2 replies to this topic

[christophecvr]

Hello,

Like some maybe now i'm currently working at a SystemTools e2plugin.

By adding some stuff like swap file on some devices, I noticed that's it's difficult to make it perfectly working whitout terminal intervention for some things.
The goal is after all to have tools easy to use for end user whitout real knowledge about mountpoints and so on.

One of the problems is that the mointpoints are a bit asigned as wildcard's. This result in error's which are easely solved if user does understand what's happening and then solve this trough terminal commands. That last one is really not for everyone.

Can there not come a kind of general agreement to take some standard's about naming the mointpoint related to specific devices ?

Like for example for long time on dm a kind of standard was held by :

The usb sticks where mounted on usb , .... (the fact that now there are the longer the more usb ports avbl this is the most difficult one that's trough.)

But we had also cf, hdd mmc

Before the cf on dm was trough e-ide two. But the more recent boxes do have sata, and the sata one is normally for internal hdd the two for dvd. The cf is now an usb device.
mmc slot is also usb device.

Since normally on boxes in pretty complete category like dm8000 with multiple tuners and so on ..

there is one cf slot one mmc slot. a couple of usb ports.
internally mostly a sata for hdd and one for dvd.

The cf was mounted to cf by standard, mmc to mmc and ...

At this time some changes made that in some case the cf is mounted to folder called compactflash (located in media) instead of cf which was standard for long time.

All this changes are a bit related to the general use of udev.

Can we not try to make some standards to use stay as much as possible close to udev default values.

Udev's auto moint will standard (if You did not modifie source )

first use the partition lable if present then if this is not present use the parttion uuid if that 's not present or zero use disk for a disk like device if that's used by other device the device node of parttion self .

Can there not be an agreement to use for cf always cf or maybe compactflash, for mmc mmc and so on, ? This will avoid a lot of p.i.s. to all developpers and maintainers

[MiLo]

OpenPLi does not use udev (it uses mdev instead).

a CF card is mounted at "/media/cf", as you can see for yourself in the /etc/mdev/mdev-mount.sh script that is responsible for that. Since the word "compactflash" does not appear in that file, it will never mount anything under that name, unless you did something yourself like filling in /etc/fstab.

I'd be happy to improve things, but I have no clue as to what problem you're trying to solve. One could even go as far as to let enigma2 itself decide on the name of mountpoints, if that would help.

[christophecvr]

OpenPLi does not use udev (it uses mdev instead).

a CF card is mounted at "/media/cf", as you can see for yourself in the /etc/mdev/mdev-mount.sh script that is responsible for that. Since the word "compactflash" does not appear in that file, it will never mount anything under that name, unless you did something yourself like filling in /etc/fstab.

I'd be happy to improve things, but I have no clue as to what problem you're trying to solve. One could even go as far as to let enigma2 itself decide on the name of mountpoints, if that would help.

I rechecked, But there is still a problem after You format a cf,usbstick or mmc card. If it was already mounted. After format it is mounted to folder located in /media/ <dev node>

where dev node is /dev/sd.. the actual device node.

when You reboot the box after format it's mounted indeed to cf(for cf device) or usb(for usb stick) or mmc1 (for SD card)