One of the things that makes me interested in having a Reform is the large presence of unorthodox operating systems. I refer specifically to images for Sculpt and 9Front - both of which are much more recent in conception than Unix - although of course there is NetBSD as well. With the hardware being open and simple in relation to mainstream computers, the Reform seemed a natural fit for similarly small, simple niche OS.
However these images are mostly for the original iMX8 processor and support for subsequent ones - even the âplusâ version of that processor - seem to be slow. I hear that Sculpt recently was ported on the Pocket Reform with the âplusâ processor, which is phased out of the current purchase option sheet.
With the Reform Next seemingly making more steps to appeal to a wider audience out of the box and putting more effort into the ÂŁinux distro I wonder if - paradoxically - we are losing that sense of the MNT being a âneutral territoryâ for alternative OS.
I get it that people want a usable OS and tools to run it out of the box, and ÂŁinux is the dominant brand, just like mainstream buyers expect their computer to come with Windows pre-installed. And there may be a wistfulness about the product leaving behind the early and experimental days and becoming more professional. I am worried that tighter integration with ÂŁinux might suck the oxygen from other OS, as system integration with Linux become more tightly integrated.
As for me, and as somebody who likes to keep old machines running, I still consider the strength of ÂŁinux to be that itâll resuscitate any computer that you rescue in the rubbish bin. It is no longer particularly novel or subversive like it was in the 1990s, and readers may notice I (childish, I know!) put a UK pound symbol, like was popular amongst the ÂŁinux community when mentioning Micro$oft, to reflect that most of its development is corporate nowadays. Modern ÂŁinux is such a large and complex beast - compared to the other OS mentioned - it is perhaps not totally in tune with the open and legible aspirations of the Reform. But then again, it is what I am typing this on.
I donât have easy answers that would solve the problems of raising support to the other operating systems to the same level as ÂŁinux enjoys. We canât just make ÂŁinux worse to as not to overshadow the unorthodox OS. Clearly the Next is not going to be fully supported by the current range of niche OS from day one, but on launch perhaps an effort can be made to get as many niche OS running on it as possible. Set bounties perhaps? When the Reform first came out it was a fixed target for a few years with one processor - so perhaps a moratorium on new processors for a year or two to give an opportunity for RK3588 to be supported by the alternative OS scene.
I would like to hear what others think, especially amongst those using alternative OS full time. Hopefully Reform continues to have a large contingent of alternative OS in the future!
If you want other OS to support MNT reform family, you should ask the developers that work on those OS.
As you said, reform is really open source friendly and every firmware are in their gitlab, so the only thing that matter is whether other OS developer would interesting in reform family.
Go to *BSD(or any OS you want)âs official mailing list/channels/IRC and ask them to support reform family.
** I donât think MNT Research team has enough workforce to support other OS, 9front and sculpt support on imx8 is done by volunteer. **
nobody working on these other operating systems even has the new hardware yet. these teams are incredibly small, and in the case of 9front entirely self-financed. with the original cpu module, 9front support was facilitated by hardware donations from third parties.
While my daily driver is Linux I also have a fondness for ânicheâ operating systems. Linux is so large it is not quite possible for a single person to understand the whole thing. It is far harder to change things. In Debian I am for example working on making it possible to allow the creation of GNU/Hurd bootable disk images (spoiler: it works since the end of last year).
I fear you have to partly blame me for that as well. Debian integration with the Reform is working so well these days, that we now have the Reform packages in Debian proper and can install Debian using the official installer (with some patches on top). Did my efforts making Debian ÂŁinux better on the Reform suck the oxygen from other OSes? Maybe.
I do not disagree. Even in Debian, one of the few remaining large distros that is not directly (but only indirectly) influenced by large corporations, this is a common theme these days. Some things happen just because a large company decided that they should exist and not because some hobbyist thought it would be fun or useful to have this thing.
This is where I loose you I think. But maybe already above when you say that maybe the alternatives were maybe better if Linux were not as good. I donât think that this is a helpful way to think about this. Different Linux distributions can also happily co-exist without one having to be worse than the others, no? I also believe that alternatives to Linux can happily co-exist with Linux without Linux having to become worse.
Unfortunately I think we are also back to the ÂŁ in ÂŁinux: MNT somehow needs to put food on the table of those it employs. MNT already made a very risky experiment to appeal to freedom purists with the LS1028A. That experiment largely failed. While people are very vocal about how bad binary blobs are for them. Somehow, sales of the LS1028A did not make it a sustainable alternative for MNT in the end. With Linux alternatives we might be in a similar territory. There are a lots of really cool alternatives out there but are the people who would happily use them instead of Linux enough for MNT to rather appeal to them than to the more popular demand of a modern Linux desktop system? My gut feeling would be that no, they are not enough. In the end that is a strategic decision MNT has to make but I could not say Iâd fault them if they continue to rather appeal to the more popular OS. Iâd be surprised if a âmoratoriumâ would be doing MNT the company any good in the long term.
If you like Linux alternatives, please put your free time there. I love Debian which is why I put my own unpaid volunteer time there. If enough people put their time into Linux alternatives, then it does not matter how good Linux is. Iâm not convinced that Linux needs to be worse for others to succeed, sorry.
