Does anyone here use Mintstick for making livUSB's? Need testers.

greenjeans · 2025-05-27 23:33:20

Aaargh, of course it would be uefi....yeah I don't think i've even got a uefi iso around here anyway, for sure I haven't tested one. crap.

All right, back to the drawing board. Like I needed another reason to dislike uefi.

I've used the utility to write over sticks that already had a second partition just fine, must be something with that uefi partition, and I guess it needs to stay, and also possibly some additional delay needed in the script.

Thanks again Altoid, this is all good info, smacking my forehead right now that I didn't even think about an efi-iso needing a 2nd partition on a liveUSB, seems like a regression if you ask me.

Altoid · 2025-05-28 00:14:48

Hello:

greenjeans wrote:

... of course it would be uefi...

I should have remembered the problems I had installing a UEFI *.iso to a USB stick.
And the multiple hoops I had to jump through to get it done
Fortunately ralph.ronnquist figured it out.

Have a read, it's fun. 8^/

greenjeans wrote:

... must be something with that uefi partition ...

A UEFI machine will not boot if there is no UEFI partition it can find on the boot device.
USB stick, HDD, etc.

greenjeans wrote:

... it needs to stay ...

Yes, at least to boot UEFI enabled hardware.

Now, if your script can look for empty/unformatted space (and find it) then Bob's your uncle.
You just don't touch any formatted space, whether it be hidden or not.
ie: with any format, just in case.

greenjeans wrote:

... possibly some additional delay needed ...

Maybe.
Or maybe it is something else.
Just a hunch, with the error being of an aleatory nature.

greenjeans wrote:

Thanks ...

You're welcome.

As to all this UEFI shit being a regression ...
Methinks we have not seen the tip of the ~~ice~~shitberg yet.

Best,

A.

greenjeans · 2025-05-28 15:11:56

Yeah after sleeping on it, i'm not sure i'm inclined to put in the work to make it work with efi stuff, I have a couple machines here that will do efi but they're both in compatibility/legacy mode and i'm quite happy with them, and not willing to change that to test some stuff i'll likely never use. Efi is just shit, and I don't feel like stepping in it if I can help it.

It's been a fun project and an awesome learning experience, and now I have a super nice tool that i'll use regularly. Bonus: I un-installed the gnome-disk-utility because now I don't need it anymore, my utility actually does a trick the GDU does not.

fsmithred · 2025-05-28 17:20:24

Altoid wrote:

devuan_chimaera_4.0.3_i386_desktop-live.iso
This *.iso file did not generate a #$%& UEFI partition when I wrote it to the SD Card.

That's right. the i386 doesn't do uefi. I'm not aware of any 32-bit uefi systems.

greenjeans · 2025-05-28 19:17:35

That's right. the i386 doesn't do uefi. I'm not aware of any 32-bit uefi systems.

Apparently Crunchbang++ does, the iso Altoid linked to is efi, and won't boot up a live-session on my machine.

Something really wrong with that iso anyway, it's bizarre, downloading it alone caused my browser to lock up at the end of the process. And just looking at it on a usb stick that I burned, it's got a whole lot of odd stuff on it I usually don't see on normal isos.

And it apparently requires an efi partition as Altoid mentioned just to (attempt) to boot.

Uefi is just shit. Pure shit. Like shit with a side of shit-sauce and a shit-shake with some turds on the side.

stargate-sg1-cheyenne-mtn · 2025-05-29 07:39:57

mostly for future thread visitors and with respect to OPOT, just wanted to make others aware of EasyOS by Barry Kauler of PuppyLinux fame.

EasyOS

https://bkhome.org/news/tag_easy.html

Altoid · 2025-05-29 10:56:27

Hello:

greenjeans wrote:

... apparently requires an efi partition as Altoid mentioned just to (attempt) to boot.

Yes but not if you are booting a box with no UEFI.
eg: Asus 1000HE
You can just delete the UEFI partition with disks albeit not with gparted which does not see it.

My guess is that because of how the installer works now, any *.iso made for a post Chimaera/Bullseye kernel will produce a UEFI partition, independently of the kernel present in the installation.

This complicates everything.
eg: try to install Daedalus or any post Chimaera/Bullseye release on a USB drive.

All this UEFI, secureboot, apparmor, SELinux et al crap being imposed manu militari by the WinTel consortium will not end well for us.

Best,

A.

greenjeans · 2025-05-29 13:07:41

No, something in that iso is making that happen, I make liveUSB's all the time with daedalus based isos, my own, vanilla Devuan daedalus, even Devuan excalibur. It's not the kernel version triggering it. It's definitely the iso itself.

greenjeans · 2025-05-29 13:11:54

Why would easyOS be pertinent to this topic? It doesn't even use the .iso format from what i've read? To be honest, from reading about it, I think Barry lost a marble or two after he stopped developing Puppy.

Altoid · 2025-05-29 17:00:39

Hello:

I wrote:

... because of how the installer works now, any *.iso made for a post Chimaera/Bullseye kernel will produce a UEFI partition ...

greenjeans wrote:

... something in that iso is making that happen ...

