saintless
am using your new base now
looks and works great
amazing work
thanks for doing all this
wanderer
Mintpup (THREAD CLOSED)
Hi, wanderer.
To skip this message after apt-get update:
install linuxmint-keyring:
To find and try special Mint utilities use:
I will post some more information about special Ubuntu/Mint structure different from Debian and Puppy. It will be useful later if you like to use scripts from DD or Puppy.
About copy on write (cow) with official mint (casper) boot we need to add showmounts to the boot code and keep it in the iso and any boot code examples. Booting with showmounts will make visible /cow and /cow-rw-backing (screenshot attached). The good news is /cow doesn't change the location using fresh boot, save file and encrypted save file. It is mounted always in /cow (in Debian Wheezy/Jessie it changes the location depending on persistence method and the same goes for the boot partition location).
Also new directory /casper appears with empty subdirectory with the same name as the squashfs module(s) in use on boot. Seems the overlay process in Ubuntu is somehow different from Debian. In Debian I can see the content of modules in use using showmounts parameter (in /live/module-name.squashfs in Squeeze for example).
Right click and Properties on /casper-rw-backing will give you the size of casper-rw save file in use.
/cow is the place where the system keeps all the changes from the session (and from the save file in use also). It makes possible to create module from changes only.
/cdrom is the boot partition (where /casper folder with main module, vmlinuz and initrd files is located).
Comparing the same locations in different live systems:
/cdrom (in Ubuntu/Mint) = /live/image (in Debian Squeeze) = /lib/live/mount/medium (in Debian Wheezy/Jessie but it could be different depending on the persistence method) = /mnt/home (link to /mnt/sdaX boot partition in Porteus) = /mnt/home (link to /initrd/mnt/dev_save or /initrd/mnt/dev_ro2 in Puppy).
/cow (in Mint/Ubuntu) = /live/cow (in Debian Squeeze) = /lib/live/mount/persistence/loop1 (or depending on persistent method /lib/live/mount/persistence/sdxx in Debian Wheezy/Jessie) = /mnt/live/memory/changes (or for changes on EXIT only /mnt/live/memory/images/changes-exit in Porteus) = /initrd/pup_rw (in Puppy).
Toni
To skip this message after apt-get update:
Code: Select all
W: GPG error: http://extra.linuxmint.com rebecca Release: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY 3EE67F3D0FF405B2
W: GPG error: http://packages.linuxmint.com rebecca Release: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY 3EE67F3D0FF405B2Code: Select all
apt-get install linuxmint-keyringCode: Select all
apt-cache search mintAbout copy on write (cow) with official mint (casper) boot we need to add showmounts to the boot code and keep it in the iso and any boot code examples. Booting with showmounts will make visible /cow and /cow-rw-backing (screenshot attached). The good news is /cow doesn't change the location using fresh boot, save file and encrypted save file. It is mounted always in /cow (in Debian Wheezy/Jessie it changes the location depending on persistence method and the same goes for the boot partition location).
Also new directory /casper appears with empty subdirectory with the same name as the squashfs module(s) in use on boot. Seems the overlay process in Ubuntu is somehow different from Debian. In Debian I can see the content of modules in use using showmounts parameter (in /live/module-name.squashfs in Squeeze for example).
Right click and Properties on /casper-rw-backing will give you the size of casper-rw save file in use.
/cow is the place where the system keeps all the changes from the session (and from the save file in use also). It makes possible to create module from changes only.
/cdrom is the boot partition (where /casper folder with main module, vmlinuz and initrd files is located).
Comparing the same locations in different live systems:
/cdrom (in Ubuntu/Mint) = /live/image (in Debian Squeeze) = /lib/live/mount/medium (in Debian Wheezy/Jessie but it could be different depending on the persistence method) = /mnt/home (link to /mnt/sdaX boot partition in Porteus) = /mnt/home (link to /initrd/mnt/dev_save or /initrd/mnt/dev_ro2 in Puppy).
