Practical Technology

for practical people.

January 30, 2008
by sjvn01
0 comments

Why companies don’t support Debian

At a recent Australian Linux conference, Sam Varghese reported that two Debian developers pointed out that the Debian Project needs more corporate support for “men, money and machines” to advance the operating system.

They’re right. It does. They also pointed out that many companies, such as HP, IBM, Silicon Graphics and Google, either use Debian Linux internally, or actually incorporate it into products. For example, HP uses Debian “Etch” 4.0 in its new t5735 thin-client device. Right again.

Debian, either directly or through related Linux distributions such as Xandros, is used both by Linux enthusiasts and Fortune 500 companies.

Of course, you couldn’t prove that by the vast majority of Debian developers who never see a thin dime from their Debian work. Or, I should add, get access to new hardware, travel expenses to Debian developer conferences and so on.

The reason for this is twofold. First, Debian, as a developer community, has never wanted any kind of “business” organization or corporate partnerships or sponsorships. It is purely a volunteer operation and woe unto any would-be developer who tries to change Debian’s ways.

For example, in 2006, several Debian developers, led by Anthony Towns, the official Debian project leader at the time, decided to arrange for the two release managers to be paid for some of their efforts. In response to this, other developers refused to work on the next release of Debian. As Andreas Barth, a Debian release manager, explained at the time, the delay has resulted because, “Some people who used to do good work reduced their involvement drastically. There was nothing I could do about, and that happened way before I started full-time on release.”

At other times, there have been efforts to create a commercial alliance of Debian groups and Linux companies that made Debian-based distributions. They have always been met with strong resistance from the Debian community. Even the most important of these efforts, the Debian Common Core Alliance, which was led by Debian founder Ian Murdock, met with Debian opposition. The group, beset by the very community it sprang from, eventually changed leaders but still failed.

Debian doesn’t just turn on its own when they try to organize Debian in a more businesslike manner. The community also turned on the Mozilla Foundation because it restricted the use of its Firefox Web browser logo. So it is that today “pure” Debian distributions include IceWeasel, which is essentially a carbon copy of Firefox with a different, “freer” logo.

So, while I think Debian’s development needs radical change and commercial support, the community seems dead set against it. A large number of the developers would rather be free than beholden to any corporate entity. As CJ Fearnley, CEO of Linux service provider LinuxForce, and a long-time Debian developer, put it to me in late 2006, Debian is “Very, very messy at times. This is democracy in action!”

This comment points to another side of Debian’s problem with corporate support. Let’s say someone, say a company like Google, does give Debian a completely unrestricted gift of money or equipment. Who does Google give it to? The current Debian project leader, Sam Hocevar? That’s a one-year position.

The group itself? A fine idea, except it’s made up of more than 1,000 developers scattered across the world. It’s also, as Fearnley alluded to, not a very organized group. For example, the Debian Weekly News, a newsletter for the Debian community, doesn’t appear to have been published since July 3, 2007.

If Google were to decide to give financial support or PCs to what it determined to be the most “deserving” of Debian developers, it would only be following the path already taken by Anthony Towns, which proved to be a failure.

In short, there’s no there there in Debian. There’s no one to write a check to, and even if you did write a check to an individual, the other developers seem likely to turn against him or her.

Now, despite all this, Debian does keep going. It just released a major security update to Debian 4 late in 2007. But, it’s going to do things its own, cranky way.

If you want Debian to behave in a more businesslike way, if you want companies to support Debian, what you really need to do is to break away from Debian and make your own distribution on a more traditionally organized basis. This, of course, is exactly what Mark Shuttleworth did when he founded Ubuntu in 2004.

Ubuntu, as its deals with Dell and Sun show, has had little trouble finding corporate partnerships. For Debian, as it’s now constituted, that will never be an option.

January 30, 2008
by sjvn01
0 comments

Come On! VMware`s Quarter Wasn`t that Bad

I’ve long thought VMware, both the company and its virtualization software, were in trouble, but come on, people, it’s not like the last quarter was awful.

When I last looked at the stock market – 2 p.m. EST, Jan. 29 – VMware’s stock was being beaten to death by crazed sellers. How badly? Try down $27.30 per share, or 32.9 percent since the market opened. Ouch!

And why were they throwing stock overboard like passengers tossing other passengers off an overloaded lifeboat? Because, dare I say it, the company only showed a 150 percent increase in fourth-quarter profit and an 80 percent jump in sales to $412 million, when the analysts had predicted that it would make $417 million.

