OSNews

Considering the iPod can be accessed as a removable storage medium, no different than an external HD, I think you're on the right track.

I'd just like to point out, it doesn't even have to be AAC. The iPods will play mp3s just as well as anything else. It'd take 30 minutes of a decent programmer's time to get this kind of thing rolling, and the payoff (could be) huge. Good point you made, and I hope the Helix/Real folk listen. I've never been a fan of Real because of their prior mis-steps, but if they'd get on the right track and prove themselves as being a new/good company, I'd stand behind them just as I would any good technology. Here's to hoping they take your advice and carry it across the goal line!

OT: Your kind of missing the point I think. Apple's iPod has its own database format. You can drag and drop as much mp3's to your removable harddisk ipod, but they'll never show up.

Yes, but you can read that database format using libgpod, which is used by a number of different music organisers for Linux (which don't compile under Solaris - damn OSS programmers and their view that the whole world revolves around Linux).

As for AAC; all my music is already in that format, I sure as heck don't want to re-rip 9gigs of music again, thank you very much!

The db format is well documented. Again, 30 minutes of somebody's time, and it's a done deal. Maybe an hour if you want it to be bug free. If I were more versed in Solaris, I'd attempt to hack it out myself just to prove it could be done. Hopefully somebody will catch on.

Not even 1/2 hour would be required; just integrate libgpod and realplayer together, extend realplayer functionality as to allow the playlist to be understood, and from what I understand, recording is being implemented in the next version.

<shakes head>

Solaris 11 is known as 'project Nevada' which will be entirely based on what you see in public.

Build 33, is obvious, its the 33build of the tree, and the 03/06, is the month of the build which is shipping - rather that me do it, you could justf--kinggoogleit or use http://www.opensolaris.org which explains the release process.

"3/06" for "March 2006", "Nevada" being the codename for what will most likely be designated "Solaris 11" eventually, and "build 33" meaning "the 33rd weekly or bi-weekly build of the Nevada codebase".

I still don't get the dates in the future though:


The October 2007 date is the expected "production" release date for this release, if I remember correctly.

Edited 2006-03-18 13:00

Ahh! Thanks.

For people who are looking for a Debian/Ubuntu experience with a Solaris Express (Nevada) kernel, the Nexenta OS is perfect. For more information and links to download, see:

http://genunix.org

You wrote:

> I wonder when Sun will simplify the installation of Solaris...

Best way to get/keep up-to-date on topics like that is via the opensolaris.org community and project pages and forums:

http://opensolaris.org/os/communities/
http://opensolaris.org/os/projects/

Also, be careful when describing problems. Report it with a complete description that includes how to trigger the problem. It can be misleading otherwise and doesn't help resolve the actual bug.[1]

Eric Boutilier
Blog: http://tinyurl.com/n59fk

[1]: Articulation accreditation: If you follow the main opensolaris forum, the last paragraph above may look familiar. See:

http://www.opensolaris.org/jive/message.jspa?messageID=27565#27565

It's important to use a good download tool, because the image files are so large. I've used Sun Download Manager ( http://www.sun.com/download/sdm/ ) with good results. At least, then, you can be sure the images themselves are good.

As far as Solaris vs. Linux installation: they are on par. Of course, Linux and Solaris each have quirks, like device naming, that people need to account for, but the installation processes are okay. I actually prefer Solaris' partitioning tools over Red Hat's, for example (keyword *prefer*--no flaming).

> Btw, is ZFS already available for the boot partition ? ...

It's in the works. Details (lots) can be found on this thread:

http://www.opensolaris.org/jive/thread.jspa?messageID=25434

On a related note, to subscribe to zfs-discuss, go here:

http://mail.opensolaris.org/mailman/listinfo/zfs-discuss

Or to receive zfs-discuss in your feed reader, use this (RSS/XML) URL:

http://www.opensolaris.org/jive/rss.jspa?feed=rss/rssmessages.jspa?...