/cow (in Mint/Ubuntu) = /live/cow (in Debian Squeeze) = /lib/live/mount/persistence/loop1 (or depending on persistent method /lib/live/mount/persistence/sdxx in Debian Wheezy/Jessie) = /mnt/live/memory/changes (or for changes on EXIT only /mnt/live/memory/images/changes-exit in Porteus) = /initrd/pup_rw (in Puppy).
Toni
- Attachments
-
- cow-rw-backing.jpg
- (61.48 KiB) Downloaded 639 times
RF_KILL disabling wifi problem
Hello Toni,
I just noticed your post about frisbee. I was trying to get wireless working last night, first using frisbee and then I tried Peasywifi. In doing so I got what I knew was a bash-related error message (export -f not recognised by sh) and planned to report back to remind you that you had to make sh a link to bash rather than dash (though I forgot you could do much the same with command dpkg-reconfigure dash). So anyway, I did get both frisbee running and in a different attempt Peasywifi working. However, I still can't get wireless to work.
The problem is (as noted from dmesg) that the underlying wireless hardware itself seems to be disabled by the OS at boot time, and I don't know how to re-enable it. The message I get from dmesg following a reboot is:
I am currently stumped on this. My wifi uses iwlwifi driver and the firmware and modules are all there and show up in lsmod list. But if I try to connect in either frisbee or peasywifi I always get error message saying wifi not enabled and the hardware wifi switch on the laptop doesn't make any different and nor does commands like ifconfig wlan0 up. Note that iwconfig reveals that wlan0 does exist but it is simply not enabled because, I think, of that RF_KILL disable at boot time.
Have you actually been able to connect via wifi or has anyone else?
Cheers, William
EDIT: I haven't yet googed "RF_KILL bit-toggled to disable radio" but will do now to see if anyone else ever had similar problem.
I just noticed your post about frisbee. I was trying to get wireless working last night, first using frisbee and then I tried Peasywifi. In doing so I got what I knew was a bash-related error message (export -f not recognised by sh) and planned to report back to remind you that you had to make sh a link to bash rather than dash (though I forgot you could do much the same with command dpkg-reconfigure dash). So anyway, I did get both frisbee running and in a different attempt Peasywifi working. However, I still can't get wireless to work.
The problem is (as noted from dmesg) that the underlying wireless hardware itself seems to be disabled by the OS at boot time, and I don't know how to re-enable it. The message I get from dmesg following a reboot is:
I also have the xfce version of Linux Mint as a live CD, and wireless connects fine with that. The difference seems to be that during booting Linux Mint first toggles RF_KILL to disable radio but later on nearer the end of booting re-toggles it to enable radio (dmesg shows this), and then wireless works. So something seems to be different even though MintPup should simply be a cut-down version of the same (isn't it?).RF_KILL bit toggled to disable radio
I am currently stumped on this. My wifi uses iwlwifi driver and the firmware and modules are all there and show up in lsmod list. But if I try to connect in either frisbee or peasywifi I always get error message saying wifi not enabled and the hardware wifi switch on the laptop doesn't make any different and nor does commands like ifconfig wlan0 up. Note that iwconfig reveals that wlan0 does exist but it is simply not enabled because, I think, of that RF_KILL disable at boot time.
Have you actually been able to connect via wifi or has anyone else?
Cheers, William
EDIT: I haven't yet googed "RF_KILL bit-toggled to disable radio" but will do now to see if anyone else ever had similar problem.
Last edited by mcewanw on Sat 11 Jul 2015, 06:20, edited 1 time in total.
Hi, William.
I don't have wifi and can't test it. Initrd.lz is rebuilded from the smaller base module and has a patch to make encrypted save to work. Boot with the unpatched initrd.lz from here (it is the same as in Linux Mint 17.1 with only change mint to root user):
http://kazzascorner.com.au/saintless/De ... .1/casper/
Try first initrd.lz (official mint boot) instead porteus. If wifi doesn't work with unpatched initrd.lz the only reason could be purging some driver or firmware package. The kernel modules are the same as in LinuxMint 17.1. No changes there.
