Slashdot Mirror

Compared to other non-Microsoft SMB/CIFS client implementations, the OS X SMB client is deficient.

It does not support DFS, which basically amounts to having symlinks from one CIFS share to another (so a hierarchy of directories spread across many servers can be accessed via one namespace.) But Linux's cifs client driver can use it and Samba can host it.

Don't believe everything you read about Samba.

Its DFS support is shoddy at best and catastrophic at worst. It just fails to mount on Fedora 6 (samba 3.0.24-7, kernel 2.6.22.14-72.fc6), and causes a kernel panic on Fedora 9 (samba 3.2.4-0.21, kernel 2.6.26.5-45.fc9.i686). Of course, I am using Windows 2003 R2, which has improved DFS (which, is absolutely killer, especially the much better replication).

As opposed to the Windows paging system? Has the author used a Windows OS lately? Swapping is a *bleeping* killer! Especially when you have more than enough memory not to swap. :-/

No OS performs well while swapping heavily, but OS X seems to swap too often. I'm at 4.65 GB of swap used right now on a system with 4 GB of RAM, 3.94 GB of it used. The sum of real memory allocated by all processes is less than 3 GB. In my experience, this is worse than Windows or Linux. It's not rare for applications or the entire system to respond slowly for seconds at a time.

Windows lets you disable or limit swap easily, and Linux gives you lots of control over how much swap you have and how it is used.

Macs support CIFS/SMB pretty darn well these days. I keep hoping that someone will come up with a better replacement, but CIFS/SMB will continue to work until that day comes.

I disagree. Compared to other non-Microsoft SMB/CIFS client implementations, the OS X SMB client is deficient.

• It does not support DFS, which basically amounts to having symlinks from one CIFS share to another (so a hierarchy of directories spread across many servers can be accessed via one namespace.) But Linux's cifs client driver can use it and Samba can host it. Closed-source DAVE supports DFS on OS X.
• By default, the OS X SMB client puts desktop folders and dotfiles on remote SMB shares.
• The smbfs driver in OS X is known to cause kernel panics.

It's not like Apple is a struggling company who can't afford to improve their SMB/CIFS client. It's not like the required protocols are so difficult to implement that no one can figure it out. It's not like Microsoft is suing anyone who dares implement certain features.

It's not like they're ignoring the SMB client altogether. In Leopard, they finally introduced SMB packet signing (which other non-Microsoft clients have supported for longer, including Samba's userspace smbclient.) And performance seems to have improved; in unscientific testing with no attempt to improve performance on the Samba server or client, I went from 45 mbytes/sec SMB performance in Tiger in any condition to performance very close to the maximum possible given the disks involved - about 70 mbytes/sec.

Those of you who still have jobs, or who (like me) are lucky enough not to have a mortgage or any debt, do the decent thing and donate to the Samba project. You know it makes sense! (I just tipped ''em $100 and BOY! do I feel smug, self-righteous and validated as a human being! ;))

I must have missed the memo about Samba being able to serve as an AD domain controller - that's huge for me.

Read this :

http://www.samba.org/samba/PFIF/PFIF_agreement.html

for details on patent issues. It's not as black as you paint it.

Jeremy.

Samba is GPL 3 licensed. I think Microsoft would have a hard time with any patent suite in light of that.

and that after only 16 years!!!

Every so often, some filters will block samba.org as being in the "Arts & Entertainment" category.

Of course, if you listen to the folks in Redmond, Samba is just for entertainment anyway, not any serious work.

If you release your invention to the public domain...

I want to be clear on this, since there's a notable difference between copyright and patent here: If you publish the details of your invention without having first filed for a patent on it, then the invention is effectively in the public domain. It's a bit like losing trade secret privilege. Once an idea is known, it can't become unknown.

Now, aspects of how you implement your invention might be copyrightable, such as source code. You could have a completely proprietary application under full copyright protection. If you publish its techniques and protocols without filing patent disclosures first, then those techniques and protocols are no longer patentable. But, the actual code that implements the techniques and protocols is still protected under copyright. So, the invention is effectively in the public domain, but your specific implementation is not.

