| Summary: | Review and test LiveUSB HOWTO for 2007.0 | ||
|---|---|---|---|
| Product: | [OLD] Docs on www.gentoo.org | Reporter: | Xavier Neys (RETIRED) <neysx> |
| Component: | Other documents | Assignee: | Xavier Neys (RETIRED) <neysx> |
| Status: | RESOLVED FIXED | ||
| Severity: | major | CC: | dcoorna, docs-team, releng |
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| URL: | http://www.gentoo.org/doc/en/liveusb.xml | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: | missing cp437, can't mount vfat | ||
|
Description
Xavier Neys (RETIRED)
2007-04-02 09:57:48 UTC
(In reply to comment #0) > 2006.1 was "broken", 2007.0 should fix mounting issues. Mounting issues haven't been fixed. Codepage 437 is defined a FAT default and selected as module but the module is missing from the squash FS Tried both 2007.0 & 2007.0-r1 Please consider releasing a -r2 It's missing from the squashfs or the initramfs? Created attachment 118655 [details]
missing cp437, can't mount vfat
Prolly the initramfs indeed.
FYI, tried LiveCD on a larger key without any issues. Only the minimal CD is affected.
Folks, Should the title of this document reflect that it is x86 only? We have a similar document for ps3s. *sigh* I really hate this. We have *never* supported booting from a USB stick. Back when brix was still around, he did the work to do this and also made sure that it actually worked during the release cycle. We simply aren't going to release *another* bump just to fix something we never supported in the first place. If Andrew wants to do yet another release bump, he can do so, but I'm saying that it was never supported and not something I plan on supporting in the future. There's just too much incidental little crap like this that gets tacked on to the millions of other things we have to check with each release. That being said, if someone were to file a bug asking us to make sure cp437 was built-in on the next minimal kernel, we could do that easily. ;] Basically, I leave it to Andrew on what he wants to do as the x86 Release Coordinator. Not particularly. Sigh. One more regression.... I guess it wasn't reasonable to expect something as trivial as this to be fixed in eight months. Doc now requires LiveCD, >=i686 or amd64, and >=1GB key First off, drop the attitude. As I said, we *never* supported it. EVER. There was a developer who dedicated his time to working with us to verify that *his* document would work. We were unaware of the cp437 requirements, as the last release had a completely different issue which prevented us from even knowing this issue existed. So no, it wasn't eight months. Nobody reported the cp437 problem to us until after the release. Anyway, if someone wants to track the bug that will fix this issue, it is bug #143774 and will be fixed in the next release. I'm sorry to buzz on this button again I initially warned brix (henrik brix andersen) of this problem on 9th august 2006 at the time he answered he left (or was about to) gentoo dvlpmnt team Anyway whatever his reasons, I've been trying as much as possible to avoid the unsupported feature 'stick-boot', but today I'm facing a situation where this sounds as my only affordable solution though. unfortunately, seems that even 2007.0 or 2007.0-r1 minimal install cd still suffer from this sickness I'm not sure of chris gianelloni's feelings about memory stick booting; however I'd enormously appreciate directions required to fiddle with either the squashfs and/or the initrd in order to make this work My problem is exactly the same as Xavier Neys's screenshot (img_1348.jpg) and I honestly don't see how this relates to a cp437 absence problem .. by the way, I've ordered a bigger stick so that I'll be able to try if the install-cd (not the minimal-cd) is works on USB as Xavier claimed. if this works, I'll have an alternative, nevertheless, I'd really enjoy whatever cause this bug fixed within 2007.1 The Release Engineering team never has supported booting from anything except for the official media using official means. The media is not and never has been tested to work on a USB stick. When brix was a developer, he did all of the work. I told him then that we (RelEng) had no interest in this and it would go unmaintained if he left. If booting from a USB stick works, it'll be out of either pure luck or because someone stepped up to do the testing as well as discovering and finding solutions for any bugs which cause this to not work. Essentially, someone needs to step up and do *all* of the work necessary for this, except for the actual CD building itself. Otherwise, it's only luck if the media boots from a USB stick. Sorry, we just have better things to do (like making sure it boots on various hardware from CD) with our limited time. 1) the "live usb how to" remains on gentoo site installdocs, get rid of it then since it's unsupported 2) only the minimal cd is affected 3) the problem is identified, "the vfat readability" module I don't see why you're going so harsh about such a simple fix U want testers ? we do test ! we reported the bug, we hope to further test may be you don't realize that booting from usb is far from uncommon is extremely convenient, and is definately future... why in hell did you move to bootcd ? why not distribute a set of 40 floppy images ? does this question sound dumb...? think again about USB support then and apply your own logic again. I'm really tired of all these rude comments anyway I'll do the test copying to non vfat partition on the stick and I'll test othe rbooting system than the one described on DOCUMENTATION PAGE that you offer, fortunately I found a solution, I turned around people's idiotic behaviour, and got a rescuecd, which is a fresher gentoo than the minimal cd anyway, and it goes perfectly at installing a gentoo, has full support of gpt/efi partitions (you know also that +2 tera file systems are more and more frequent ..don't you? that's called future man) and they care about their users, and have a nice page describing how to put their thing on a usb stick installing gentoo I'm not. If the vfat thing is all that's needed, then it'll work. My point is not that I don't want it to work, it is that taking the time to make it work is something Release Engineering can't do. If someone were to step up to do it, we'd be appreciative. I'm sorry if you read anything else into my comments. I was merely explaining the case, not being rude. Sometimes it helps to give people the benefit fo the doubt, especially online, since you can't get tone from text. I've tested it with a 2007.0-r1 minimal CD and its not working! Error postet in the jpg image from xavier! Please reopen the bug sorry! didn't read all comments! tested on install-x86-minimal-2007.0-r1 as well but am getting the same error. As listed in img_1348.jpg. I wouldn't mind donating some of our staff time to test this for minimal-2008. Is there something available I we can test with? Can you re-open? 2008.0 isn't even in the planning stages yet. Hi, I have just created a "install-x86-minimal-2007.0-r1" based LiveUSB using the information found at http://forums.gentoo.org/viewtopic-p-4639760.html#4714286 perhaps the best way to resolve this BUG is to start using the ext2 filesystem for all Gentoo LiveUSBs and ask that the Offical "Gentoo Linux LiveUSB HOWTO" be changed appropriately. Regards, theDuck I've just successfully created an install-amd64-minimal-2008.0.iso on a USB stick with grub as the bootloader, worked like a charm. Since this issue looks fixed, can this page: http://www.gentoo.org/doc/en/liveusb.xml be changed to mention that the FAT mounting issue has gone away now? I expect there is a shortage of necessary volunteers to maintain these docs - if that's the case, then I would be very willing to lend a hand. I really benefit from the quality of Gentoo documentation, so I feel I should contribute. Thanks for the info William. It would be best if you could provide us with a patch for the doc with all the proposed changes included in it. Just open a new docs-team bug (2008.0 liveusb review for that purpose) and drop the patch there. |