Linux, LSB, UnitedLinux - MozillaQuest Magazine - UnitedLinux, a Divisive Weapon for Caldera's Darl McBride -- Part 1

MozillaQuest the on-line computer magazine

September 4, 2002

Digital Photography

Graphics

Multimedia

Web Page Design

Web Tools


UnitedLinux, a Divisive Weapon for Caldera's Darl McBride -- Part I
By Mike Angelo -- 4 September 2002 (C) -- Page 2

Article Index

Part I.

Introduction
Caldera's Divisive McBride Attacks Red Hat and Mandrake
LSB and UnitedLinux
Summary for Part I

Part II.

Caldera's Divisive McBride Attacks Red Hat and Mandrake

In mid-August at the LinuxWorld Expo in San Francisco, Caldera's new CEO Darl McBride made some very divisive statements. He fired full salvos at Red Hat Linux and ignored Mandrake Linux. Moreover he did this under the color of the UnitedLinux banner. Simply put, McBride seeks to put the fledgling UnitedLinux in a fight with Red Hat (by name) and Mandrake and other Linux distribution providers (not named) to squabble over grabbing shares in the same, existing Linux market pie.

UnitedLinux . . . announced the imminent availability of the product for beta testing, promising to breathe new life into the challengers to Red Hat Inc. UnitedLinux Set for Beta; Caldera Eyes New Era, Scot Petersen, eWEEK, August 13, 2002. (Link in Resources section at end of this article.)

Going forward, McBride said, the UnitedLinux group will try to offer Linux users and enterprise customers an "alternative" to Red Hat Linux. (ibid.)

"If we combined our market share, we have more than Red Hat," said new Caldera president and CEO Darl McBride . . .. "If we combine our developer count, we have more than Red Hat." (ibid.)

However, it appears that at least one UnitedLinux member does not agree with McBride's statements -- or the in-your-face, anti-Red-Hat, anti-Linux-community marketing policies for UnitedLinux that McBride articulated in San Francisco. More about that further down and in Part II.

Gaël Duval is the founder of Mandrake Linux and co-founder of MandrakeSoft SA and MandrakeSoft, Inc. Mandrake is not a member of the UnitedLinux consortium, however, Mandrake is a participant in the LSB project and was the first Linux distribution to achieve LSB certification. Moreover, Mandrake and Gaël Duval have clearly stated that Mandrake will not join the UnitedLinux consortium. In e-mail discussions we asked Gaël Duval about McBride's market share claims.

Gaël Duval: It's easy to say things. Now we need proof.

Note: Actually, McBride and Stowell have been ducking our questions about Caldera, LSB, and UnitedLinux since 2 August 2002. Then we sent an e-mail interview to Darl McBride via Blake Stowell. In that interview we asked McBride if and when Caldera OpenLinux would be LSB compliant, if LSB is important, how Caldera would fit into UnitedLinux, if UnitedLinux and LSB can get along with each other, and so forth. We never received answers to those questions.

However, in light of McBride's San Francisco attack on Red Hat, which has an LSB certified Linux distribution, and the more recent superficial change in Caldera to a UNIX company, it is no wonder McBride and Stowell wanted to duck our 2 August questions. (Superficial because it seems that Caldera in fact has been a UNIX company rather than a Linux company for some time now.)

Then, after we decided to do this story we again sent a set of questions to Blake Stowell on 23 August 2002. The 23 August questions were generally similar to the questions we asked Mandrake's Gaël Duval, Red Hat's Mark de Visser, and SuSE's Holger Dryoff -- about McBride's statements, LSB, and UnitedLinux. We also specifically asked what data McBride and Caldera had to back up McBride's statements: "If we combined our market share, we have more than Red Hat," and "If we combine our developer count, we have more than Red Hat."

Stowell responded by requesting a week to answer our questions stating that he was busy preparing for the then upcoming Caldera/SCO forum scheduled for the then following week in Las Vegas. However, a typical PR person trick is to try to stall-off what looks like something that will be bad press for the PR person's company. Is that what Stowell was up to?