An example of this might be the SMB protocol that Microsoft uses for sharing files over a network. Portions of the protocol may be protected by patents. Had they instead chosen to document the protocol without filing for protection, then these aspects would become unpatentable, and everyone could implement the protocol without getting granted special exemptions. In either case, though, Microsoft retains the copyright to its Windows implementation.

I wanted to make this clarification, since copyrights and patents so often get confused.

(And then there's chips with their "mask work" protection...)

--Joe

For parallel compilation on unix check out distcc: http://distcc.samba.org/

One would think that "The easiest way to do it" would be to install Winbind, LDAP or Kerberos and use those to authenticate against AD.

The advantage here is that you're dealing with free software, included and supported by default in most Linux-based operating systems, and in many cases integrated so tightly that you only need to run one command and tick a few check boxes to make it work.

What does this third party solution add to that besides the $250 per seat price tag?

Checkout ccache, it caches compilation results and then on recompile it reuses the results when nothing has changed in the preprocessed file. Because it works on preprocessed files instead of the original source files it can ignore quite a few changes that would otherwise cause a recompile and thus can speed up recompilation a lot. It also happens to be very easy to use, just compile with "ccache gcc" instead of "gcc" and you are done.

"Boundary failure when parsing SMB responses can result in a buffer overrun"

Does this apply to a particular CPU/MMU compiler combination or is it generic across all systems? Is it technically possible to design a system that is immune to buffer overruns or, by default, fails safe, as in not allowing any old code to walk all over the address space.

Here's the assigned Common Vulnerabilities and Exposures - "Boundary failure when parsing SMB responses can result in a buffer overrun"

No, we did report it. The answer at the time was "this is allowed by POSIX, deal with it", can be seen in the bug report here :

https://bugzilla.samba.org/show_bug.cgi?id=4715

I did point out that no other POSIX system behaved like that, but that didn't seem to make much difference :-). Eventually I just added a parameter that allowed our open directory cache to be turned off on *BSD. Once it got into the hands of Marc Balmer he took us seriously and fixed the bug.

Jeremy.

People like you are the reason why, despite a solid technical foundation, BSD sucks ass. Finger point at Samba all you want. Whine, bitch, moan and complain about other people doing everything except sending you a stripper-gram to inform you of the bug. Your attitude is exactly the reason why so many of us have left BSD behind and moved on to better things.

Yes, Samba did pass on what it found and it appears they were promptly shot down by someone on the *BSD side.

The Samba e-mail archives contain a message from over 3 years ago, but it doesn't give attribution to the *BSD source.

The Samba Bugzilla also has a bug reported more recently involving the same issue. Reading through the bug history, you can see there was one FreeBSD dev involved in the bug discussion, and he referenced a prior conversation between Tridge (Samba) and PHK (FreeBSD) where PHK said there was no bug in FreeBSD.

Yes, Samba did pass on what it found and it appears they were promptly shot down by someone on the *BSD side.

The Samba e-mail archives contain a message from over 3 years ago, but it doesn't give attribution to the *BSD source.

The Samba Bugzilla also has a bug reported more recently involving the same issue. Reading through the bug history, you can see there was one FreeBSD dev involved in the bug discussion, and he referenced a prior conversation between Tridge (Samba) and PHK (FreeBSD) where PHK said there was no bug in FreeBSD.

I agree, it does seem like they're trying to imply that there's only a 1 out of 5 chance that anything related to the Samba technical detail licensing is patented.

Here is a relevant link:

http://samba.org/samba/PFIF/ - The Samba licensing announcement.

The announcement has a lot of ambiguities (and IANAL), but it appears hey agreed that:

1) Samba Team members would receive access to protocol documentation. This information would only be available to Samba Team members, and available only under NDA
2) Access to information would not restrict CODE that could be produced using this information
3) It does not provide any patent coverage.
4) However, Microsoft would provide a list of patents covering the protocols used by Samba, and keep the list updated. This provides Samba folks a way to understand exactly what methods to avoid - which infringe patents.
5) Microsoft agreed that any patents not detailed in this list and found to be infringed cannot be "asserted" by Microsoft.

Presumably, there are items that MS will provide for #4, so there are patents that relate to Samba.

Are you mad? Samba fought long and hard to get those documents. MS fought back in all but appearance, to the best of my knowledge.

You might find How Samba was Written interesting.