Toni
I don't have wifi and can't test it. Initrd.lz is rebuilded from the smaller base module and has a patch to make encrypted save to work. Boot with the unpatched initrd.lz from here (it is the same as in Linux Mint 17.1 with only change mint to root user):
http://kazzascorner.com.au/saintless/De ... .1/casper/
Try first initrd.lz (official mint boot) instead porteus. If wifi doesn't work with unpatched initrd.lz the only reason could be purging some driver or firmware package. The kernel modules are the same as in LinuxMint 17.1. No changes there.
Toni
Just some links with similar wifi problem in Mint and Ubuntu:
http://billauer.co.il/blog/2014/08/linu ... eless-lan/
http://forums.linuxmint.com/viewtopic.php?f=53&t=142863
http://ubuntuforums.org/archive/index.p ... 53072.html
http://ubuntuforums.org/showthread.php?t=2168840
http://ubuntuforums.org/showthread.php?t=2168840
http://billauer.co.il/blog/2014/08/linu ... eless-lan/
http://forums.linuxmint.com/viewtopic.php?f=53&t=142863
http://ubuntuforums.org/archive/index.p ... 53072.html
http://ubuntuforums.org/showthread.php?t=2168840
http://ubuntuforums.org/showthread.php?t=2168840
Thanks Toni, I'll try your suggestion shortly. I note now on googling about RF_KILL that there have been issues with this at other times, though I'm still to read more about what these messages say.
William
EDIT: Note that my laptop is an HP Elitebook 2530p, and some of the links about RF_KILL problem involve HP Elitebooks I notice. However, Linux Mint itself clearly works around the problem. However, I expect some of these RF_KILL-related links may contain a solution - I'm still reading... :-)
William
EDIT: Note that my laptop is an HP Elitebook 2530p, and some of the links about RF_KILL problem involve HP Elitebooks I notice. However, Linux Mint itself clearly works around the problem. However, I expect some of these RF_KILL-related links may contain a solution - I'm still reading... :-)
github mcewanw
Hi Toni,
Rather than Porteus boot method I tried that alternative initrd.lz you suggested above, but result was the same - RF_KILL blocks my wifi during boot. Once again I could free the block using:
rfkill unblock all
Linux Mint official must do that or some other config elsewhere since the problem does not exist in official Linux Mint boot.
I'll keep looking.
William
Rather than Porteus boot method I tried that alternative initrd.lz you suggested above, but result was the same - RF_KILL blocks my wifi during boot. Once again I could free the block using:
rfkill unblock all
Linux Mint official must do that or some other config elsewhere since the problem does not exist in official Linux Mint boot.
I'll keep looking.
William
github mcewanw
Thanks William!
Checking both initrd.lz (patched and official) I see nothing different in files containing "rfkill" line. The smal base-minpup module and official Linux Mint 17.1 filesystem.squashfs has the same /lib/firmware and /lib/modules content.
It should be some special mint package or configuration file removed with apt-get purge, but I'm afraid you are alone in this fix searching. I can't help much without wireless.
Edit: Not a fix since wifi in official Mint works from boot for you, but interesting similar problem with Linux Mint 17 and the simple solution on page 2 pressing: Fn+F11 (airplane mode):
https://www.pcspecialist.co.uk/forums/s ... Mint-17%29
Toni
Checking both initrd.lz (patched and official) I see nothing different in files containing "rfkill" line. The smal base-minpup module and official Linux Mint 17.1 filesystem.squashfs has the same /lib/firmware and /lib/modules content.
It should be some special mint package or configuration file removed with apt-get purge, but I'm afraid you are alone in this fix searching. I can't help much without wireless.
Edit: Not a fix since wifi in official Mint works from boot for you, but interesting similar problem with Linux Mint 17 and the simple solution on page 2 pressing: Fn+F11 (airplane mode):
https://www.pcspecialist.co.uk/forums/s ... Mint-17%29
Toni
Hi Toni,
I've been looking through Linux Mint 17 config files (best I can) but can't find anything. I presume MintPup and Linux Mint both use systemd, which I know nothing about. I wondered if systemd was somehow unblocking wifi in Linux Mint, but wasn't doing so in MintPup?
Anyway, I give up for now and will just use rfkill unblock all
William
I've been looking through Linux Mint 17 config files (best I can) but can't find anything. I presume MintPup and Linux Mint both use systemd, which I know nothing about. I wondered if systemd was somehow unblocking wifi in Linux Mint, but wasn't doing so in MintPup?
