Twice a year, when it comes time to upgrade to the newest Ubuntu release, I’m reminded of how nice it is to have my /home directories on separate partitions from the system directories–except when they’re not, because I chose the default Ubuntu partitioning scheme instead of configuring it manually.  Indeed, life would be much simpler if the Ubuntu installer gave /home a dedicated partition automatically.  Here’s why.

One of the nicest features of modern Linux is its strict separation of user data from system files, which makes it easy to upgrade the operating system to a newer version without losing personal files and settings.  By keeping /home on a dedicated partition, upgrading is trivially simple–users need only to run the Ubuntu installer, configure partitions manually so that / is wiped out while /home is not, and then boot into the new system without having to reconfigure any personal settings.  Even the desktop wallpaper is left untouched.

Granted, even without a separate /home partition, it’s still possible to back up the directory to removable media, reinstall the system, then put the files from /home back into place.  But that’s a tedious process that requires access to a sizeable external hard disk and generally means fighting with file permissions in order to get things working.  Definitely not for non-geeks, or for geeks who have better things to do.

Separate /home by default

In most cases–the exceptions being systems with very small hard disks–there’s no reason not to set up the /home directory on a dedicated partition when first installing Ubuntu.  It makes things much easier if a user decides to upgrade to a newer version of the operating system at a later date, and doesn’t entail any negative consequences besides slightly reducing the amount of disk space available for personal files–which matters less and less in a world where storage is becoming cheaper by the day.

Unfortunately, separate /home is an option that only geeks are likely to choose, since it requires manual partitioning, which will frighten most normal people away (with good reason, since there’s a real possibility of destroying data on a multi-boot system if the user doesn’t know what she’s doing).

As Alan Pope wrote on Ubuntu brainstorm almost two years ago, the lives of many users would be simpler if the ubiquity installer placed /home on a separate partition by default–or at least offered that choice as an autoconfigured option, rather than one available only through manual setup.

Recent releases have seen excellent improvements to ubiquity’s default behavior–the Ubuntu 9.10 installer even allows one-click setup of an encrypted /home directory–and easy configuration of a separate partition for personal files would be another great step on the usability front.  Let’s home the Ubuntu developers see the light soon.

A build of the nouveau video driver for nVidia cards was included in the Ubuntu repositories beginning with Jaunty.  I’ve been meaning to test it for a while, and finally found some hardware recently to give it a shot.

Much to the chagrin of free-software stalwarts, nVidia has always refused to provide a full-featured open-source Linux driver for its video chips.  Instead, it offers the nv driver, which supports only 2D acceleration–meaning no desktop effects–and whose source is ostensibly open, but is partially obfuscated and not very useful to third-party developers.

nVidia also offers a free (but not Free) closed-source driver that provides 3D support, and although the process of installing this driver in Ubuntu has been greatly simplified in recent releases, the fact that it’s proprietary presents a number of philosophical, legal and technical obstacles for Ubuntu developers and users.

To solve this problem, the cleverly named nouveau project was launched a few years ago to develop a full-featured, open-source video driver for nVidia chips.  As its feature status chart demonstrates, it’s still maturing and doesn’t yet offer any real 3D functionality, but 2D support is implemented.

Fedora 11 uses nouveau instead of nv as the default driver for nVidia chips.  In Ubuntu, nouveau is not installed by default, but is available from the repositories starting with Jaunty.

Testing nouveau

I was curious to see how well nouveau works on Ubuntu, so I installed it (using the xserver-xorg-video-nouveau package) on a Karmic system with a GeForce2 MX200 card (PCI ID 10de:0111).  The results were not exactly impressive, but not dismal either.

In terms of basic 2D functionality, nouveau gets the job done, albeit not yet as well as nv.  My monitor’s native resolution was set automatically, and I could move and resize windows and switch between virtual desktops smoothly.  There was a little choppiness during certain operations, but the desktop was usable.

In terms of performance, however, it’s clear that nouveau has a ways to go, at least on my hardware.  I used glxgears (yes, it’s not a good tool for benchmarking overall video performance, but it’s a useful basis for standardized comparison of FPS rates under different video drivers) to measure video frames per second under the nouveau, nv and nvidia (closed-source) drivers.  The tests were run with desktop effects turned off and the system idling.  The average FPS rates were as follows:

• nouveau: 355.3
• nv: 475.8
• nvidia (version 96, from Ubuntu repositories): 1993

Clearly, nouveau came out well short of the performance of even the nv driver, and was bested many times over by the proprietary nvidia driver.