MS has NEVER done anything yet that is pro open source

You'd better tell the Samba people that. They think they've been given the documentation for the protocols they implement under a reasonable license which will significantly aid development:

    http://news.samba.org/announcements/pfif/

My main reason for using Linux is that I don't need it to be shiny if I don't want it to. It may seem a bit strange, but sometimes I like to be able to just sit back and stare at the white text, rather than spend an hour making a new style for XFCE, or looking through my images for the perfect background. More important than that, with a GUI, it's hard to display (verbose) information without someone crying 'bloat'; in a CLI, it's only going to push some old text up which you can request again at any time.

Close behind this (but probably more important than my freakish love of text) is the ability to share a single machine among multiple users at the same time, without even having to be near the machine, thus (potentially) getting much more use out of that single machine. Oh, and logfiles for when one of these multiple users screws something up, or when I just want to know if anyone's still using the OpenArena server.
Speaking of maximizing use, there are plenty of tools for setting up a cluster of Linux boxes to handle anything from rendering a short film to compiling in record time.

tl;dr - I'm crazy for minimalist interfaces, want to keep my machines loaded down when in use, I want to know when and how everything screws up, and I want my strange side-projects to be over quickly.

MS is obviously not going to give away filesystem specs or the other interoperability roadblocks ...

Really? You might want to ask Andrew Tridgell over at the Samba project about that. It seems he has gotten exactly that from MS. In such a way that he can show it to anyone who joins the Protocol Freedom Information Foundation.

Yes.

The Software Freedom Law Center got the protocol documents for Microsoft workgroup networking, which they were supposed to make available in 2004.

The EU agreement also weakens Microsoft's FUD about Linux and other FOSS violating its patents. They now have to disclose patents covering its workgroup protocols so developers will be able to show their code doesn't infringe.

Yes they're true. Please help us. See here :

http://samba.org/samba/devel/

for details.

Thanks

Jeremy.

Actually, the Samba team already received the documentation from Microsoft.

Here you go...

There are two properties of the HFS+ filesystem which rsync had problems with (that I am aware). metadata and ACL's. Neither of them are essential for data recovery itself and rsync had issues with them on any platform, not just osx.

It bears noting that the reason no one in the linux world cared about metadata until rsync began developing patches for it was due to the fact that linux didn't have them until 2.6 (or 2.5 for the brave). In fact, the idea of metadata (file attributes in linux parlance) was taken from osx's success and proven utility in using them. Reference: http://lists.samba.org/archive/rsync/2004-June/009937.html

Further as soon as metadata and acl support was available to the linux world, it was to the osx world as well.

The last niggling problem with osx is resource forks. There have been patches for resource fork support in rsync since at least 2002 (when rsyncX was copyrighted).

If I am mistaken about the features that are apparently lacking via HFS+ and rsync (and the fact that they existed for all users of rsync regardless of platform), please correct me.

Finally, we've just been talking about some unix utils that needed a bit of time to catch up with the current trends in filesystem design. There have been command line utilities for creating bootable restores and preserving metadat for a while in osx besides the ones listed above such as ASR and PAX.

I believe the mistake your making is to assume that an osx admin is limited by the pre-compiled software offered on osx by default, but this is simply not the case. No more then it is the case for any *nix install that leverages open source. I also find it peculiar (unless you bring to light issues which I have not covered here) that your major complaint about said backup solutions is that they could not back up extra information with HFS+ that was not even existent in prevailing *nix file systems (e.g. metadata). This is more of a reflection on the lack of features in said *nix solutions then in OSX.

I will now use the opportunity to promote my favorite hard link / cp based backup solution which I use for both Tiger and Panther backups without issue: BacupPC

Microsoft say their indemnity doesn't cover GPL3 software, perhaps someone from Novell can comment on that? Last I looked Windows interop is largely via Samba... Speaking of which, perhaps Novell are also in a position to comment on all the anti-GPL3 FUD we've recently been subjected to?