Knowing nothing of what goes on inside an *.iso or how it works, I did not accurately write what I meant to say.

So I will rephrase:

"... because of how the installer works now, any *.iso made for a post Chimaera/Bullseye kernel will [have the instructions to] produce a UEFI partition ..."

That would cover the making that happen you are making reference to.

From what I gather, it seems that it is not something you can weed out just like that.
ie: clip out a piece of script.

Which is why I said how the installer works now, thinking it is part of something more complicated.

I seriously doubt that the #!++ i386 *.iso has any/many UEFI enabled i386 boxes / laptops to be used in.
So why is it a UEFI enabled *.iso?

It is also systemd enabled, so it is a no go for me.
I just wanted a peek at the Openbox config files.

@greenjeans
Maybe you could take a minute to look at them and see if they can be 'dropped' into a Devuan Chimaera with an Openbox setup.

Best,

A.

fsmithred · 2025-05-29 18:11:11

Install openbox and do a diff on the files you're curious about against the ones in /etc/xdg/openbox. (probably menu.xml and rc.xml)

FWIW, not related to the installer isos, but refractasnapshot has a config option for enabling uefi boot or not. (you always get legacy boot) It's just a matter of turning on or off one (I think it's just one) function. Oh yeah, the devuan-live isos have the same option because the same function is used in live-sdk.

To test those other config files, drop them in ~/.config/openbox/ and fire up an openbox session.

Altoid · 2025-05-29 22:09:31

Hello:

fsmithred wrote:

... a diff on the files you're curious about ...
... probably menu.xml and rc.xml ...

Right, those are the ones I'veen looking at.

fsmithred wrote:

...drop them in ~/.config/openbox/ and fire up ...

Yes. I was planning on doing that in my present 1000HE setup.

Re: enabling UEFI

fsmithred wrote:

... the devuan-live isos have the same option because the same function is used in live-sdk.

So the function to enable UEFI in the *.iso, be it installer or live, is in some part of the respective *.sdk?

If so, maybe the installer could have an option to turn it on/off, depending on the default setting.

As always, thanks for your input.

Best,

A.

ralph.ronnquist · Yesterday 00:21:54

I don't understand what you talk about.

Any choice between using UEFI bios and legacy bios is a choice on and for the target system, because it's the target system that decides how it boots, and at best a choice for the person installing.

An installer ISO image may be prepared to support either or both means of booting.

The installer software, when executed, may offer setup for either one or both means of booting into the installed system. One of the primary differences is then the requirements that the means of booting has on the target system's disk partitioning and filesystem choices. That is in the hands of the person installing.

Altoid · Yesterday 11:50:35

Hello:

ralph.ronnquist wrote:

... don't understand what you talk about.

Should not be a surprise. 8^)
I have already owned to "Knowing nothing of what goes on inside an *.iso or how it works ..."
Among other things.

ralph.ronnquist wrote:

... using UEFI bios and legacy bios is a choice on and for the target system ...
... at best a choice for the person installing.

Right.

Remember my post asking about dd'ing devuan_daedalus_5.0.1_amd64_netinstall.iso to a USB stick?

It was eventually solved but not before you figured out all the hoops I should jump through to be able to get a bootable USB device.
Not precisely a choice for me or for any other Devuan user with a non-UEFI box.

ralph.ronnquist wrote:

The installer software, when executed , may offer setup for either one or both means of booting ...

Yes, if it boots.

@ralph.ronnquist: as always, thanks for your input.

@greenjeans: I think the original topic has gone MIA after the testing / project ended.
For neatness' sake, you may want to consider marking the thread as solved.

Notwithstanding, the problems this UEFI crap has brought to the table remain.

Best,

A.

greenjeans · Yesterday 16:27:04

@greenjeans: I think the original topic has gone MIA after the testing / project ended.
For neatness' sake, you may want to consider marking the thread as solved.

Not necessarily, the conversation has pivoted somewhat, but it has turned to discussion of the stumbling block itself, and perhaps a better understanding of that will help with expanding the script to include efi stuff.

I did spend a little time working on re-writing, but after several attempts i'm not quite there, so it's on the back-burner simmering while I try to gather more data and do more reading. I have a working prototype (mostly) that extends it to bypass the need for Mintstick, and does everything I want it to, wipe-format-write-add data partition. But again, that's gonna be useless for efi-only isos that may need the boot partition on the liveUSB.

Gonna be twice the amount of code just to deal with the useless efi crap.

The officially official Devuan Forum!

#26 2025-05-27 23:33:20

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#27 2025-05-28 00:14:48

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#28 2025-05-28 15:11:56

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#29 2025-05-28 17:20:24

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#30 2025-05-28 19:17:35

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#31 2025-05-29 07:39:57

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#32 2025-05-29 10:56:27

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#33 2025-05-29 13:07:41

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#34 2025-05-29 13:11:54

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#35 2025-05-29 17:00:39

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#36 2025-05-29 18:11:11

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#37 2025-05-29 22:09:31

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#38 Yesterday 00:21:54

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#39 Yesterday 11:50:35

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

#40 Yesterday 16:27:04

Re: Does anyone here use Mintstick for making livUSB's? Need testers.

Board footer