i.MX8M comments?

7 respostas [Última entrada]
tonlee
Desconectado
Joined: 09/08/2014
jxself
Desconectado
Joined: 09/13/2010

They switched to one that needs a blob. I have lost interest in this now.

Re-designing the whole thing from scratch before you even have a shipping thing seems a huge waste of time. It's like "yeah, I know people were waiting but we're throwing it all away and starting over."

Doing so it seems a disserve to those that have been waiting while also putting a larger priority on going newer and with more performance over software freedom, even going to far in the announcement to try to minimize it by saying it's "only" (notice the word "only" to try to minimize it) a few kilobytes in size. They want a third party to try reverse engineer it, which is of course nice assuming that the third party is both interested and can actually pull it off. I will re-evalaute the situation if things change but this change has caused me to lose both interest for now and trust of them in the future. I don't feel that I can trust their future decisions anymore if these are the kinds they are shown to make.

chaosmonk

I am a member!

I am a translator!

Desconectado
Joined: 07/07/2017

> notice the word "only"

Yes, downplaying the problem does not demonstrate a serious intent to fix it. They are still advertising it as "our libre laptop" so their attitude should be that the project will be a failure if they don't actually liberate it. If they treat it as a stretch goal then I won't at all trust them to deliver.

tonlee
Desconectado
Joined: 09/08/2014

> reverse engineer it

Should they manage, do they then provide a better cpu than RK3328 and
h6? Thanks.

chaosmonk

I am a member!

I am a translator!

Desconectado
Joined: 07/07/2017

> Unfortunately, during the boot process, i.MX8M requires a closed-source firmware for an embedded ARCompact processor in the Synopsys DDR4 PHY.

> We are in contact with reverse engineers with the goal of analyzing what the capabilities of this so called PHY Utility Block (PUB) are, and to find out if we have a chance to replace this firmware at some point in the future.

I would need more information about the "chance to replace this firmware at some point in the future" before contributing. If it's just a vague promise to address the problem later so that they can get a non-free product out the door while advertising the hope of a free system, well, we've seen how that's worked out so far with the Purism laptops. On the other hand, I appreciate that they are being transparent about the problem. If they actually put someone to work on the problem who has a solid track record of successful reverse-engineering and are equally transparent about their progress, I would be willing to contribute if that would help fund the reverse engineering effort.

mntmn
Desconectado
Joined: 06/10/2019

Hello,

I found this thread yesterday searching for ARCompact information, ironically. I was initially a bit frustrated that I got almost no direct feedback about the i.MX8M choice.

The DDR PHY firmware is a major concern for me or I wouldn't have put it front and center in the article. The main puzzle piece that is still missing for me to make more sense of the disassembly (which I already have -- it turns out IDA Pro can disassemble the firmware files) and especially to be able to gauge potential security issues of this IP block, is the documentation for the PUB (PHY utility block). These files (PUB databooks, for example) do exist in Synopsys SolvNet, but I don't have access to that. I'm currently trying multiple different ways to get access to the docs.

Also, it is totally possible to make a i.MX6QP version of Reform. We would need to adjust the 0.4 motherboard to the form factor of Reform v2 and replace the charging/power circuit with the new one. I'm just not sure if it's a good idea to mix 2 different motherboards as options in the crowdfunding campaign, because it adds complexity.

jxself
Desconectado
Joined: 09/13/2010

I had never herd of the Reform laptop until I started using Mastodon and found your toots. After looking in to the device specifications it seemed that it might have been possible to use it without any proprietary software and I became more interested. I have to admit that my interest was limited to that one particular aspect of it and none other.

> I was initially a bit frustrated that I got almost no direct feedback about the i.MX8M choice.

Since originally learning about it through Mastodon it was clear that you were gathering feedback from beta testers that actually had the device because that's directly mentioned at https://www.crowdsupply.com/mnt/reform/updates/assembling-beta-units-feedback-35c3-etc but until I saw your message on Mastodon where you mentioned finding us talking amongst ourselves about what we thought, I had no idea that you were also interested in feedback from the general public as well. Since you do seem to be interested in feedback my first will be: Please update the project pages to make it clear that you are looking for feedback from the general public. Because that wasn't ever clear to me.

