Linux Community's Journal|
[Most Recent Entries]
Monday, July 19th, 2004
|samba and permissions
At least I feel my questions are getting more intelligent. =)
Setting samba permissions in the smb.conf for the homes directory, this is what I have
comment = Home Directories
browseable = yes
writable = yes
create mode = 4755
right, so this works great and any files that get created in that top directory have the perms of 4755 applied to them and all is hunkydory. So any time I actually create a file there it has all the right permissions and all that crap. If however I move or transfer a file into this directory it retains its original permissions when in fact I want it to be dumber than that and just assume that I'm making something new and assign it the 4755 perms like it would if i were actually making a file there, locally. Does that make any sense? The only thing I can come up with is a cron job that goes off every 5 minutes that chmods the whole dir and brings everything into spec. That seems a crass way to go about it if you ask me. Is there a more elegant way that I'm missing? Something in fstab mount options perhaps?
In specific, the ability to upload files to their ~/public_html directory via samba share for online viewing needs this permissions functionality if possible. I also have a media server partition (/srv) that many people access locally... Needs the same treatment so everyone has at least read access to it.
A large part of the reason this is important is because the majority of my users are not linux people, don't know the first thing about linux permissions, and its not very realistic of me to expect them to chmod this and chown that... It also feels like this is my job as an administrator to have this kind of stuff squared away so they don't have to mess with it.
Thanks all for any possible solutions you can provide.
|Articles To Help Promote Free Software
I had been aggregrating some articles that would help evangelize Free Software to different organizations. It seemed like whenever I tried to just save the links that they would degrade over time. I felt it necessary to capture the entire article. I started thinking perhaps others might be interested. I don't keep very on top of it, but try to when I get a chance. You can see them at fsarticles
Some of you may recall my regularly repeated GUI crash trick. it's amazing and stupendous but it doesn't bring in the bucks like pulling a rabbit out of my shared libraries so I thought I'd dump it.
Origionally I thought it was because of hyperthreading, but I don't know that that's the case anymore. I think i've narrowed it down, so let me give you the rundown.
When I change from my GUI (windowmaker, though it happens in anyhting that is XWindows or similar) to a command line window (ie ctrl-F3) and
(usually) there is a window or series of windows actrive (esspecually if they are changing like an Opera browser) then the problem can usually be assured to happen (if not the first or second time soon after that. Sometimes, when this occurs, I have to ctrl-F8 or ctrl-F9 to get back to teh GUI (it moves up).
Thoughts on repairing this?
|Installing Linux to SATA Drive
I'm working on installing Debian to my new system using a WD Sata Raptor Drive... I realize I'm going to need kernel support for it to install it, however I don't have another IDE drive to install to.. any ideas? I blew away Gentoo due to it breaking all too often and their project team is all out of whack... and i refuse to use RedHat due to the whole (too much like MS issue), and slackware.. well i'm just not into it.. anyone successfull or have a nifty technique to install Debian to SATA?
|Woe be SCO
SCO, shorthand for The SCO Group and not related to the original Santa Cruz Operations, have placed themselves in quite a bind. For over a year now, they have been blasting the media about "millions of lines of source code" in Linux being theirs, and how there is this vast mountain of proprietary code improperly contributed to Linux. They've claimed to have incontrovertible evidence about this (still do, in fact, in Germany, violating the spirit, if not the exact wording, of a court injunction against them there).
Well, seems they've got one-upped by IBM here. Not surprising that the Nazgul lawyers managed that, but rather amusing to watch. The following chain of events has taken place:
- IBM asked SCO, as part of discovery, to provide a list of ALL Unix V code found in Linux, regardless of who put it there.
- SCO signed off under oath saying they have complied with all IBM discovery requests.
- IBM presented the list of Unix V code to the court, asking the court to declare that Linux is free from Unix V code and that there therefore exists no copyright controversy *at all* between Linux and Unix V.
SCO is currently in the process of throwing a hissy fit over this. Who knows, maybe they'll succeed in divert this oncoming freight train, but they have placed themselves in a rather marvellous bind.
Go David Boies!
|The SCO case and Linux
It appears, judging from responses to my last post in linux
, that there is still some confusion over what legal theories SCO are invoking, and what they actually can get through them.
SCO are claiming several things. The most interesting claim is that they are entitled to royalties from all Linux users, inclusive. The question then is, under what legal theory is it possible to make that kind of claim? The only theory this is possible under is the so called "intellectual property" theory.
Intellectual property, or IP (not to be confused with the network term) is actually a catch-all phrase for several very different legal theories; namely Copyright, Patents, Trademarks and Trade secrets. Let's look at these, one by one.Patents
No claim about SCO patents in Linux have been made.Trademarks
No claim about SCO trademarks in Linux have been made.Trade secrets
Trade secrets are only protectable by contract, and leaked trade secrets only provide allowance for relief from the party leaking them. Once they are no longer secrets, they afford no protection - and most certainly no need for licensing and royalties. SCO has posturized some about this, but for purposes of royalty payments it is irrelevant.Copyright
This is what IBM has cornered SCO as telling the court they have no claim to regarding Linux.
These are the only legal theories (outside of individual contracts) where royalties have to be paid. If IBM loses every single point in the case, but the declaration that there is no SCO copyrighted code in Linux is passed, SCO has no right what so ever
to royalties from anyone, except those it has a contract specifying royalties from with.Any other information, from SCO or any other source, is FUD. Please help stop the FUD by only citing FACT and not unwarranted speculation helping SCO and undermining lawful use of FOSS software.
Disclaimer: IANAL, although I could be playing one on TV. If you need legal advice to decide how to handle SCO knocking on your door, pay a lawyer for it. If you have any corrections to this post, please do provide them; it is intended as anti-FUD, but I am merely human and it may contain mistakes.
Needless to say, I finally got my Ipw2100 card working with the Intel driver. However, I can't seem to get ACPI to work propperly...Every time I put the machine into the S1 state, I can't seem
to wake it up...Has anyone else had a similar problem? My only recourse is to hold down the power button, wait until the machine completely shuts off, and then reboot. I also can't seem to get
out of S3 propperly either. I end up having to reboot as well after getting out of S3, since I lose my Speakup review keys (Speakup is the screen reader I use). Is this situation somewhat
improved in 2.6.7? I'm using 2.6.6 right now. Also, any Inspiron-600M users know how to get the machine out of S1? Because hitting the power button doesn't seem to work. Thanks!