Anyway, I give up for now and will just use rfkill unblock all
William
github mcewanw
The MintPup base doesn't use systemd but upstart. I think the official Mint 17.1 also boots with upstart by default.mcewanw wrote:I presume MintPup and Linux Mint both use systemd, which I know nothing about. I wondered if systemd was somehow unblocking wifi in Linux Mint, but wasn't doing so in MintPup?
But in official Mint 17.1 you have installed systemd-services, libsystemd-daemon and other packages missing in MintPup-base module. Inside initrd.lz (patched or not) there is /scripts/init-top/udev calling systemd-udevd daemon (but probably nothing to do with wireless):
Code: Select all
/lib/systemd/systemd-udevd --daemon --resolve-names=neverEdit: Seems systemd-udevd has much to do with wireless:
https://forums.gentoo.org/viewtopic-t-9 ... art-0.html
https://bbs.archlinux.org/viewtopic.php?id=156539
/lib/systemd/systemd-udevd is included in udev and it is installed in Mint-pup-base module, but maybe it needs systemd-services, systemd-shim, libpam-systemd and libsystemd-daemon to work.
Toni
Edit 17.07.2015: The iso image replaced with new initrd1.xz and changes in some scripts for porteus-boot to keep the same system structure as DebianDog.
The script from here removed since it will not work well with the changed system structure.
I wrote we need to add showmounts parameter for casper-boot (initrd.lz) to be able to create module from changes from /cow directory. But after adding showmounts we need to change also remaster-mint-cli script to exclude /cow, /casper-rw-backing and /casper. Attached archive with remaster-mint-cli mods.
I will update the script in this post if we need to change it again.
With porteus-boot and without showmounts parameter for casper-boot the included in the iso script works well but it is recommended to replace it with the attached version in /opt/bin
Toni
The script from here removed since it will not work well with the changed system structure.
I wrote we need to add showmounts parameter for casper-boot (initrd.lz) to be able to create module from changes from /cow directory. But after adding showmounts we need to change also remaster-mint-cli script to exclude /cow, /casper-rw-backing and /casper. Attached archive with remaster-mint-cli mods.
I will update the script in this post if we need to change it again.
With porteus-boot and without showmounts parameter for casper-boot the included in the iso script works well but it is recommended to replace it with the attached version in /opt/bin
Toni
Last edited by saintless on Fri 17 Jul 2015, 19:49, edited 1 time in total.
saintless and everyone
I am posting this from the base iso that saintless provided
everyting works and looks great (except I guess wireless)
making the hybrid iso available (as you have done)
is the best way for me to use the system
this mint (and debian) system
is a great gift not only to the puppy and mint world
but to the whole linux community
It finally makes these large distros accessable and modifyable
I cant thank you enough for all your work
eventualy I would hope for a mint base with
firefox (as well as dillo already in)
a media player
jwm (already in)
a terminal (already in)
gtk1-roxfiler (already in)
gtk1-emelfm
gtkedit geany leafpad (already in)
and wireless
the other utilites from puppy debian or mint as needed
many people (myself included)
would probably not need to add much more
I am reading your posts and trying to follow along
eventually even I should be able to learn and use your system
once again thanks
wanderer
I am posting this from the base iso that saintless provided
everyting works and looks great (except I guess wireless)
making the hybrid iso available (as you have done)
is the best way for me to use the system
this mint (and debian) system
is a great gift not only to the puppy and mint world
but to the whole linux community
It finally makes these large distros accessable and modifyable
I cant thank you enough for all your work
eventualy I would hope for a mint base with
firefox (as well as dillo already in)
a media player
jwm (already in)
a terminal (already in)
gtk1-roxfiler (already in)
gtk1-emelfm
gtkedit geany leafpad (already in)
and wireless
the other utilites from puppy debian or mint as needed
many people (myself included)
would probably not need to add much more
I am reading your posts and trying to follow along
eventually even I should be able to learn and use your system
once again thanks
wanderer
Hello Toni,saintless wrote:
Edit: Seems systemd-udevd has much to do with wireless:
https://forums.gentoo.org/viewtopic-t-9 ... art-0.html
https://bbs.archlinux.org/viewtopic.php?id=156539
/lib/systemd/systemd-udevd is included in udev and it is installed in Mint-pup-base module, but maybe it needs systemd-services, systemd-shim, libpam-systemd and libsystemd-daemon to work.
