Scan a fixed directory for .sfs, .3fs, .swp, .iso at boot

[sunburnt]

Someone in another post said they had success using a newer "fs" type.
There's a lot of development going on at Debian & other "houses" along these lines.
I haven't kept up with all of it, but time permiting I'll do an assessment of it.
I'm hoping that this problem & many others will be solved in this next gen. of driver handlers.
My knowledge of Linux modules & how they work is at a bare minimum.
The finding the USB drive problem could probably be solved with some scrutiny.
It's not a mounting problem, just a recognition problem, that's why hacking the code worked.
When Puppy booted to only mounting the USB & nothing else, it work very well.

My troubles in Puppy with newer USB devices that are only for Win2K & XP may not be so easy.
I have a Mp3 player & a Kodak 3.2mp camera that Puppy won't recognize.
Win98 will do the camera, but the Mp3 player doesn't have a Win98 driver.
I don't know enough about Linux drivers to fix this, installing a new module is easy.
But which modules to use for the enormous amount of USB products out there?
And how does Linux decide which module to use for each paticular device?

[Dougal]

Someone in another post said they had success using a newer "fs" type.

What does that refer to? what do you mean ""fs" type"?

I don't know anything about modules either, never needed any

[joki]

Joki: I thought you wanted to have /pup1, /pup2 etc. directories, so I only made it one dir deep.

yeah, i originally thought 1 deep just a /pup directory to scan for on every media. no version either. but looks like you've managed to do that

thankyou for the info and specifics. will it get into future releases? hope so

[sunburnt]

Dougal; I'd have to look for the post, but it was a newer replacement for devfs I think.
I hadn't looked at the activity in the /usb dir., just the /devices file that gives lots of info.
I'm not sure what the 00x files are for, so I don't know if they're for mounting.

The problem is that just the card reader is recognized not the cards in it.
I tried the brute strength method of just attempting to mount all the /sd devices.
That didn't work, but the dmesg has part of it's report that lists the partitions on the USB.
The problem was that my PC didn't reliably give that info, & other PCs varied as well.
It came down to writing lots of exceptions for possable info that may or may not exist.
That's when I shelved any further development until I find another way to do it.

usb-auto works well for single cards in the reader & most other USB devices.
Since then I've learned that the mount command can mount "synchronus" or "asynchronus".
So if that works as advertized, then the sync command in the timed loop isn't needed.
At some point I may release a new version with that improvement.
I'm hoping that I'll find a better way to do this, making the old usb-auto obsolete.

Right, I think the "uhci_hcd" module is one of the drivers for M.B. USB controller chips.

[Dougal]

The problem is that just the card reader is recognized not the cards in it.
I tried the brute strength method of just attempting to mount all the /sd devices.
That didn't work, but the dmesg has part of it's report that lists the partitions on the USB.
The problem was that my PC didn't reliably give that info, & other PCs varied as well.
It came down to writing lots of exceptions for possable info that may or may not exist.
That's when I shelved any further development until I find another way to do it.

Does MUT recognise the cards in the reader? Maybe you should look at the MUT code. That Jesse fella seems to know what he's doing...

[sunburnt]

I've looked at Jesse's code but I"m not sure if it'll pickup multiple cards.

I don't have a card reader, but I'm getting one for the daughter's Mp3 player.
So currently I have no hardware to continue testing my code on.

[Gn2]

Random_Ramblings & Related_ Rituals:
Re any "Wizardy"
*******************************************
EG ~ Why does it seem best - to use dmesg parsing - vs perhap a script to probe applicable bus locations

Reasoning : If DMA is OEM implemented - & ISO standards adhered, then - device output should give all relevant capabilities

The changes from kernel 2.4_xx Devfs to 2.6_xx Udev altered hot/cold-plug related probes
This generally suggests the generic Scsi character F/Sys interface be used.
(sg/pg utilities queries)

http://www.google.ca/search?hl=en&ie=UT ... ds&spell=1

http://www.die.net/doc/linux/man/man8/sg_map26.8.html

http://www.enterprisenetworkingplanet.c ... hp/3635686

HMmm ~ Biggest headache seems to come when any device does not have a partition file table
OR - how to be persistent in device naming between re-mounts.
Esp if another is then used as well (first found "claims" resources)
Growls - Distribution inconsistency to views on mount points utilizing badly chosen "symlinks")

EG >> "NOT a Block Device" = How many can create a device (MAKEDEV/MAKENOD)
> True block device vs File system devices

http://www.google.ca/search?hl=en&ie=UT ... od&spell=1

Use of /Proc task numbering is not feasible - they are dynamic to system events !
(top output < > ps ax < > system event logs)

Mount parameters : Sync & Async options > Relate to writes to device:
Default behaviour is to wait for idle CPU cycles (async)
Sync means "write through"- but all is regulated by - defaults chosen Re kernel latency prioritizations
in defined queues awaiting forked parent process thread or:
Altered slightly by use of "nice"
Defaults to interupt process syscalls () are configurable - NOT recomended if not fully aware of how all work !

http://www.win.tue.nl/~aeb/linux/lk/lk-3.html

http://lwn.net/Articles/201682/

Lastly -if a device is open to view by any utility - then unplugged prior to umounting > Corruption may occur.
Pity no present easy method to stop physical removal such as when a media "tray" is used

(I know, all very easy to say) > how to best apply anything.

Or - is it better (design philosophies) to attempt automated wizards > adding yet another layer hiding what is taking place-
Vs including pointers to comprehensive documentation on use of any Linux administration task.

Hackers love playing with code - hate to document.
**************************************************************
" Up in the sky - is it a bird or a plane.... NO it's Super-Dawg Goofy":
> In his scratchy red flannels crashin' & burnin' agin'

I have no solution- but sure admire the problems

YMMV

[kleung21]

Well, just to revive the threat for this and maybe have Barry add back the pupsave variable to the bootup cheatcodes

I often wish to specificy which copy of the pup_save I would like to use in testing and as it is right now, I have rename or move things which is a bit - awkward.

Any help would be appreciated.

Thanks,

Kevin

[Dougal]

I often wish to specificy which copy of the pup_save I would like to use in testing and as it is right now, I have rename or move things which is a bit - awkward.

Why do you need to rename them? If you have more than one on the same partition it gives you an option to choose between them.

[kleung21]

You are right that if you have more than 1 pupsave in a partition, it gives you a list to choose from BUT; I prefer that the load up process be automated.

Anyways, I'll take a look at your init script or just hacking away at the init script so it works.

For me, the importance is that I have an older laptop with one of those loud hard drives. If I run off of the USB key only (and ram), it will keep the system much quieter.

Thanks,

Kevin

[pakt]

Well, just to revive the threat for this and maybe have Barry add back the pupsave variable to the bootup cheatcodes

I often wish to specificy which copy of the pup_save I would like to use in testing and as it is right now, I have rename or move things which is a bit - awkward.

It may not be that hard to add say a 'psave=' boot option to init and just have it look for that particular pup_save file on the machine. If it's not found, the normal menu would show (if there are multiple pup_saves) or just use the one available. I would find that useful too

Paul