Welcome
Welcome to refracta

You are currently viewing our boards as a guest, which gives you limited access to view most discussions and access our other features. By joining our free community, you will have access to post topics, communicate privately with other members (PM), respond to polls, upload content, and access many other special features. In addition, registered members also see less advertisements. Registration is fast, simple, and absolutely free, so please, join our community today!

Alternative usb installation method, part 2

Stuff that has not yet gone into the official build.

Re: Alternative usb installation method, part 2

Postby fsmithred » Fri Feb 21, 2014 9:17 pm

One more. I made a few minor tweaks and fixed the password logging problem. I've used enough parts of it enough times that I think it may be safe to let the kids play with it, so I'm letting go of the mickey-mouse numbering system I've been using, and I gave it a new version number.

There are still some minor bugs in it. I occasionally have leftover loop devices after I mount and unmount loopfiles several times. I think it happens when the task is terminated early. And sometimes there are files or mountpoints leftover in /tmp.


refracta2usb-0.9.5.deb
http://distro.ibiblio.org/refracta/files/Testing/

Code: Select all
  * (refracta2usb)
  * Changed some wording in settings window.
  * Running refracta2usb in debug mode automatically runs subscripts
  * in debug mode.
  *
  * (mkloopback)
  * Passprase is no longer logged.
  *
  * (select-loop-mount)
  * Passprase is no longer logged.
  * Mounted loopback file opens in non-root file manager.
  *
  * (functions_r2u)
  * Added function find_file_manager for select-loop-mount
  * Added $other_file_manager setting to config file.

User avatar
fsmithred
 
Posts: 1987
Joined: Wed Mar 09, 2011 9:13 pm

Re: Alternative usb installation method, part 2

Postby dzz » Sun Feb 23, 2014 10:14 pm

Gets better every time!

Don't know what to do about leftover mounts or loop devices. If the script is terminated other than cleanly that can happen.

Regarding the RW initrd patch: We have been using files based on live-boot_4.0~a7-1 which is now "old experimental". I updated the patch to use live-boot-4.0~alpha18 and posted here: http://exegnulinux.net/refracta/experim ... 8+4.tar.gz

It includes a fix for swap not automounting (which they say is a bug in util-linux not live-boot) Tested only here and on sid so far, seems to work fine.
dzz
 
Posts: 629
Joined: Wed Apr 27, 2011 11:53 am
Location: Devon, England

Re: Alternative usb installation method, part 2

Postby fsmithred » Sun Mar 30, 2014 12:38 pm

@dzz:
I just started to look at putting the newer patch-initrd (with files from alpha18) into r2u, and I see that alpha19 is now in experimental. Have you played with that at all? There appears to be a fix for the swap in it. The 9990-misc-helpers file is the same as in alpha18, so your patch for that should still work.
User avatar
fsmithred
 
Posts: 1987
Joined: Wed Mar 09, 2011 9:13 pm

Re: Alternative usb installation method, part 2

Postby dzz » Sun Mar 30, 2014 7:37 pm

live-boot (4.0~alpha19-1) experimental; urgency=low

* Updating copyright notices for 2014.
* Updating swap bootparameter in manpage to match code (Closes:
#738146).
* Adjusting internal file of the swap component to match its own boot
parameter.


That's the changelog. The correct swap parameter is "swap" not "swapon" but it doesn't work. They say the bug is in util-linux because fstab.d, where it puts the swap config, exists but is not sourced at boot: https://bugs.debian.org/cgi-bin/bugrepo ... bug=666163 and won't fix (I reported it but who am I to argue.. ). My patch does work, with either "swap" or "swapon"

I didn't bother with alpha19-1, which came out after my new patch for 18. Who cares about a copyright notice and tweaking a man page for something that doesn't and may never work?