Toni
I haven't as yet tried to get wireless working on my system apart from using "rfkill unblock all" to activate the wireless led on my system (some systems may not even need that). I will try tonight to see if I can now use wireless via the appropriate commands (or frisbee, if I can get that working properly). I am reasonably confident it should now work, but time will tell... I would prefer to not add any systemd services and so on and leave such additions as an option. I'll report back progress later.
William
github mcewanw
Hi Toni,
Sorry, I must have made a mistake first time I tried frisbee a few days ago (and I had already arranged for bash to be called by symbolic link sh). I don't know what is going on, but I deleted my changes folder and used porteus boot to boot up MintPup. Using DebianDog, I had already downloaded the DebianDog version of frisbee you gave the link to. Then I did the following:
and to my huge surprise, wireless was working. At no stage did I type "rfkill unblock all" - it was not needed!
I quickly checked the frisbee script to see if perhaps it had an rfkill line inside it, but it did not. But anyway, it appears that something inside frisbee itself re-activated my wireless, which is great of course!
Main news is that wireless thus works out of the box for me following the above two commands.
It would be good if Rox could have some right-click options to, for example, install a deb package rather than the way I did it with dpkg -i.
Anyway, I've typing this via wireless on my HP Elitebook 2530p using the default Dillo browser in MintPup... :-)
I will now repeat the above, just to confirm that wireless will come on again and also I'll reboot to check it stays... All good so far though!
William
Sorry, I must have made a mistake first time I tried frisbee a few days ago (and I had already arranged for bash to be called by symbolic link sh). I don't know what is going on, but I deleted my changes folder and used porteus boot to boot up MintPup. Using DebianDog, I had already downloaded the DebianDog version of frisbee you gave the link to. Then I did the following:
Code: Select all
dpkg-reconfigure dash
then, from the directory I had frisbee deb file stored:
dpkg -i frisbee_1.2-20140602_i386.deb
I quickly checked the frisbee script to see if perhaps it had an rfkill line inside it, but it did not. But anyway, it appears that something inside frisbee itself re-activated my wireless, which is great of course!
Main news is that wireless thus works out of the box for me following the above two commands.
It would be good if Rox could have some right-click options to, for example, install a deb package rather than the way I did it with dpkg -i.
Anyway, I've typing this via wireless on my HP Elitebook 2530p using the default Dillo browser in MintPup... :-)
I will now repeat the above, just to confirm that wireless will come on again and also I'll reboot to check it stays... All good so far though!
William
github mcewanw
Yep, wireless works fine now, just by following the steps in my above post - I just verified it again from a clean install.
As before, on first boot, using Porteus boot method, the bare MintPup blocks my wifi from working (the wireless led stays off and pressing the wireless switch has no effect). On running "dmesg | grep RF_KILL" from the commandline I find the message RF_KILL disables radio but it never gets re-enabled at this stage. Then I simply used "dpkg-reconfigure dash" (answering NO) and installed frisbee using "dpkg -i".
Note, that something during frisbee installation enabled wireless (I was watching the wireless led carefully as dpkg installed it and at a certain point the led came on). Then I checked "dmesg | grep RF_KILL" again and this time noted that there was now an additional RF_KILL message saying "radio enabled". Perhaps their is a frisbee pre-install script included in the frisbee deb but I haven't checked for that yet [EDIT: checked in deb postinst but saw nothing relevant], but some command somewhere during installation caused that "radio enable" switch [EDIT: apparently something installed as part of frisbee; the simple act of installing frisbee activates wireless, before even running frisbee itself - weird to me, but that's how it is... Note that nothing extra was installed, since no network connection at that time though I noticed message about resolvconf being removed (as far as I remember) by the install process]. I did also note that in the terminal output the wireless led appeared to come on when the screen output said: "ioctl [SIOCSIWENCODEEXT]: Invalid argument" but I have no idea if that has anything to do with it.