There are definitely some devs in the niche OS world who would be happy to contribute a port, but who donât have access to the hardware. A community-led effort to find interested devs and get hardware into their hands (by pooling funds to purchase said hardware, or by arranging to lend out end user hardware) would probably make this happen more quickly. But it canât be put onto MNTâs shoulders, as they likely donât have the resources at the moment. Building and supporting hardware is expensive and time-consuming, and they are not a large company.
If you arenât already working on a project and you want to contribute to the ecosystem, taking the lead on something like this would be a great service.
Thank you @mountain, I wholeheartedly agree with what you wrote. As the person who made large improvements to the Debian integration for the Reform, having the hardware in my hands was and still is what allows me to do what I do. 2 years ago, the Reform was not my primary device, so I was able to tinker with it to my hearts content which enabled the creation of system image v3. These days, @2disbetter borrowed me his imx8mq classic Reform and once again this enables me to fix a lot of stuff (see the reform-tools 1.65 release).
I believe that if we want Reform support of other OSes than Linux, we have to put our money where our mouth is.
In the past there have been talks about organizing some sort of crowd funding to pay a developer to enable suspend/resume support for A311D or RK3588. If something like this were to happen for 9front or sculpt, Iâd certainly donate money to such an effort as well.
My initial post is based much more on âvibesâ than anything material, and a sense perhaps that the Reform as âneutral territoryâ might be slipping away as the inevitable need to make it more professional, and although I was not sure if I had successfully elucidated this I feel the community as a whole and @josch in particular was very sympathetic. We have had a few stimulating replies, so thanks!
I suppose that the remedy might not involve a major effort but simply reassuring that the niche OS have not been forgotten. This might only require making this niche SO more visible in the Reform scene (even if not plastering OS logos on the homepage is not the style of MNT as a business!).
This is a good idea, providing it does not cause perverse incentives amongst the operating systems concerned. I have been a couple of years donating to Haiku a small amount each month, and it is to the board how they spend it. If a developer from Genode or 9front were paid to work on the Reform specifically, would that be a better use of their time than improving the code generally, or making it work better on, say, Think pad series? Given that Reform is a drop in the ocean, probably not, and this misallocated effort might actually impose an opportunity cost on that community.
As you say in your posts, the balance between neutrality and a reasonable experience for the end user is not easy to reconcile, and this is before we consider the âsecurityâ features that big brands are adding these days!
Just to echo @josch here, I think those of us who are interested in alternatives need to be willing to spend our time, energy, and money to make it happen.
Especially in the ARM ecosystem, it seems like every particular board and configuration needs a bit of attention to get a boot loader, kernel, and drivers working well.
Iâm really hoping that once I have access to a Reform I can try to get NetBSD running on it. My sense is that it should be possible, especially based on all the OpenBSD work done already, but I donât have enough experience to be sure. Weâll see I guess!
Iâd love to run OpenBSD on my Reform Next. The OpenBSD arm64 port already supports the original MNT Reform and the RK3588, so I canât imagine that supporting the Reform Next is out of the question. Iâll be glad to help test/troubleshoot stuff, but Iâm no C developer so I likely canât help with implementation.
This is a recurring topic we have in Debian. We receive donations but we only use them for things like paying for our server infrastructure or travel expenses. We do not use this money to pay somebody to work on a certain task. How would you select that task? Who would select it? Why spend donation money on that task instead of others? Why would one person get the money and not the other to work on this task? So instead: nobody gets the money to implement things.
On the other hand, I frequently observe that great project is made if somebody is paid half or fulltime to solve an issue. These are then mostly things that a company is interested in and thus is throwing money at the problem. Believe it or not: a lot of problems are really hard and people with the right skill also need money to live and they work better when they can put their full concentration on a task instead of just doing it half an hour twice a weekâŚ
Yes, this has always been the bane of ARMs existence. It is very easy for platforms to drift into obscurity, even with Linux, because there is only this version 5.12 vendor kernel that works on the device and thatâs it.
Patrick Wildt seems to be working on that: 2025-02-11.log
going back to the âclassicâ reform, I had hopes that there might end up being support for RISC OS, which I (and any other British forumites of a similar age will also) remember well from primary school, as it ran the Acorn computers that were prevalent in schools in the 1990s, and were in fact one of the original users of the ARM ecosystem thatâs now everywhere, including our Reforms. RISC OS development has continued/been restarted and it now supports Raspberry Pi among others.
When I investigated I was disappointed to find that the RISC OS developers consider the Reform to be too âeccentricâ to put it kindly (they were actually more unpleasant about it) and niche to be worthy of consideration, and therefore they were firmly against any support for it.
That said, this was before the Pi adapter came into being so by using that and an actual Raspberry Pi module, RISC OS would probably mostly work, and the same is likely true of most OSes that support Rasberry Pi. so that likely works in favour of the lesser well known distros and those with less development horsepower as well.
Having SoCs that use a more standardised method of booting might help as well. uboot works but it seems to be subtly different for every single arm device, vs something like UEFI that for all its faults (of which there are many) and shoddy implementations (of which there are also many) does seem to be largely consistent across boards and manufacturers.