Yes, the issue is not so simple. It really depends upon the company, its situation in the market, and the like. But, generally speaking there is significant cost savings in using some things as open source. In the case of a small contract call center in my area, open source was the saving grace for the company. Their IT overhead was so great that the company felt it could not longer be competitive and was considering closing doors. Indeed, the IT department shrank to three people. But these three intrepid people replaced the proprietary Nortel Telephone system that was bleeding them dry on maintenance, support, and just plain babysitting with two Asterisk servers and SNOM telephones. The second largest expense was on the maintenance of their exchange server. So, exchange was phased out in favor of Zimbra. Zimbra was brought online in a week's time and has seen 99.999% uptime with only looking at the logs once a week versus babysitting an exchange server every day. This is not some case study, this is my friend that achieved remarkable results. Asterisk and Zimbra have put this call center back in the black. My friend does see some merits to proprietary, i.e. Active Directory. Simply put, he needs it to adaquetaly manage his workstations. He thinks once Samba4 hits a release, there is potential for phasing out the windows domain controllers. Soon, Windows will be relegated to a SQL server. My friend says that programmers are working furiously to convert to an *AMP solution.

Samba is open, it is based on reverse engineering of the SMB NMB CIFS protocol accretum. See: http://us4.samba.org/samba/docs/SambaIntro.html gg nm.

Besides, F/OSS systems still have quirks. For example, I deployed OpenFiler a couple of weeks ago, and I can't access it by its DNS name (\\foo\bar) but can if I use the IP address (\\10.0.0.201\bar). DNS itself works fine (can dig and ping.) I did some RTFM and Googling and found that the issue may be in the Kerberos library that needs to be upgraded by compiling from source... Clearly I'm not going to do that on a NAS appliance that is supposed to just work. A Windows server has no such issue, it just works. Anyone but a geek would just throw the OpenFiler out because the issues involved (Kerberos, WINS, time syncronization (NTP) etc.) are just way above the pay grade of an average technician.

Combine that with ccache and the pains of compiling large C/C++ programs will slowly disappear right before your eyes. :)

Especially when you can use distcc.

I remember hearing that Samba has some pretty elegant code. Apparently several projects like talloc have also fallen out of it. Later versions also have very extensive test suites...

I remember hearing that Samba has some pretty elegant code. Apparently several projects like talloc have also fallen out of it. Later versions also have very extensive test suites...

Mr. Shuttleworth is to be applauded for not bending to Microsoft. Bending to Microsoft is capitulating to the FUD climate and ultimately does more harm to Linux than good. Interoperability is a good thing, but at what cost? Have software patents and measly threats turned us all into scared little rabbits? I am not much of a Linux fan, instead favoring BSD, but I have to give credit for Mark Shuttleworth challenging Microsoft to put its money where its mouth is. These thinly veiled threats by Microsoft represent nothing more than a company in the beginning of its death throes. Microsoft is loosing its ability to innovate. Open source may actually save Microsoft and its own executives see it as nothing more than a cancer. Once Samba releases version 4 and the Open Change Project makes its first release, Microsoft will have a serious threat to its Active Directory and Exchange dominance. Face it, MS SQL server isn't as irreplaceable as Microsoft would have you think, Share Point Server is purely redundant, and Apache is the web server Howitzer. Microsoft has an excellent chance to open source its protocols, streamline its business model, and take advantage of all the free community development to work out the myriad of bugs and problems. Microsoft does not have the problem of market penetration so, by open sourcing its protocols and using its marketing machine, there is no serious threat to long term profitability. Conversely, its products would be made that much better.

Thats Jeremy Allison. Perhaps you've heard of him.

Andrew Tridgell's rzip wasn't on there either.

http://samba.org/junkcode/

Tridge is one of the smart guys behind samba. And rzip is pretty clever for certain things. Just ask google.

Want a Windows Home Server? Load a copy of Linux/*BSD and Samba on to a spare PC. There you go, all the power of a basic domain without all the costs associated with an M$ product.

Oh, the EU has fined so many companies for price fixing, I don't even know where to begin--Bayer & Chemtura, Siemens, Dow, escalator firms, Heineken, Aventis, animal feed companies, the Deutsche Post, many vitamin producers, Nintendo and, of course, the well known case of Microsoft.