Say what? My calculator tells me that VMware missed its earnings prediction by about 1.2 percent and that by all objective measures it had a great quarter. Look at those numbers again: 150 percent increase in profit. 80 percent jump in sales. I don’t know about you, but if I were a software vendor, a MacDonald’s franchiser or a plumbing supplies store owner, I’d settle for an 80 percent jump in sales.

Now, I think VMware, the company, was always a bad buy. What did you get for buying VMware? I’ll tell you, you got a small fraction of 13 percent of the company. The lion’s share, 87 percent, was kept by EMC. I’m sure EMC will put the best interests of VMware stock holders over those of EMC. Oh, and would you like to buy a slightly used bridge between Brooklyn and Manhattan? Only one owner.

Besides, VMware is a perfect example of a proprietary software company–think SCO, think Sun before it got the open-source religion–about to be overrun by open-source software. In 2006, Sunbelt, a Microsoft partner, and Yankee Group reported that VMware had a total virtualization market share of 55 percent.

That was then. This is now.

Today, open-source software virtualization programs such as Xen, OpenVZ, KVM, VirtualBox and UML (User Mode Linux) are quickly gaining traction. Others such as Parallels, which is based on open source, enable users to do things such as run multiple instances of Mac OS X on a single system that’s beyond VMware’s current ability.

While VMware offers free, low-end virtualization programs, VMware Player, for example, its higher-end software, is anything but free. Thus, the open-source programs, in the hands of administrators who know what they’re doing with virtualization, are a better (read cheaper) deal for companies.

In addition, because VMware isn’t open source, its development speed lags behind the manic pace of the Darwinian, only the best code survives, free software world.

The final nail in the coffin is that operating systems companies are integrating virtualization into their operating systems. The Linux kernel now has KVM baked in. Red Hat and Novell both offer Xen virtualization with their Linux distributions.

There’s also a little company named Microsoft – heard of them? – they’ve decided that you can virtualize any version of Vista after all. Microsoft also just bought Calista Technologies, a desktop virtualization company. Their good buddy Citrix also bought XenSource, the company behind Xen, over the summer.

Now, maybe in the long run, VMware could deal with the open-source threat. After all, look at how well Unix on x86 has done…well, maybe not. It’s also possible that it could handle Microsoft building virtualization into the operating system. I mean, aren’t we all still using the Netscape browser? No?

OK, so I’m convinced that, in the long run, VMware’s in trouble. But, in the short run, like today’s market panic about VMware? Come on, people. Save it for early 2009, when I think you’d need to be an idiot to have VMware in your stock portfolio. For today, though, get a grip! The current news is good.

P.S. I don’t own a share of any technology stock and never will so long as I’m in the technology journalism business.

A version of this story first appeared in eWEEK.

January 29, 2008
by sjvn01
0 comments

Sun`s Best Buy Ever: MySQL

Some people can’t believe Sun actually spent a billion dollars for MySQL; I can’t believe they got such a great deal.

If you believe my long-time colleague John C. Dvorak, Sun purchasing open-source database power MySQL was “perhaps the worst single event I have ever witnessed in the history of tech mergers and acquisitions.” Nonsense! This is Sun’s best deal ever.

Besides, the worst technology acquisition of all-time is still Compaq buying DEC. Let us have a moment of silence now for Digital. Sun buying MySQL, on the other hand, isn’t just a good deal, it’s a great deal.

Some observers are horrified that Sun paid a billion — $800 million in cash and picking up $200 million in pre-existing MySQL stock options–for MySQL, a company with a 2007 revenue of a mere $53 million. I think this says more about MySQL’s failure to effectively transform their open-source software and services into cash than it does about the almost limitless potential of the MySQL DBMS (Database Management System).

Look at where MySQL is already used. Almost every blog and many social networks and content management systems run with MySQL at their heart. When people talk about running Linux as an enterprise server platform, nine times out of 10 what they’re actually describing is a LAMP (Linux, Apache, MySQL, Perl/PHP/Python) software stack.

Yes, only a few companies currently pay MySQL for support, but do you know who those customers are? MySQL’s customer list includes Google, Nokia and Yahoo. If I were a software vendor, I’d like to have those companies as my customers, wouldn’t you?

If you take a long hard look at how MySQL is used in business, here’s what you’ll see. Besides LAMP stacks, you’ll also see it being used in SOA (service-oriented architecture) and Web 2.0 applications. Enterprises, top analyst Dan Kusnetzky of The Kusnetzky Group told me long ago, buy software for one of two reasons: the operating system and associated software stack supports the applications they need or the database supports the applications they need. MySQL supports the applications that 21st century companies need.