Eric
http://opensolaris.org/

Correction: This is the correct feed URL (RSS/XML) for zfs-discuss:

http://www.opensolaris.org/jive/rss/rssthreads.jspa?forumID=80

> First, Solaris download sucks. Why not let the iso images on a ftp server or bittorrent ? I would like to use wget to download from my job ftp server but I can't get a simple url to download the images.


I completely agree with this.

> Second, it is ridiculous have to pay for a commercial use of a beta version. Even Microsoft wouldn't something like this. I can imagine that Solaris 11 will not free ( US$ 0) for commercial use also.

You are sadly misinformed - Solaris is free for all purposes.

> I prefer to use linux or any *BSD instead of Solaris. Both are solid and you can get more information freely on internet about them.

Did you know that you can also get the source code to Solaris?.

You may prefer Linux...good for you - now STFU and go write some code

Edited 2006-03-18 18:31

Why would you want to benchmark a production file system (ReiserFS) against something that is still in development (ZFS)? Have you talked to anyone at Sun about your idea or presented them any test data?

Robert Escue: "Why would you want to benchmark a production file system (ReiserFS) against something that is still in development (ZFS)?"

---

As stated in the subject title, I mentioned Reiser4, which is also not in the mainline kernel sources.

Reiser4 is not the same as ReiserFS (also known as version 3).

And rightly so even if I did mentioned ReiserFS, or XFS, or ext3 for that matter. Everywhere I see an article about a new Solaris release, I see users wanting to try ZFS. Why _shouldn't_ users have this information?

Regarding test data - explicit test data is not needed as it's typically generated by one of many artificial filessystem benchmark tools.

Regarding talking to Sun: I'll let them do their own work to market their products. However, their license agreement already dictates that any derivative reports created by myself fall under the category of "confidential information", which must be kept as such for 5 years after agreeing to the license, and therefore cannot be publically shared with fellow interested parties like yourself in a timely fashion.

Sun's license agreement muzzles the voice of people who want to compare their operating system again others'.

I, for one, do not appreciate these business tactics, and will take my business elsewhere.

Edited 2006-03-18 21:31

Sun's license agreement muzzles the voice of people who want to compare their operating system again others'.

I, for one, do not appreciate these business tactics, and will take my business elsewhere.

Standard business practice. Whether or not you agree with it is your personal option. However, I'm not surprised by it. The problem is if you don't have restrictions like these, people end up spreading incorrect information. Filesystems are especially difficult to benchmark correctly, most people do it wrong. That doesn't mean I agree with this policy either, I personally think that it shouldn't exist. However, I'm not about to "kick a gift horse in the mouth."

Considering that Solaris Express is essentially a beta product, I can see Sun's position in not wanting people to publish performance data on something that is not production ready. There are any number of vendors that do not desire to have what they would call "preliminary data" published, Sun Microsystems is not unique in that regard. Or is this another attempt to make Sun out as "evil" because of their stance for or against Linux or OSS software?

Right - I think binarycrusader mentioned this is standard practice, and from what I understand has been the cases for all releases; beta and production.

Even paying customers are under a similar muzzle.

Image a car company that does not allow the publishing of milage and performance information for their vehicles? (unless tested under their desired conditions?) - on the flip side I agree with binarycrusader; you certainly want a fair and competent test.

How hard would it be for Sun to perform these tests inhouse, against other operating systems running on identical hardware (Windows 2003, Linux, FreeBSD, etc..)? If they want to tout ZFS as a screamer - as mentioned here by Sun employees in 2004: http://blogs.sun.com/roller/page/val?entry=zfs_faqs_freenix_cfp_new
Why not show the progression over time?

Everyone who cares about filesystem performance understanding that things improve - we've seen the case with Reiser4 over the years.

Edited 2006-03-18 23:15

The performance would depend on the type of storage and systems used, at best any benchmark result would be a generalization of the performance potential of a given filesystem. If you have a specific performance need then you should be talking to Sun about testing ZFS and Reiser4 in one of their testing centers on hardware that you actually use.

> *yawn* another random release of the soon to be extinct solaris... how boring.

yawn, yet another linux fan boy soon to grow up and move out from his parent's basement.