The second would be: If you are looking for feedback from the general public (and not just the beta testers that have the device) then please even go so far as to solicit it. I say that because I never knew the change to i.MX8M was even being considered in the first place, despite that I signed up to get emails from crowdsupply regarding the project. The only way I knew was when I got the email about the update "Re-Introducing Reform", which announced that the decision was already made. And without asking for any feedback. If you are interested in feedback then a (in my opinion major) change such as this should have had its own update saying "I'm thinking of this; please let me know." So, to add on to my first feedback, not only make it clear that you're looking for it but actively solicit it via those Crowd Supply updates. Now that I've said those two things, I'm moving on.

> The DDR PHY firmware is a major concern for me or I wouldn't have put it front and center in the article.

To be clear my concern wasn't about the placement of the information about the proprietary software; I would have figured out that problem and came to the same conclusion regardless. Although, yes, it was good that you were upfront about it. Rather it was that the decision was made at all.

It is, of course, your project and I don't mean to tell you how to run it. I had seen the Reform laptop as a way to have a computer without any proprietary software on it; and perhaps that was never your goal. (Because if it was then this seems a way to really fall off the wagon by making this change regardless of whether the reverse engineering is going to be successful or not.) But it is one of *my* goals in the computers I buy and since my interest was limited to that one particular aspect of it (being able to use it without any proprietary software) my interest in the Reform laptop ended once that was no longer possible. Perhaps I was disappointed for the project not having a goal it never had from the start but I suppose the outcome is the same either way. Earlier in the thread I did say that I would be willing to reevaluate this should the situation change in the future and I do stand by that.

> I'm just not sure if it's a good idea to mix 2 different motherboards as options in the crowdfunding campaign, because it adds complexity.

It is your project of course and you must decide such things yourself but I will add (as an additional piece of feedback since you say you are interested in such things) that if it were possible to order one with the i.MX6 I would consider that the situation around the Reform laptop to have sufficiently changed to warrant that reevaluation I mentioned earlier. I would also do that if the proprietary software were successfully reverse engineered. I say "if" because in reverse engineering projects like that there can be no timelines and no promises that anything will come of it. I do hope that something comes of it because the Reform laptop could be another way for us to continue to evacuate the Titanic: https://jxself.org/titanic.shtml

I wish you well in your endeavors.

chaosmonk

I am a member!

I am a translator!

Desconectado
Joined: 07/07/2017

Thanks for reaching out. Perhaps you can clarify some things.

> This firmware, which is only a few kilobytes in size

I appreciate that you are at least transparent about the existence of this flaw, but I also initially read this sentence and your use of the word "only" as downplaying its significance. If we misunderstand you, can you please clarify what you meant?

Can you also state whether you consider reverse-engineering the firmware to be an necessary step in this project, or a stretch goal? In order words, is there a chance that you would ship any units before reverse-engineering is complete? You tone makes it seem like you would, and yet you advertise this project as "our libre laptop," which will only be a truthful description upon successful reverse-engineering of the firmware.

Before you officially launch the Crowd Supply campaign and begin accepting pre-orders, I urge you to make your goals clear. While this recent change was disappointing news to me, if I had already pre-ordered one of the laptops I would have considered it a bait-and-switch. Before accepting pre-orders, I recommend that you either

(1) Go back to the original board that does not require non-free firmware, and save the more ambitious goal of liberating the new board for your next campaign.

(2) Stick with the new board for now, but promise not to ship any units before reverse-engineering is complete, and to fall back to using the original board if reverse-engineering is unsuccessful.

(3) Stick with the new board, regardless of whether reverse engineering is successful, but unless and until it is, do not advertise a "libre laptop."

I would love to see (1), and in that case would be happy to help fund your campaign once it launches. I would also be content with (2), but you would need to be very clear and consistent with your messaging to avoid confusion later on if you have to return to the original board. If you go with (3), I would personally lose most of my interest in the project, but still wish you the best.