Sun is a winner though on both counts. With MySQL, Sun can finally offer companies a complete operating system, DBMS, and software stack. Besides the already existing LAMP stack, Sun can also do well by finally revive Solaris’ flagging sales. How? By creating what I’m going to call the SAMJ (Solaris, Apache, MySQL, Java) stack.

By themselves neither Solaris nor Java has been doing that well in the enterprise market lately. Even the belated open-sourcing of both proved to be a case of too little, too late. I believe though that OpenSolaris and Java have the potential to matter much more both for the open-source community and CIOs. When you team them up with MySQL, Sun has the potential to put together a very compelling alternative to LAMP or Microsoft’s hodgepodge of servers, Live services and .NET developer stacks.

Now, Sun could blow it. Over the years, I’ve given Sun a lot of grief for going back and forth about whether they were really supporting Linux and open source and also for what was really the company’s biggest acquisition flop, its complete waste of a billion on the Linux appliance company Cobalt Networks.

This time though I think Sun has it right. More to the point, as I look at Sun over the years, I see a company that’s been at war with itself.

There was the fight between those who saw Sun as a hardware company and those that saw it as a software one. There was also the conflict between those who believed in Linux and/or open source and those who believed in proprietary software as the one true way. Over the last two years Sun appears to have settled these internal conflicts.

The winners were the software and open-source crew. Now, with one voice and the purchase of MySQL, they have the chance to show that they were right all along. I don’t see these people, some of Sun’s best and brightest, blowing their chance.

Sun’s worst move? No way. This is Sun’s best move, and certainly its best chance, to become once more a major IT power. And, at the same time, it will give corporate customers the kind of pricing and support they want and need and aren’t likely to find from Microsoft or Oracle.

A version of this story first appeared in eWEEK. >

January 28, 2008
by sjvn01
0 comments

Will a Cheap Vudu be a Successful Vudu?

Apple TV is hardily the only player in the IPTV (Internet Protocol TV) game. One other significant player, Vudu has a similar product, the eponymous Vudu Box.

The Box, which had been going for about $395 and includes a 250GB hard drive, has been dropped to $295. This is, doubtlessly, a direct response to Apple repricing of the Apple TV. Today, the 40GB Apple TV goes for $229 and the 160GB Apple TV will run you $329.

Comparing the two is an interesting exercise. On the one hand, the Vudu supports HD (high definition) all the way up to the top of the line 1080p HD, while Apple TV can only support 720p HD. Until recently, you had to use a Mac or a PC with iTunes to place movies or TV episodes on an Apple TV. Any day now a new firmware upgrade will enable you to download rented or bought video directly to the Apple TV. Vudu has always come with direct to the TV functionality.

Continue Reading →

January 28, 2008
by sjvn01
1 Comment

Has Microsoft Disavowed Vista?

Technically, Vista is pure misery. It eats system resources like an elephant does peanuts, Windows break and its so-called improved security is a joke. I know it. You know it. Even Microsoft’s most devoted yes-men know it–although they won’t admit it–and perhaps Microsoft knows it as well.

What else can explain why Microsoft is now leaking news about Windows 7, the next version of Windows? Oh, officially Vista SP 1 is still the big upcoming news, although I think most businesses are actually more interested in XP SP 3. The simple truth is that no matter how Microsoft and its partners like CDW spin it, Vista is not being picked up by corporate users. Even Bill Gates’ vaunted 100 million Vista users number should be taken with a large—very large—grain of salt.

Most of the information is dripping out of the blog, Shipping Seven. But, it’s more than just Shipping Seven, which may, or may not, be real. Microsoft is hard at work, harder than one would expect, with Vista just over a year old, in getting its next desktop operating system ready for action.

As Directions on Microsoft analyst Michael Cherry recently told eWEEK’s Peter Galli, “I don’t think Vista is as bad as Microsoft has convinced people it is.” What should Microsoft do then? Cherry recommended that Microsoft “discuss the next version of the operating system, currently referred to as Windows 7, and what it will do.”

Could Vista have missed its shot? Yes, yes, I know, how can I say this when there are tens of millions of copies of it out there? Easily. It’s one thing to drop copies of Vista Home Basic and Premium on Best Buy customers who don’t know any better. It’s another thing entirely to get CIOs and IT managers to spend—or should I say waste?—billions on Vista.

For now, whether Microsoft likes it or not, XP, and not Vista, is the Windows those businesses will continue to use. And the companies that want to move on to a truly better operating system? They’ll be moving to Linux or Mac OS.

A version “Has Microsoft Disavowed Vista?” first appeared in eWEEK. >