I'm not saying that none of these fines are unjustified but I am saying that, if I may opine, the EU has been issuing a lot of fines. With this recent Apple one, it does seem as though Apple had no choice and if they aren't given an alternative to losing their contracts with record companies for the sake of running one Europe encompassing store, then I don't blame them. On the surface, the EU Commissions seem to be discouraging big businesses from selling things like XBoxes, PS3s or iTunes inside all of the countries. Is this a good thing or a bad thing? I guess time will tell ...

And there is a host of companies out there getting paid to do Samba support:
http://us1.samba.org/samba/support/us.html

* "PERSONAL FABRICATION: A Talk with Neil Gershenfeld"

* Democratizing Innovation

Bring on the future, where things like fab@home are in every home, where people no longer have to wait for companies to develop products, the people as a community develop them together, with the same spirit/philosophy of FOSS.

I don't want a Win/Mac box, I don't care how easy either of them appear, I want a free and open source box and neither Win/Mac provide me with that freedom. Here's a brief article I recommend everyone read:

The Land of "Nothing for free" by Jeremy Allison .

The fact that our society today is filled with people who would rather consume than fiddle is one of the reasons why gas guzzling cars with proprietary internals are still used by the majority. Eventually this will all change as people will more easily be able to develop their own hardware themselves (think something like fab@home in every home) with free/open hardware designs shared and improved upon.

The question is: do you want to support the FOSS movement or do you want to support companies who provide closed source software? I don't care if hardware from Microsoft or Apple can run Linux, I don't want my money going to either company, period. If other people enjoy tinkering with said hardware, cool. I believe we all should (and will, eventually) be developing hardware on our own. Those who would respond with, "I don't care about all that, I just want X,Y,Z" are the focal point of blame. Unwind the philosophy from the person and the soul is nothing but another bag of peas to scan at the check stand for Company A,B,C.

Confirmation that you're an idiot just came through on the Samba mailing list:

-----Original Message-----
From: samba-bounces+xxxxx=xxxxxxxxxx.com@lists.samba.org [mailto:samba-bounces+xxxxx=xxxxxxxxxx.com@lists.s amba.org] On Behalf Of Jeremy Allison
Sent: Monday, December 25, 2006 12:03 AM
To: david rankin
Cc: samba; opensuse
Subject: Re: [Samba] Re: [opensuse] Open-source leader leaving Novell forGoogle

On Sun, Dec 24, 2006 at 11:11:18PM -0600, david rankin wrote:
> Jesus,
>
> I hope it isn't true. But if it is, we will miss him dearly, both
> here and very much so on the Samba list. However I can under stand and
> respect the decision. Good luck and God speed Jeremy. Jerry, can you
> pick up the slack?? An ill wind blows for us all as a result of the MS deal....

It's true I'm leaving Novell, but why do you think this means I'm not going to be on any Samba lists ? I'm joining Google on 2nd Jan, and believe me when I tell you they're *very* interested in me spending all my time on Samba :-) :-).

As Herb once said to me, "Same job, different office" :-) :-).

Jeremy.
--
To unsubscribe from this list go to the following URL and read the
instructions: https://lists.samba.org/mailman/listinfo/samba

Well, they took a first step with this, but how much further will they be willing to go? What does SAMBA stand to lose if they lock out Novell?

"Incorrect. Ballmer stated that the purpose of the Novell-Microsoft agreement was to protect customers against patent litigation (from any possible company). Microsoft made no such statements that they would sue Linux users. If you have a quote from Ballmer or any other Microsoft representative to support your ignorant claim, please provide it."

God I hate shills. Sigh ok then.
http://news.samba.org/announcements/team_to_novell /
http://www.groklaw.net/article.php?story=200611032 01234813

Note the quote from Ballmer:

  "Microsoft CEO Steve Ballmer said his company is open to talking to other Linux distributors about reaching mutual patent coverage deals similar to the agreement signed Nov. 2 with Novell.

Such talks would be a good idea, Ballmer suggested, since now only Novell's SUSE Linux customers are the only Linux vendors that have any assurance that Microsoft won't sue for patent infringement...."

"Done. Here's the first relevant link: (the first link was actually an article about how Apple (a litigious company itself) sued Microsoft over user interface issues: "

Ah the classic astro turfers tactic. When proven wrong try to blame apple!. The first link I got was about how MS sued people over their FAT patent and how the patent was challenged in court and eventually upheld.