EDIT2: I also note that /etc/frisbee/frisbee.conf contains the line "wireless_enabled=1" so maybe frisbee has some command that uses that to reactive wireless, but I don't know which command as yet or if that is the reason (but out of curiousity, I'm looking through frisbee script again...). EDIT3: Hmmm but that doesn't explain things anyway since wireless becomes activated on installing frisbee and before even running it... EDIT3: I currently suspect it may be something to do with the bash script called "frisbee" installed into /etc/init.d, since wireless_enabled appears in that script in relation it seems to wpa_supplement and being in init.d I could imagine that /etc/.init.d/frisbee script automatically being activated on install (?). EDIT4: Alternatively, I note that /etc/init.d/frisbee scripts sources a whole lot of other scripts and the frisbee functions, so maybe something in them re-enabled wireless.
EDIT5: Ah, yes... I think this is it in /usr/lib/frisbee/func lines 1003 to 1005:
EDIT6: NOTE WELL therefore, that if anyone wants to use a different wireless tool (such as Peasywifi) they may need to check that different tool also contained some such mechanism for re-enabling wireless (Peasy may or may not already, I haven't re-tried it)... (assuming I am correct in my detective work above...) EDIT7: I've mentioned the matter to rcrsn51, the author of PeasyWIFI, for any confirmation or advice or in case it is useful info for his own program. EDIT8: I have since tried PeasyWIFI, but couldn't get it to detect wireless networks on Scan, but I have never used the program before so perhaps it is another problem or that I didn't install it properly; frisbee is working well for me though.
To start frisbee, I just entered the command in a terminal. It automatically then scanned for available wireless networks, and found my home one. I then highlighted that, then pressed 'Connect' button and entered my wireless pass key. Important to then wait for quite a while until frisbee does all its stuff in the background with dhcp and so on. Just be patient and wait... could take even a minute or so and available wireless screen becomes empty for a moment but then message about checking wlan0 comes up and soon a green box shows the new IP address for the machine. Then ran Dillo and am posting here!
Okay, so wireless is now confirmed as working fine for me in MintPup without doing anything other than install frisbee!
William
As before, on first boot, using Porteus boot method, the bare MintPup blocks my wifi from working (the wireless led stays off and pressing the wireless switch has no effect). On running "dmesg | grep RF_KILL" from the commandline I find the message RF_KILL disables radio but it never gets re-enabled at this stage. Then I simply used "dpkg-reconfigure dash" (answering NO) and installed frisbee using "dpkg -i".
Note, that something during frisbee installation enabled wireless (I was watching the wireless led carefully as dpkg installed it and at a certain point the led came on). Then I checked "dmesg | grep RF_KILL" again and this time noted that there was now an additional RF_KILL message saying "radio enabled". Perhaps their is a frisbee pre-install script included in the frisbee deb but I haven't checked for that yet [EDIT: checked in deb postinst but saw nothing relevant], but some command somewhere during installation caused that "radio enable" switch [EDIT: apparently something installed as part of frisbee; the simple act of installing frisbee activates wireless, before even running frisbee itself - weird to me, but that's how it is... Note that nothing extra was installed, since no network connection at that time though I noticed message about resolvconf being removed (as far as I remember) by the install process]. I did also note that in the terminal output the wireless led appeared to come on when the screen output said: "ioctl [SIOCSIWENCODEEXT]: Invalid argument" but I have no idea if that has anything to do with it.
EDIT2: I also note that /etc/frisbee/frisbee.conf contains the line "wireless_enabled=1" so maybe frisbee has some command that uses that to reactive wireless, but I don't know which command as yet or if that is the reason (but out of curiousity, I'm looking through frisbee script again...). EDIT3: Hmmm but that doesn't explain things anyway since wireless becomes activated on installing frisbee and before even running it... EDIT3: I currently suspect it may be something to do with the bash script called "frisbee" installed into /etc/init.d, since wireless_enabled appears in that script in relation it seems to wpa_supplement and being in init.d I could imagine that /etc/.init.d/frisbee script automatically being activated on install (?). EDIT4: Alternatively, I note that /etc/init.d/frisbee scripts sources a whole lot of other scripts and the frisbee functions, so maybe something in them re-enabled wireless.