These numbers shouldn’t be read as evidence of nouveau’s performance overall.  It may do a lot better on other video card models.  Unfortunately, I don’t have other nVidia hardware available to test on.

Moreover, the simple fact that nouveau works reliably is good news for Ubuntu users, even if the module has a lot of catching up to do to match the performance of nVidia’s closed-source driver.  nouveau may not be much yet, but it’s a solid start to what should one day allow more Linux users to take full advantage of their computers’ capabilities without having to rely on closed-source software.

If all goes according to plan, Ubuntu 10.10 will sport Gnome 3, which represents a radical overhaul of Ubuntu’s default graphical user interface in the form of Gnome Shell, when it debuts a year from now.  In order to get a taste of what this desktop of the future will look like, I’ve spent the last few days using the development version.  Here’s what I’ve found.

Gnome has been around now for a decade, and its approach to the desktop hasn’t changed remarkably in that time.  Gnome 1.x doesn’t differ in any fundamental way from the 2.2x versions available in the latest Ubuntu releases.  Traditional Gnome also behaves similarly to the interfaces of most proprietary operating systems.

That will all change with Gnome 3, however, which will bring a totally new interface to the table in the form of Gnome Shell.  As the Gnome developers described it:

…the shell idea is not just about changing the panel and the window manager. It’s about changing the way you start an activity and how you switch between two different activities.  Or more generally, how you manage your different activities on the desktop.

Changing the way we access documents (via a journal, like GNOME Zeitgeist [3]): having to deal with a filesystem in their daily work is not what makes users happy — on the contrary, they generally just want to access their documents and not to browse their hard disk. Providing new solutions to this problem (using timelines, tags,  bookmarks, etc.) is something that has been of interest in our community for a long time, but we never completely jumped in. We simply should.

Gnome Shell in action

Gnome Shell is a lot harder to describe in words than to demonstrate, so here’s a screenscast of the new interface in action, based on a build I completed a few days ago using the latest code:

The development version of Gnome Shell still has some significant bugs that need to be worked out and can be choppy at times, but the video nonetheless demonstrates the major changes to the interface.  Taskbars and docks have been replaced by an overlay that allows all windows to be managed from one location.  Virtual desktops can be added and removed as needed.  Recently accessed files are accessible with one click.  And Gnome Shell is its own window manager, rather than relying on metacity or compiz.

Too intuitive?

These new features are great and offer a lot of potential for improving productivity.  After using Gnome Shell for a few days before switching back to traditional Gnome, I’m already missing the former, and wish it were stable enough now to be used every day.

While it offers great new approaches to managing workflow, however, the new interface of Gnome Shell takes a lot of getting used to.  It may arguably be more “intuitive” (whatever that means) in the long run, but users accustomed to drop-down menus and taskbars will likely find themselves quite uncomfortable the first time they use Gnome Shell.

The beauty of Gnome 2.x is its simplicity.  Anyone who’s used any kind of GUI-equipped computer before–or who hasn’t used a computer at all–can sit down in front of traditional Gnome and figure things out pretty quickly.  There’s not much ambiguity in the well labeled Gnome menu, and Gnome manages windows and tasks like every other mainstream desktop environment created in the last fifteen years.

But since “intuition” for most people has more to do with what they’re used to than anything else, I worry that Gnome Shell will turn normal users off to Ubuntu and Linux by forcing them to learn a new interface.  Maybe that won’t be the case–maybe non-geeks are less set in their ways than I tend to believe–but this obstacle should nonetheless be a point of consideration for Gnome and Ubuntu developers.

For the time being, Gnome Shell remains in heavy development, and may see significant changes before it reaches maturity.  As it exists now, however, it looks very promising, provided it can reconcile a radically progressive interface with normal users’ desire for familiarity.

Ubuntu is poised to take another small but significant step into corporate enterprises, thanks to Bomgar, the appliance-based remote support software specialist. When Bomgar 10.4 debuts November 16, it will support a growing list of Linux distributions, including Ubuntu, Red Hat and SUSE Linux Enterprise Desktop. Here are the implications for Ubuntu administrators.

