Slashdot Mirror

I'm sure RedHat wouldn't mind taking over exclusive maintenance of the kernel. Maybe they have a developer or two that are finished working on certain mature projects. I guess they'd rather have them as kernel developers than what they are doing right now.

Just kidding (maybe)

strace -p $PID_OF_DBUS_DAEMON

As for your network, it's probably something related to this http://avahi.org/

That may be useful to some, but I always assign non standard ports. I have a HTTP server, but port 0 is not forwarded. Even when I had one computer, I used nonstandard ports. Now I assign ports based on what PC is running the service, so one PC may get ports 11xxx, another 12xxx and so on.

That's fine for you, but imagine having to explain that setup to your grandmother. With IPv6, all she has to do is plug the device into the network, everything else is auto-setup. (in case you're wondering why she'd need something like that, just think about VoIP and video chat applications).

Well, I didn't know that. This is good, but having two or three IPs per host on the same network makes configuration a bit difficult

No it doesn't, because that's done automatically. You just have to plug the device into the network.

especially since (again, I don't know how v6 handles this) if DNS returns all addresses, the host may pick the public one to connect to an internal host.

That's what DDNS is there for (ahavi and similar implementations). You only have to remember the host name, the lookup is done for you without any server whatsoever.

Especially since v6 IPs are harder to remember.

Again, DDNS. Trying to remember IPs is so 1990ies.

Also, those applications that send the IP address as payload are badly designed. The IP address already is in the header of the packet, why send a copy of it in the data section?

I agree on that, but not sending it in the data section still doesn't help with NATs.

Yes, FTP does this wrong, though I used a few FTP servers and they could find out the external IP by either DNS or some other method.

The original idea behind that was that you can hook up two FTP servers to each other (one passive one active) and copy files directly. Nowadays FTP servers don't allow that any more, since it's a bit of a DoS vulnerability. I personally consider FTP to be deprecated, there are much better protocols now, like WebDAV.

IM, like Skype, already just works, I just need to forward the required port (the same amount of configuration as would be to allow that port in the firewall).

Port forwarding is a huge hack that requires the user to do security-related stuff that shouldn't be touched by a layperson. Skype has a different solution: When you actually have reasonably fast direct internet access while running Skype, you act as a supernode. This means that you act as a proxy when two NAT-using persons want to talk to each other. If you pay for your data transfer (and you always do, even when it's indirectly), you pay for two persons you don't even know being able to audio/video chat.

Evidently, it's already in the works.

Bonjour for windows is Apple's port of zero configuration networking for windows. My debian etch box at home has some zeroconf networking; it's called avahi.

Zeroconf is a good thing and should be embraced. There's a book on zeroconf by Daniel Steinberg and Stuart Cheshire, published by O'Reilly.

I'm not much of a fanboi, but Apple contributed a few interesting things to OSS, used today in most Linux distributions :

- the Objective-C front-end to GCC, essential to GNUStep
- Bonjour (aka Rendez-Vous), although most distributions have switched to Avahi.
- Lots of patches to konqueror's back end (the same as Safari)
- Darwin, by itself quite significant and in particular allowed the development of the hfs+ driver in Linux.

There are other examples.

Apple has also done some serious innovating in OS X: expose, ubiquitous zeroconf, system services, launchd, etc. What I find interesting is that major Linux developers don't copy those features from OS X. [emph mine]

Let's have a look at your examples.

Zeroconf - Avahi
Expose - Kompose
System services - errr, that's been replaced by Launchd on your list.
Launchd - ReplacementInit - Considered Apple's LaunchD and Sun's SMF amongst others, but none were quite the right fit.

I have several theories as to why this is.

Out of the three feature examples you've given, two are implemented in linux, your premise is incorrect, your theories are utter bollocks.

Bonjour for services hasn't taken off in a big way on Linux (or Windows outside of Adobe Products/iTunes for that matter). However, most Linux distros now ship with Avahi which is fairly mature but there are comparatively few programs that can use it (its main use currently seems to be for autoip configuration). Some distros also firewall it off by default (but Kubuntu isn't in that list).

I've noticed music programs (Rhythmbox, Amarok) often support it but they are trying to interoperate with iTunes which is another issue again.

By the way I think someone said they might work on a Kopete bonjour plugin a few weeks ago.

I'm also a little sad that OSX has dropped default support for printers annouced over CUPS broadcast but thems the breaks. If you know what you're doing it's possible to renable it (and set your Macs to broadcast too but that's another story).

Bonjour for services hasn't taken off in a big way on Linux (or Windows outside of Adobe Products/iTunes for that matter). However, most Linux distros now ship with Avahi which is fairly mature but there are comparatively few programs that can use it (its main use currently seems to be for autoip configuration). Some distros also firewall it off by default (but Kubuntu isn't in that list).

I've noticed music programs (Rhythmbox, Amarok) often support it but they are trying to interoperate with iTunes which is another issue again.

By the way I think someone said they might work on a Kopete bonjour plugin a few weeks ago.

I'm also a little sad that OSX has dropped default support for printers annouced over CUPS broadcast but thems the breaks. If you know what you're doing it's possible to renable it (and set your Macs to broadcast too but that's another story).

First do you have a personal firewall on either side ? Look into your iptables on the Linux server side, do "sudo iptables -L" and make sure the rules are empty. If not stop the iptables service. On the mac side look into the "Sharing" system preference tab and disable the firewall.