EDIT5: Ah, yes... I think this is it in /usr/lib/frisbee/func lines 1003 to 1005:
Code: Select all
function start_wpa {
rfkill unblock wlan 2>/dev/null #140221
ifconfig $INTERFACE up
To start frisbee, I just entered the command in a terminal. It automatically then scanned for available wireless networks, and found my home one. I then highlighted that, then pressed 'Connect' button and entered my wireless pass key. Important to then wait for quite a while until frisbee does all its stuff in the background with dhcp and so on. Just be patient and wait... could take even a minute or so and available wireless screen becomes empty for a moment but then message about checking wlan0 comes up and soon a green box shows the new IP address for the machine. Then ran Dillo and am posting here!
Okay, so wireless is now confirmed as working fine for me in MintPup without doing anything other than install frisbee!
William
github mcewanw
Hi William, Wanderer.
William, thanks for testing and investigating this wifi problem. I had a chance to test the mintpup-base hybrid iso without frisbee on friend's HP laptop and after boot and typing "rfkill list all" shows the wifi only and both (hardware and software) have no. I think wifi is not blocked on boot in this laptop but didn't have time to create save file and install frisbee to test. I think there is no airplane mode function in this laptop also.
Anyway frisbee will be included and since it works for you it is the best fix.
I guess the network manager in official Mint also has some similar rfkill line in some script.
Most important thing for me is to make GUI remasterdog and remastercow working in MintPup base. I think it is already done but more testing is needed. New links /live/cow and /live/image are created depending on the boot method and this helps to adapt DD scripts easier.
And mods for debdog-installer to work with MintPup-base (still nothing done here yet).
So far I have DD-Jwm menu system working, frisbee, wallpapers changer from Terry, xhippo, gnome-mplayer-1.0.7, transmission-2.31, calculator_fox (I will add xfe too because I don't feel comfortable witout it), emelfm2.
About rox configuration I'm not the best man to do this. I can add some settings from DD but not everything will work. I will see if I can setup rox to have some default run actions for special scripts like install deb packages. Such settings working well for user and root account take time and much testing.
If someone likes to help here and has experience with rox it will be better.
I don't plan to make another iso soon, but I can upload new testing module to be replaced in frugall install if you like to test the changes (same as we started with DD).
Wanderer,
why do you like to have included Firefox? It takes over 100Mb and in few months the included in the iso version will be old. This means you will have to upgrade it in save file and the new version will be installed again in your save file (keeping the same size older firefox in main squashfs module)?
Isn't it better to keep only dillo and let the user choose better browser? It is simple terminal line typing to have firefox working in few minutes:
Or we can make extra sfs module working in MintPup? It could be replaced with new firefox version easy.
We need to decide to keep included or as separate module for extra download /usr/share/locale files. We keep them out for extra download in DD. No problem to keep them included in the main module for MintPup but the iso size will be bigger. Locales package is included anyway and I prefer not to remove it. Unlike Debian Mint/Ubuntu complains lacales is missing when installing/removing packages.
Edit: I still have troubles to download the mintpup wallpapers. If someone can mirror them inside archive here for example I can include them:
http://www.datafilehost.com/
Toni
William, thanks for testing and investigating this wifi problem. I had a chance to test the mintpup-base hybrid iso without frisbee on friend's HP laptop and after boot and typing "rfkill list all" shows the wifi only and both (hardware and software) have no. I think wifi is not blocked on boot in this laptop but didn't have time to create save file and install frisbee to test. I think there is no airplane mode function in this laptop also.
Anyway frisbee will be included and since it works for you it is the best fix.
I guess the network manager in official Mint also has some similar rfkill line in some script.
Most important thing for me is to make GUI remasterdog and remastercow working in MintPup base. I think it is already done but more testing is needed. New links /live/cow and /live/image are created depending on the boot method and this helps to adapt DD scripts easier.