Bomgar is a virtualized help desktop platform that allows companies to remotely manage systems over the Internet. Since 2007, Bomgar’s software typically required an IT manager to use a Windows, Mac OS X or SUSE Linux desktop as their remote administration platform. Going forward, support technicians will also be able to run Ubuntu when managing remote systems, Bomgar says.

Bomgar’s pledge to increasingly support Ubuntu is noteworthy because Bomgar itself has a growing following among corporate IT managers. In fact, Bomgar is among the fastest-growing privately held companies in the United States, according to the annual Inc. 500 survey and report.

Meanwhile, Canonical has been striving to improve Ubuntu’s enterprise positioning. The new Ubuntu 9.10 Server Edition includes a private enterprise cloud option — known as EUC (Enterprise Ubuntu Cloud). Plus, Canonical continues to polish Landscape, a remote management and administration tool for Ubuntu desktops and servers.

Follow WorksWithU via Identi.ca, Twitter and RSS (available now) and our newsletter (coming soon).

ZaReason is taking a small but significant step to help strengthen the Ubuntu brand. To help celebrate the Ubuntu 9.10 launch, the PC maker is offering “a stick of Ubuntu aluminum case badges with each computer ordered” October 29 through November 5. Here’s why you should care.

Let’s be honest: Many consumers and businesses have yet to discover Ubuntu. But those who do frequently turn to ZaReason or System76 — two Ubuntu-centric PC makers — for help and guidance. I mentioned last week that both ZaReason and System76 planned to offer Ubuntu 9.1o preloads upon the operating system’s debut.

ZaReason’s efforts are particularly interesting because the PC maker always seems to be thinking about promoting and evangelizing the Ubuntu brand — placing Ubuntu logos on keyboards and Ubuntu case badges on netbooks, notebooks, laptops and servers. And now, ZaReason is offering those free case badges Oct. 29 through Nov. 5 with system purchases.

Such branding isn’t a big deal. Or is it?

I think you already know my answer.

Follow WorksWithU via Identi.ca, Twitter and RSS (available now) and our newsletter (coming soon).

When Canonical launched Ubuntu 9.10 earlier today, Centrify Corp. — maker of Microsoft Active Directory-related tools — raised its hand and vowed to support Ubuntu Server Edition 9.10. It was a small but significant announcement that could (potentially) help Ubuntu gain a stronger foothold within corporate Windows networks. Here’s why.

Centrify offers “Active Directory-based auditing, access control and identity management for Unix, Linux, Mac OS, Java and web platforms.”

I’m not positive if Centrify overlaps with such solutions as Likewise Software. But I was impressed to see Centrify front-and-center with Ubuntu-oriented news today.

Frankly, I wish more software development firms raised their hands and pledged to support Ubuntu Server Edition 9.10 — but I suspect most Canonical software partners are waiting for Ubuntu Server Edition 10.04 (Lucid Lynx), a Long Term Support (LTS) release scheduled for April 2010.

Centrify’s Offering

Centrify claims its software suite for Ubuntu 9.10 Server Edition enables “an open source cloud computing environment that will be easily integrated with Microsoft Active Directory.”

David McNeely, director of product management at Centrify, offered this prepared statement:

“Centrify helps organizations deploying Ubuntu by enabling granular, role-based access controls and privilege management that goes well beyond access controls to ensure that their administrators have only the rights they need without giving them extended privileges that, if misused, could compromise systems and by extension the business.”

Centrify claims to turn a non-Microsoft system or hypervisor into an Active Directory client, enabling administrators to secure that system using the same authentication and Group Policy services currently deployed for their Windows systems.

Does Centrify work as advertised? I haven’t tested the software, which arrives for Ubuntu 9.10 Server Edition (starting at $350 per server) in late December.

Interestingly, Centrify promotes its software through a growing channel partner program. As you likely know WorksWithU has two channel-focused sister sites (MSPmentor.net and TheVARguy.com). So we respect channel-friendly companies.

It’s good to see an ISV raising its hand and supporting Ubuntu 9.10 Server Edition. Hopefully more ISVs will raise their hands during Ubuntu 10.04’s release cycle.

Follow WorksWithU via Identi.ca, Twitter and RSS (available now) and our newsletter (coming soon).

