Beta Mountmenu (Installer)

[lomaxx]

It's been quite some work for me. I struggled with bash, I struggled with Zenity, I struggled with the Pandora-OS, I struggled with vi ... and most of all I struggled with a scatter-brained myself and my lacking concentration. But after something like 3 weeks of rather intense work, I can finally release the test-version of Mount-Menu. It's just a small application, but might be helpful for someone out there.

Quote from the PXML-file:
"This package will install Mount-Menu, a semi-automatic mount-mechanism for both SDHC-card-slots of the Pandora. It can be configured to launch on card-insertion and can mount, unmount, do read-only checks and simple automatic repairs of filesystems. It was made to be an alternative method to the default, fully-automated mounting behaviour of UDEV. See Documentation for details."

It SHOULD be safe to use. I tested a lot of things, removed all bugs that I could find and never managed to damage any of my file-system with it. However as usual I can not guarantee anything as your system- and/or card-configuration differs and might in theoretically cause trouble. The only known issue that is left: If you mount a partition with Mount-Menu, then it does not work to unmount it by using xfce's unmount-option in the context-menu of the cards desktop-icon. Unmount it with mount-menu instead. That's what it's for.

download: http://userpages.uni-koblenz.de/~bbrink/tmp/mount-menu-installer.pnd

The PND will show up at Startmenu->System

Any feedback is welcome. Especially from people who are good at bash-scripting and who browsed the scripts.

 

[lomaxx]

I celebrate the 100th view of this thread without a single comment with a very small update. Fixed a small, harmless bug.

I knew that not many people would be interested in this small application, but still I am a little bit surprised that not even a single comment appeared. Probably it's not worth releasing this on the repo then. Or my precautious warnings scared everyone off.

 

[Emnasut]

Thanks, this looks interesting! Although I just stumbled upon it by accident, if it would have been in the repo I would have found it a lot sooner...

 

[Esn]

It was made to be an alternative method to the default, fully-automated mounting behaviour of UDEV. See Documentation for details.

Just to understand this a bit better... what is wrong with the default one?

(also, I added this to the software projects list)

 

[lomaxx]

I wouldn't exactly speak of right or wrong. It's more a matter of personal 'taste' how things are handled. I also explained this in the documentation that you can read before starting the installation, but here is my point of view on things again anyway. =)

The default mount-mechanism of the PandoraOS, which is based on UDEV, does always mounts the SDHC-cards directly on insertion or on boot. I myself don't want that. One reason is that, because I always run my OS from SD-card and UDEV in that case nevertheless mounts the card again, I was left with a system where the card was mounted twice. Once as root ("/") and once as /media/root (as I labled the card "root"). That's not dramatically as far as i can tell, but still I consider it to be a bit unclean.
Also i boot the system without the framebuffer-splash, so i can read the messages of the boot-processes. However the UDEV-mount-mechanism littered the boot-messages with quit a lot of messages because it tries to mount all partitions first as fat-filesystems (instead of "ext", which i use) and even tries to mount /dev/mmcblk0 (in opposite to the partitions themselves like mmcblk0p1 or ...p2). Again this is not really a problem, just annoyed me because of the error-messages.
Another reason is that, although i always keep the second card inserted, I do not always want to mount the second card. Either because i want to test it or because I don't need it at the moment.
Sometimes when i insert a card, i nearly manage to push it in completely, but than it snaps of my finger. At that point the mount-procedure already started with the UDEV-mount-mechanism and because the card was directly ejected by the mechanical spring, the mount-procedure is interrupted. It's unlikely that the cards file-system will be damaged because of this, as i doubt that during this period data is written to the card, but it always leaves me with a bit bad feeling. The semi-automatic concept of mount-menu avoids this.
There are a few more reasons mentioned in the documents.

While writing the scripts i then came up with additional thoughts like file-system-checking and -reparation. For that it uses two programs, that were not written by me, but are commonly used in pretty much any linux-distribution. However: Note that the repair-function TRIES to repair the card. It might happen that the filesystem is still readable before repairing the card, but completely unreadable afterwards, if the disk was damaged beyond repair. So it's important that you backup as much data as you can before trying such a repair. There are cases where i still risk to do a repair without a backup - when i check does not show any errors, but the filesystem was not cleanly unmounted sometime in the past. That's still my personal risk though.

If you got additional question, then feel free to ask.