Why Dell Should Upgrade Past Ubuntu 8.04 LTS

[themiddaysun]

The tests were run using an SSD, some of the write performances were low for 9.04. From the article I read on anandtech degridation of writes will happen on SSD's after the drive has been used for a while. I guess my point to this is....Was a different/new SSD used for each install of Ubuntu? If not, maybe the block and page formula that is explain in that article are in play here causing bad performance. Just curious.

[deanjo]

One interesting aspect of Dell's decision to sell PC preinstalled with Ubuntu, is how Dell competes with Canonical. Both Dell and Canonical are in the business of selling support services.

They aren't competing at all. Dell offers hardware support on the systems with ubuntu pre-installed. They QVL and support their hardware. When it comes to OS support, that still goes to Canonical.

Hardware Support

Dell supports its hardware sold with Linux. The standard hardware warranty applies.
If you purchased a Dell system with Ubuntu factory-installed, we have a special hardware support queue per country:

• U.S Ubuntu Technical support - 866 622 1947
• U.K Ubuntu Technical support - 0870 353 33 53
• France Ubuntu Technical support - 0825 387 167
• Germany Ubuntu Technical support - 069 9792 15726

Software Support

• If you purchased Canonical support when purchasing your Dell system, see Canonical for details.
• To purchase Professional support for running Ubuntu on these systems, see Ubuntu.
• Online documentation and community-based support is available through the Ubuntu web site, and the Dell Community Forum: Linux message board.

http://linux.dell.com/wiki/index.php/Products/Consumer

[DanL]

Michael, while your benchmarking is great, I think you're barking up the wrong tree on this one.
Translation: Dell wants to minimize the amount of times they have to deal with customers and send them new hardware. They want to make money off the Linux market while dedicating minimal support resources to it. They don't want to pay for Linux support gurus, just shadetree nerds they can hire off of the street. They believe that having older, more "stable" software accomplishes this.

[pipe13]

Dell wants to minimize the amount of times they have to deal with customers and send them new hardware. They want to make money off the Linux market while dedicating minimal support resources to it. They don't want to pay for Linux support gurus, just shadetree nerds they can hire off of the street. They believe that having older, more "stable" software accomplishes this.

Good point Dan. Leaving whatever Dell can hire off the street aside, anytime Customer Support picks up the phone Dell runs a (hopefully small) risk that a software problem will be misdiagnosed as hardware, and an RMA issued. Nobody wants that. LTS clearly is the winner here.

[tyroeternal]

Here is one MAJOR show stopper for Dell, and a big lesson learned on my part about assuming they would only use hardware with proper support:

The Dell Mini 12 (offered with Ubuntu) and the Mini 10 (soon to be offered with Ubuntu) both use Intel GMA 500. As of right now, and apparently for a good while to come, their drivers for linux are no good with anything beyond the 2.6.24 kernel.

Until Poulsbo drivers are developed, the Mini 12's and 10's cannot upgrade to anything newer than 8.04. Following the LTS releases may be their only option when it comes to proper support for some of the hardware they provide.

[RoboJ1M]

In my experience, anything that is not an Ubuntu LTS release is not stable enough for use by anyone.

I've had my gf's 8.10 PC fail and require re-installation and Bluetooth (BlueZ 4) is just flat broken. 8.04 uses BlueZ 3. Less advanced, but at least I can OBEX files around. Bluetooth in 9.04 is still broken.

J.

[NickF]

Although this was a great review, I really don't understand why the i386 was tested. The Dell-mini comes with a custom version of Ubuntu, built on the lpia architecture. That apparently was removed for the review and a regular i386 installation was preferred.

I was hoping to see how much was gained from the lpia optimization vs the regular i386, including battery time, which isn't tested.

That's a pity...

[curaga]

@NickF:

There's already an article about the power consumption:
http://www.phoronix.com/scan.php?pag...er_usage&num=1

[NickF]

@NickF:

There's already an article about the power consumption:
http://www.phoronix.com/scan.php?pag...er_usage&num=1

Thanks for the link. However I was interested to see how those test compare with the performance of jaunty i386...