Across the board, Nine Lives Media Inc.’s web traffic is up 140 percent in October 2009 vs. October 2008. We thank you for your growing readership. We’re taking two steps to improve site responsiveness: A hardware upgrade from our service provider plus a few of our own software architecture tweaks. I will post a follow-up comment to this blog entry as soon as the adjustments are completed. In the meantime we appreciate your feedback. We’re listening and responding.

As Ubuntu 9.10 debuts today, I’d like to spend less time talking and more time listening — to you. What is your initial reaction to Karmic Koala? Desktop and mobile feedback is always welcome. But I’m particularly interested in server, cloud and Landscape feedback from IT administrators. Please feel free to post a comment or email me directly (Joe [at] NineLivesMediaInc.com). WorksWithU intends to develop a series of user case studies exploring Ubuntu 9.10’s business performance across mobile, desktop, server and cloud systems. I look forward to your thoughts.

As Ubuntu 9.10 nears its official Oct. 29 debut, the hype around this new Ubuntu is reaching a fever pitch. Some folks wonder if it can be the de facto alternative to Windows. I’ve obviously bet a portion of my IT media career on Ubuntu’s continued advancement. But let’s keep Ubuntu 9.10’s debut in perspective. Here’s a reality check — including a longer-term look at Canonical’s server and cloud strategy, which hinges far more heavily on Ubuntu 10.04 (Lucid Lynx).

There’s a lot to like about Ubuntu 9.10, and its emphasis on design and ease of use. Pundits say confidence is running high within Canonical. And Internetnews openly wonders if Ubuntu 9.10 can become the “default alternative to Windows.”

I prefer not to make such lofty predictions. Especially since Ubuntu 9.10 is like an appetizer before the main course: Ubuntu 10.04, a Long Term Service (LTS) release, could define Canonical’s success — or failure — on the server and in the cloud.

Beyond Desktop Linux

Let’s remember: Canonical CEO Mark Shuttleworth has an end-to-end strategy for Ubuntu. It involves mobile systems, desktops, servers, private and public clouds, and a growing list of sister services (Landscape, UbuntuOne, etc.).

I get a bit frustrated when readers criticize Canonical for thinking big and looking beyond the desktop. Let me state it another way:

• What if Microsoft had stopped at Windows 3.1?
• What if Microsoft never built out Windows NT Server, Microsoft BackOffice, SQL Server, Exchange Server and now Windows Azure and Microsoft Business Productivity Online Suite (BPOS)?

My point: By diversifying its product line and building end-to-end solutions, Microsoft became infinitely more powerful. And let’s be honest: Microsoft’s server software  freed many corporations from Unix systems running on very expensive RISC processors. Canonical hopes to deliver infinitely more freedom, if the company can execute on Shuttleworth’s end-to-end strategy.

Getting the Message

Already, I’m hearing from WorksWithU readers who intend to give Ubuntu 9.10 a close look on a range of systems. Consider the situation Rezitech, a managed service provider and IT consulting firm in La Habra, Calif. Notes Rezitech Senior Systems Engineer Troy Ready:

“I obviously like the neat things slowly changing on the desktop, but I think the UEC [Ubuntu Enterprise Cloud] focus on the server side is what’s really fascinating now. Our (Rezitech) UEC plans are still up in the air (pardon the pun), but I’m really excited for how the aggressive Ubuntu development will (hopefully) shape up for a stable LTS next year — it might really change the layout of our infrastructure.”

More Ways to Run Ubuntu

A year ago, Ubuntu and cloud computing didn’t belong in the same sentence. But now, a respectable 3% of WorksWithU readers say they’re going to run Ubuntu 9.10 in a cloud configuration, according to our recent reader poll (nearly 700 readers have so far participated).

In some ways, the Ubuntu 9.10 hype is ironic. Even if the release is rock-solid and well-received, Canonical’s server and partner teams already have their eyes on Ubuntu 10.04 (Lucid Lynx), a Long Term Support (LTS) release expected in April 2010.

At Atlanta Linux Fest in September 2009, Canonical’s John Pugh told me he was focused entirely on Ubuntu 10.04 — since LTS releases represent a prime opportunity to attract more ISVs (independent software vendors) onto the Ubuntu Server Edition platform.