But hey no need to face reality when you can pull out the apple wookie heh?

"Did you actually read the article? It clearly disproves the claim that Microsoft sues over its patents."

But it did sue. It sued over FAT patents. It also threatened many companies over other IP and they stopped without going to court. Hell MS sued a 16 year old kid over a domain name!

Well, for all intents and purposes, to anyone that really believes in FOSS and is informed about the deal, Novell is now a pariah.
I've lost count of the number of people calling for a boycott, or reporting that they have switched away from, or are in the process of switching away from Novell products.
I think that it is essential that this is continued. The community is the strength of FOSS. If we cannot stand together against what in essence is a form of corporate blackmail Microsoft will continue to drive wedges into the community. It's classic divide and conquer tactics.

We need to continue to spread the truth about this deal so that people have the information they need to see it for what it is, and shun Novell for he traitor in our midst that they have become. Hopefully Novell will come to their senses and abandon the deal. If not, the boycott needs to be as absolute as we can make it. We cannot allow stabbing the entire community in the back to be profitable. Currently, Novell is the new SCO, and should be treated as such.

http://www.groklaw.net/article.php?story=200611030 73628401
http://technocrat.net/d/2006/11/2/9945
http://www.softwarefreedom.org/news/20061109a.html
http://news.samba.org/announcements/team_to_novell /

I moved to Linux in 1994 as my primary desktop and server OS. About three years ago I decided that I wanted to produce some video content. Video editing was theoretically possible in Linux - I hooked up my camcorder to my Linux box and did some editing, but the tools were primitive and cofiguration was unusually difficult.

Eventually I looked at OS X and iLife. I decided to jump to a Mac. What a great move!

I found that Linux made it possible to do some things, but OS X made it simple to do them.

Fast forward a few years. I now have a few macs at home - their licensing policy makes it affordable to have several machines and a five user license for the OS and tools. My family loves the power and usability of the Mac.

Recently my linux server at home began acting a bit flaky. I did some analysis and determined that hardware replacement was needed. After checking prices for CPU/motherboard/RAM (and potentially hard disk) I figured out that I'd need a few hundred bucks to replace the CentOS box with a new one. After thinking about whether to drop a few hundred bucks or not on this server, it occurred to me that I might be able to move all of the services hosted on linux to OS X.

I found that samba,
  hotwayd,
  dansguardian,
  uw-imapd,
  fetchmail,
  procmail,
  spamassassin,
    rsync,
  rsnapshot,
  apache2,
  MySQL4,
  PHP,
  perl,
  java, and
  squid were all available for OS X.

Most of these are "in the box" with OS X. The only ones that I need to compile from source are uw-imapd and squid! Of course I need the bundled developer tools to get a compiler, and the Apple/BSD startup mechanism and the netinfo wierdness require some tweaks - but since when did Linux *not* require any tweaking?

What this means to me is that after more than a decade of running Linux at home (and work) I am *this* close to shutting down Linux for good at home.

Hope your experience is similar.

Regards,
Anomaly

PS - I share your recent comments about the loss of a pet. :(

Ah, the beauty of reverse engineering. Or the "french cafe" technique: http://www.samba.org/ftp/tridge/misc/french_cafe.t xt

Wrong question - because it is not what they are doing, actually. Let me translate Microsoft's offer: there are patent problems with linux. That's what Microsoft's offer means, no more, no less. A subtle, distressing and unfair FUD machine. Your question is understandable, because they offer doesn't make sense at all, unless you examine not what it says, but the message it conveys. That message is clear: linux might be encumbered with patents belonging to MS.

It is such a pity that Novell has become a partner to this for perceived short term gains. No wonder that the free software community is up in arms (ranging from groklaw through Perens to the Samba team) - MS simply tries to single out commercial linux companies to support its own FUD propaganda. They offer these distributions a new tool to compete with: patents. So far, commercial linux distributions competed on two fronts: technical excellence and quality of support and services. Even Oracle. Novell, by accepting Microsoft's offer, introduced a new tool: patents. This is against the spirit - if not the letter - of the GPL, which tries to enforce a level playing field, and was successful until the Novell-MS deal it was successful. (That's the main gripe of the Samba team with Novell. Microsoft is fishing for others now.