And mods for debdog-installer to work with MintPup-base (still nothing done here yet).
So far I have DD-Jwm menu system working, frisbee, wallpapers changer from Terry, xhippo, gnome-mplayer-1.0.7, transmission-2.31, calculator_fox (I will add xfe too because I don't feel comfortable witout it), emelfm2.
About rox configuration I'm not the best man to do this. I can add some settings from DD but not everything will work. I will see if I can setup rox to have some default run actions for special scripts like install deb packages. Such settings working well for user and root account take time and much testing.
If someone likes to help here and has experience with rox it will be better.
I don't plan to make another iso soon, but I can upload new testing module to be replaced in frugall install if you like to test the changes (same as we started with DD).
Wanderer,
why do you like to have included Firefox? It takes over 100Mb and in few months the included in the iso version will be old. This means you will have to upgrade it in save file and the new version will be installed again in your save file (keeping the same size older firefox in main squashfs module)?
Isn't it better to keep only dillo and let the user choose better browser? It is simple terminal line typing to have firefox working in few minutes:
Code: Select all
apt-get install firefoxWe need to decide to keep included or as separate module for extra download /usr/share/locale files. We keep them out for extra download in DD. No problem to keep them included in the main module for MintPup but the iso size will be bigger. Locales package is included anyway and I prefer not to remove it. Unlike Debian Mint/Ubuntu complains lacales is missing when installing/removing packages.
Edit: I still have troubles to download the mintpup wallpapers. If someone can mirror them inside archive here for example I can include them:
http://www.datafilehost.com/
Toni
Re: RF_KILL disabling wifi problem
If you have time to test better we can make working Peasywifi deb package. I guess some conflicting files will appear but we can solve this if you can test the deb till we build it safe and working well in Debian.mcewanw wrote:So anyway, I did get both frisbee running and in a different attempt Peasywifi working.
Toni
Personally, I think it is better to keep the base release small, which basically means leaving main browser as an option to be added by user, and similarly for locales. I suppose there could be something like Quickpet or whatever it was called, which automates the browser choice by using the likes of apt-get install firefox underneath?
Of course, Fred also preferred to include main browser in the distribution, which is a slight pity I feel, since I see no real advantage to forcing that large and major choice. But MintPup was started by Wanderer, so his call.
I do note that Wanderer seems to want to pass development mainly over to you Toni, which makes sense since you are so familiar with cutting down Debian-based system. I wonder if the alternative, adopted by Fred would be good for wanderer too, where MintPup becomes part of the overall DebianDog stable such that development of all three: DebianDog-jwm, DebianDog-openbox, and MintPup can be more easily kept in sync. Another advantage I see in that is that Fred might be more interested to also contribute to MintPup (openbox version?). Particularly important, I feel, that such a project has as many main developers involved as possible - that far increases the chance that the distribution will be maintained.
William
Of course, Fred also preferred to include main browser in the distribution, which is a slight pity I feel, since I see no real advantage to forcing that large and major choice. But MintPup was started by Wanderer, so his call.
I do note that Wanderer seems to want to pass development mainly over to you Toni, which makes sense since you are so familiar with cutting down Debian-based system. I wonder if the alternative, adopted by Fred would be good for wanderer too, where MintPup becomes part of the overall DebianDog stable such that development of all three: DebianDog-jwm, DebianDog-openbox, and MintPup can be more easily kept in sync. Another advantage I see in that is that Fred might be more interested to also contribute to MintPup (openbox version?). Particularly important, I feel, that such a project has as many main developers involved as possible - that far increases the chance that the distribution will be maintained.
William
github mcewanw
Hi Toni,
I did get PeasyWIFI running, but not actually working as it turned out, otherwise I was hoping a deb could be made for that alternative. As I said, I don't know much about PeasyWIFI so hope its author might check MintPup out since I maybe just didn't install PeasyWIFI correctly.
William
I did get PeasyWIFI running, but not actually working as it turned out, otherwise I was hoping a deb could be made for that alternative. As I said, I don't know much about PeasyWIFI so hope its author might check MintPup out since I maybe just didn't install PeasyWIFI correctly.
William
github mcewanw