Here again, Canonical faces a situation that’s similar to Microsoft’s old challenges. When Windows 95 arrive in August 1995, everyone hyped Microsoft on the desktop. But the real “Windows Everywhere” synergies surfaced when Windows NT Server 4.0 arrived in August 2006 1996.

Apply the same example to Canonical and you’ll see some striking similarities. The open source media is hyping Ubuntu 9.10 on the desktop. But keep a close eye on Canonical’s Ubuntu 10.04 server and cloud efforts. When Ubuntu 10.04 arrives, we’ll get a far better feel for how Shuttleworth’s end-to-end vision is playing out.

Please note: I realize I skipped a key topic of discussion — whether Canonical can actually be profitable offering all of these different Ubuntu releases and add-on services. I posed the question briefly to Shuttleworth on Oct. 26 during a media call. I recapped his thoughts briefly in this blog post.

As we begin to hear about more corporate Ubuntu deployments, I’ll be sure to ask if the projects involved revenue flowing in Canonical’s direction.

Follow WorksWithU via Identi.ca, Twitter and RSS (available now) and our newsletter (coming soon).

During a phone briefing today, Canonical CEO Mark Shuttleworth described the Ubuntu 9.10 desktop, server and cloud strategy to members of the IT media. WorksWithU tuned in and posed some key questions to Shuttleworth. Here are 10 highlights from the call.

As you’ll recall, Ubuntu 9.10 is scheduled to debut Oct. 29. Shuttleworth’s thoughts from today’s call included:

1. The User Experience: “We wanted to bring design and user experience to [the Linux] desktop.” Shuttleworth believes Ubuntu 9.10 achieves those goals.

2. Competition with Microsoft, Windows 7: Shuttleworth concedes that Windows 7 is impressive but “it’s still proprietary and expensive.” Also, he says, OEMs have “no desire” to go back to a single-vendor operating system market.

3. On the Netbook Market: Shuttleworth concedes that Microsoft “clawed its way” back dramatically in the U.S. netbook market with Windows XP. But he hopes once the Windows 7 dust settles, vendors and users alike will realize Ubuntu netbooks are a natural choice.

4. On Oracle’s Buyout of Sun (and MySQL): Shuttleworth sees no reason for regulators to block Oracle’s buyout of Sun Microsystems nor, by association, the MySQL open source database. Shuttleworth doesn’t see an opportunity for Oracle to abuse its database power. Plus, he notes that open source code can quickly fragment if there’s community concern about the governance and leadership of a project.

5. On Canonical’s March Toward Profits: Shuttleworth says he has “no concerns” at this stage about Canonical’s ability to achieve profitability. If necessary, Canonical could quickly focus on specific business areas that are self-sustainable in the near-term, he asserted. But at this point in Canonical’s five-year business journey, Shuttleworth remains convinced that the best strategy is to make Ubuntu an end-to-end solution.

Translation: Shuttleworth isn’t ready to say how soon Canonical could achieve profitability. But he’s basically saying the company is willing to lose money in certain areas as new businesses ramp up and strengthen the overall Ubuntu ecosystem.

It’s a daring bet considering how many different areas of focus Canonical has taken on. But then again, I doubt we would have launched WorksWithU if Canonical only wanted Ubuntu to be a desktop operating system.

6. Simple Commercial Software Installs: The Ubuntu Software Center will gradually gain more and more commercial software options — though the exact details of those efforts are still being worked out.

7. Closed Source ISVs — Still A Challenge: Shuttleworth conceded that Canonical’s biggest weakness in the enterprise remains a lack of traditional closed source ISVs. The elephant in the room was Oracle…

8. But There Is ISV Progress: Still, Shuttleworth was quick to note continued ISV progress with such companies as IBM’s information management group, Alfresco (open source content management) and JasperSoft (open source business intelligence).

9. Amazon As A Cloud Standard: Shuttleworth said Amazon’s cloud APIs are emerging as a de facto cloud standard, though he expects open, vendor-neutral cloud standards to gradually materialize.

10. Sever Hardware Relationships: Shuttleworth conceded that he had no new server relationship or certification deals to disclose.  From where I sit, it seems as if Amazon Web Services (particularly, the Elastic Compute Cloud/EC2) and Eucalyptus cloud technology are emerging as Canonical’s most reliable doorway into the server market.

Follow WorksWithU via Identi.ca, Twitter and RSS (available now) and our newsletter (coming soon).