, however, I trust the likes of FSF and Stallman. So far, they haven't given approval to any laptop built after mid 2000s. I wonder why.
In our case, I didn’t kick off the process because I would have to send a free laptop to the FSF and we can’t afford that yet. Maybe later.
Also I’m not sure what the stance of RYF is regarding controller (non-CPU) firmware, as iMX8M needs a blob _in_ the DDR4 controller PHY.
Also I’m not sure what the stance of RYF is regarding controller (non-CPU) firmware, as iMX8M needs a blob _in_ the DDR4 controller PHY.
The Librem folks added a read-only flash for the PHY code - apparently firmware that can't be updated is "part of the hardware" for RYF: https://puri.sm/posts/librem5-solving-the-first-fsf-ryf-hurd...
Points for creativity, but in my opinion this borders on rule-lawyering with not much benefit. The code that tells the ARM/M4 to load stuff from that SPI flash can be easily changed, it's in the bootloader, so it can change that code before putting it into the DDRC.
Yeah, I found that a pretty odd design choice (and tbh also not sold on a blanket "proprietary firmware is better if it can't be changed" angle)
I like the placement of the control key. I usually remap caps lock to control.
https://remmina.org/the-patricians-choice/