This is important because live sessions on low-end machines freeze up without swap.
dzz
 
Posts: 629
Joined: Wed Apr 27, 2011 11:53 am
Location: Devon, England

Re: Alternative usb installation method, part 2

Postby thwak » Tue Apr 01, 2014 1:17 am

notes regarding files within the debs I retrieved tonight from http://distro.ibiblio.org/refracta/files/Testing/

refracta2usb-0.9.5.deb 2014-Feb-21
/usr/share/doc/refracta2usb/readme.refracta2usb.txt
conf file within this deb indicates version is 9.0.4
vs outdated readme.refracta2usb.txt still states refracta2usb v.07
-=-
help_r2u.txt states refracta2usb version: 9.0.5 (hmm, packaged within a deb named 0.9.5)

Several months ago, I had suggested merging the documentation into a (one) single helpfile.
This time 'round, I'm second-guessing that suggestion.
Instead, I think users would be better served by providing multiple .desktop files in the package
(to display in the DE menu immediately following, alphabetically, the refracta2usb launcher)
and
by placing a boilerplate block of comment lines atop each of the multiple files
which likely need the user's attention (or congfiguration tweaks) prior to initial refracta2usb operation
# ==== THESE MERIT YOUR ATTENTION
# ==== PRIOR TO PERFORMING refracta2usb OPERATION:
# /usr/share/doc/refracta2usb/readme.refracta2usb.txt
# /usr/share/doc/refracta2usb/help_r2utxt
# /usr/lib/refracta2usb/r2u_exclude.list
# ==== ALSO (POSSIBLY, OPTIONALLY)
# /usr/lib/refracta2usb/patch-live-initrd/readme.txt
# /usr/lib/refracta2usb/examples/live.cfg
# /usr/lib/refracta2usb/syslinux/menu.cfg
BTW, the left-side indenting along with less than 80 column line
length creates a difficult reading experience. The long strings within
the documentation representing paths don't fit this imposed
constraint anyhow, so why expect/demand the user
slurp 700+ lines of chunky documentation through a straw?


====================

file: /usr/lib/refracta2usb/r2u_exclude.list
still states refracta2usb 0.7
(if nothing has changed since, maybe this is proper/correct?)

====================

As is, it's difficult/impossible to follow the link ( http://www.ibiblio.org/refracta ) at
https://github.com/fsmithred/refracta
and find one's way to
http://distro.ibiblio.org/refracta/files/Testing/
^--- suggest the github page should provide links to both
and, optionally, cite the github url within the Testing/readme.txt document


==========================

Reiterating something else I mentioned months ago:
For refracta2usb, the exclude list is MUCH more sparse than the default list provided for refractasnapshot.
If, as dzz pointed out, it's advisable to to exclude the file containing "machineID" when performing a snapshot...
...surely it's equally important to consider such details during refracta2usb configuration
refracta2usb.conf
# This only gets used in mkusbcrypt if you copy /home from the running live system instead of from an iso file.
excludes_file="/usr/lib/refracta2usb/r2u_exclude.list"
Above, I've added bolding.
Point of confusion: "This only gets used (in mkusbcrypt) if..."
Without mkusbcrypt in the operation, the exclude file isn't referenced regardless of other variable factors?
thwak
 
Posts: 165
Joined: Tue Nov 20, 2012 3:58 am

Re: Alternative usb installation method, part 2

Postby fsmithred » Tue Apr 01, 2014 6:38 pm

Thanks, thwak.

Version numbers:
9.0.5 is a typo.
0.7 in readme.refracta2usb.txt is correct. I haven't changed that file since then. The only reason I left it in the package is because it has instructions for manually creating an encrypted partition to use with the custom hooks. That function (mkusbcrypt) will be gone when I modify mkpersist to make an encrypted partition and use it with a patched initrd instead of with the custom hook.
9.0.4 in the conf file is a typo, too. I just changed it in my build directory, so it'll at least say 0.9.5 when 0.9.6 comes out. Guess I should automate a find and replace for that.