Next, have you tried to export your Linux disks as Apple File Shares? I do that now instead of NFS, it works fine.

You need to setup your exports with netatalk, and advertise them like "bonjour/zeroconf" shares through Avahi. Works great, easy to setup. Your disks will show up in the Network->My Network section of the finder. Both avahi and netatalk are most likely already available via your favourite distribution.

Last thing, is there a route between your two machines ? If the mac mini connects to your network wirelessly, this can be a bit challenging to setup. Can you ssh to one machine to the next ? Either way ?

Best of luck with your settings.

No one can tell if this affects all systems with mDNSResponder, there's just not enough information being released. What I want to know is: could this exploit Avahi as well?

I don't know what project the parent post was refering to, but it is not only GIMP (with some interesting ideas) that got rejected.

Other projects that were not selected include interesting improvements to the desktop infrastructure, such as GStreamer (list of ideas) or Avahi (list of ideas).

I don't know what project the parent post was refering to, but it is not only GIMP (with some interesting ideas) that got rejected.

Other projects that were not selected include interesting improvements to the desktop infrastructure, such as GStreamer (list of ideas) or Avahi (list of ideas).

So in the end you have no response to the OP except to say "I don't care about that feature, so people who do are stupid." Bzzzt. Thanks for playing.

If you actually knew much about where Linux on the desktop is going, you might have mentioned AVAHI.

I have setup a Linux server to server to both Mac and PC clients on the same volumes/shares using AFP with the Netatalk package, and SMB with Samba. Netatalk, in its new incarnations is by far the best non-apple AFP server available. It works seamlessly with modern OSX clients (10.3 and 10.4), supporting precomposed UTF-8 charactersets, long file names (most commercial NAS devices still only support the ancient appletalk implementation with 32 MacRoman charactersets and glacial unreliable performance) and even Bonjour/Zeroconf support.

Netatalk works surprisingly well with modern Samba versions (post 3.0) that support UTF-8 (and now even includes a netatalk module to ease compatibility), and both samba and netatalk hide one another's specific data from the other so that resource forks are kept and if the mswindows option is enabled in netatalk, the worst character problems (?\ etc in filenames) are safe.

What I would really love to see is a system that reliably combines these, PLUS NFS for Linux shares. The FreeNAS looks good, but seesm to be a bit on the young side without decent Mac support, and god knows there are enough Mac using companies that don't want to have to fork over money for XServes.

As far as the product, hasn't Microsoft, Novell, and an ungodly amount of other smaller companies tried to do this before?

Novell has historically not been strong on IP networking; more recently they've figured out that IP is the way to go, but I haven't heard of any cross-platform, open-standard, widely-supported IP-network technology from them. Or from Microsoft, for that matter. (How many UPNP printers can you name?)

Has anyone used Bonjour?

Only pretty much every Mac user (Safari, iTunes, iPhoto, iChat, ...). Oh, and lots of GNOME users. And maybe a few Windows iTunes users.

What's network traffic like? ActiveDirectory and Novell are both rather chatty applications when it comes to the network.

It uses caching, duplicate message suppression, and exponential backoff. Traffic is unnoticably light.

If we can find a way to keep things quiet, this is a great idea. However, there's the challenge.

Good thing those engineers at Apple figured it out 5 years ago, then!

Zeroconf is the only service of its type that I've heard of. It's certainly the only one that runs on pure-IP networks, whose standard is open, which has multiple independent implementations, which has support from both proprietary and open-source camps, and is supported out-of-the-box by many major hardware manufacturers. If there's any competition in this area, I don't know what it is.

This is also available for linux, but I haven't gotten it working properly (or really tried, for that matter). I believe the packages are called Howl and mdnsResponder.

I have a Fedora Core 4 system that advertises Netatalk shares and HTTP via mDNSresponder. Fedora 5 has dropped Howl in favor of Avahi -- another zeroconf implementation -- though I haven't done anything with it yet.

Well, as you can probably guess, it is not only harder to write such an in-depth review, but it also takes more time. So by the time such a review would be released, it would not matter anymore, because FC5 would be close to its end-of-life.

Also, such review would be biased anyway: Hardware support? what hardware? the reviewer cannot have all the hardware FC5 runs on. Performance? I think few percent difference do not matter for most people, and you can sacrify this barely visible amount of power for something more interesting (like virtualization) without anyone (but a review writers:-) noticing. Yes, people would probably be happy they can brag about how their distro runs an infinite loop 5% faster than the other distro :-)

People choose their distributions not by reading some review, but based on whether their friends use it, whether they like the packaging system (RPM vs. Deb vs. plain tgz vs. portage), whether they like the process under which the distro is being developed, whether the distro provides updates for long enough, whether it contains stable (old, boring:-) software, or a bleeding-edge technologies, etc.

I agree with you that the review in TFA is meaningless, and nobody should choose the distribution based on screenshots and the install process. However, I don't think a better or "a real in-depth" review can be written. It is too personal and requires maybe years of daily usage to discover all the pros and cons of the particular distribution. The only thing the reviewer can do is to point out the new technologies behind the distro (such as Avahi or Xen), but that's all. If people are interested in "what's new and interesting" in a distro, they should read the release notes.