80-column-wide text:
That's for the people who read these things in a console. When I made help_r2u.txt, I copied and pasted a lot from the older readme. Then I indented it to make it more readable when you open it in a text box from within the program. I'm looking at it now, and the text can certainly be wider - there's a lot of white space on the right. Without the indentation, the text is slammed up against the left margin of the window.

rsync exclude list:
This also is going away with mkusbcrypt. The excludes file gets used if and only if you run mkusbcrypt and you choose to copy your home from the running system. It serves little purpose - how much stuff will someone accumulate in their home directory between booting the read-only live system and transferring their home to the encrypted partition? Maybe some desktop settings will get changed and copied, but even that's not important, because you can just change those settings when you reboot with persistence, and the changes will stay.

machine-id:
This does not get copied in current versions of refractasnapshot, therefore it does not exist in the iso file or in the copy of filesystem.squashfs that's in /lib/live/mount/medium/live, which gets used if you select "Create live usb from a running live system." Note the misnomer - you're not copying the running system, but you are copying a copy of the iso contents while you're running the live system.

github links:
You lost me on that one. It's just a few clicks from github to the testing folder, and a single click from the testing folder to github. But yeah, it might make sense to point to the testing folder for newer versions. In fact, it's probably more important to do that from the sourceforge download site than from github.

desktop file(s) for refracta-tools help:
I kinda like that idea. It would be right there where people look for the programs.
User avatar
fsmithred
 
Posts: 1987
Joined: Wed Mar 09, 2011 9:13 pm

Re: Alternative usb installation method, part 2

Postby thwak » Wed Apr 02, 2014 12:57 am

if you select "Create live usb from a running live system." Note the misnomer - you're not copying the running system, but you are copying a copy of the iso contents while you're running the live system.
serves as a reminder that, in reading, my interpretation is often WAY too literal.

Thanks for the clarification.
thwak
 
Posts: 165
Joined: Tue Nov 20, 2012 3:58 am

Re: Alternative usb installation method, part 2

Postby dzz » Fri Apr 04, 2014 1:57 pm

live-boot_a21-1 just arrived in sid. Here is an updated patch-live-initrd: http://exegnulinux.net/refracta/experim ... 1-1.tar.gz

Need to check this will (hopefully) work for wheezy also
dzz
 
Posts: 629
Joined: Wed Apr 27, 2011 11:53 am
Location: Devon, England

Re: Alternative usb installation method, part 2

Postby fsmithred » Sat Apr 05, 2014 11:42 am

Just tried the a21 patch-initrd with a stick that already had the a18 patched initrd on it.
rw-media works
encrypted loopback file on second partition works
swap works

:)
User avatar
fsmithred
 
Posts: 1987
Joined: Wed Mar 09, 2011 9:13 pm

Re: Alternative usb installation method, part 2

Postby fsmithred » Mon Apr 07, 2014 1:54 pm

In refracta2usb-0.9.5, there's a problem with the Update task (replace_live function.) I'm trying to sort it out now. Part of the problem is that the window where you select the directory that contains the files you want to replace should have the '--directory' option. The selection will be the directory that's highlighted. So if, for instance, you select /live by double clicking it, you'll go into that directory, and the first subdirectory inside live will be highlighted. (if it's a current Refracta image, that will be the hooks directory.) If there is no subdirectory, then the first file will be the selection. Pay attention to what is highlighted in that window.

Another part of the problem is that the first CD image you add to the usb drive will likely have /live in the root of the device, but if you make a multiboot stick, the additional images will have their live folder one directory level down, inside the directory that you named when you made the multiboot. That makes it tricky to manipulate the paths such that the new image will be written to the right place. I probably need to add some more tests.

And the directions for what to select might be wrong or might need to be changed.
User avatar
fsmithred
 
Posts: 1987
Joined: Wed Mar 09, 2011 9:13 pm

PreviousNext

Return to Experimental

Who is online

Users browsing this forum: No registered users and 0 guests

suspicion-preferred