Also, the questions we asked amounted to something that should have taken at most a half-hour if the answering were straightforward rather than evasive hype. It takes much longer and much more work to lie than to tell the truth.

Moreover, if data-based claims are actually made on real, existing data, it does not take long to answer questions about the data. However, if the data does not actually exist -- well then it can take longer to get the data together. Moreover, if DeVisser, Dryoff, and Duval could take the time to answer our questions there was no reason why Stowell and/or McBride could not take the time to answer our questions -- unless perhaps there was something they were hoping to cover up.

We thought we were just about ready to go with this story at that time (23 August). So all things considered, we told Stowell: We already have comments from other people about these issues ready to go. This story is scheduled to go up today. You are more than welcome to follow up next week. However, the story goes up today with or without your (Caldera's) response.

We had all the intentions of publishing this story on 23 August. However, AOL-Netscape decided to release Netscape 4.8 that day. Then over the next week (last week) Mozilla 1.1 and 1.0.1-RC2 and Netscape 7.0 were released -- four major, breaking news, release-stories to cover. Add to that many additional follow-up discussions with the Mandrake, Red Hat, and SuSE people and it turns out that this story is being published about ten days or so later than we expected when we told Stowell that the story was going up on 23 August.

So, Blake Stowell had more than the week he said he needed to answer our questions and get his answers in to us. Thus, McBride, Stowell, and Caldera/SCO were afforded more than ample time to respond to the issues raised in this story but failed to do so. If their strategy was that by not answering the questions the issues would go away, they were sorely mistaken.

Additional Note: A pre-publication draft of this article was e-mailed to Blake Stowell yesterday to afford Stowell, McBride, and Caldera/SCO the opportunity to respond to this story, We received no reply to that request for comments.

So, we asked Blake Stowell, Caldera's Director of Corporate Communications What data do McBride, UnitedLinux, and Caldera have to back up those statements? Stowell did not answer that question. Instead, he wanted us to hold off until late next week (now last week), which we would not do. However, it is more than a week since that time and Stowell still has not sent his answers to our questions to us. (Please see the explanation note in the sidebar to the right.)

However, it is doubtful that Stowell could back up McBride's market-share statement with any data -- particularly if market share is Linux distributions only -- lets' compare apples to apples and bananas to bananas. A source told us that according to the NPD Intelect US Retail Software Report for July 2002 Red Hat had 50.1% of the Year-to-Date U.S. Retail Linux pie. SuSE came in second with 29.8%, followed by Mandrake with 19.7%, and all the rest 0.4%. If the entire 0.4% goes to a combination of Caldera, Conectiva, and Turbolinux and you add that 0.4% to SuSE's 29.8%, the UnitedLinux companies get 30.2% of the U.S. Retail Linux market -- and that falls woefully short of beating Red Hat's 50.1% market share -- as McBride claims.

In discussing market share, Mark de Visser mentioned: Recent IDC data puts Red Hat in the US at about 75% market share in units and revenue. The UnitedLinux distribution does not exist yet so we do not know how well that is going to do - you cannot buy it yet.

MozillaQuest Magazine: Is this 75% because the IDC data includes more than just retail sales?

Mark de Visser: Yes, a significant part of Red Hat's revenues come from direct transactions with corporate buyers, not from retail distribution.

Al Gillen is IDC's research director for system software in Framingham, Ma. He told MozillaQuest Magazine that Adding up the market share of client operating environment and server operating environment new license revenue shipments for UnitedLinux companies accounts for just over 20% of the market. Red Hat has over twice that share.

IDC (International Data Corporation) tracks all shipment channels at the origin of the shipment (so, for example, we count Microsoft shipments from Microsoft, not from retailers, OEMs, ISVs, etc., etc.) . . . Linux retail shipments are not necessarily reflective of the entire Linux market . . .

You can look at the marketing research data until your eyes get bleary and you want to just scream. There will be some variations in the numbers. But chances are that you will find that Linux-wise, Red Hat always comes out on top -- with enough of a margin to beat the combined numbers for the four UnitedLinux distributions. Moreover you will find SuSE usually has the number two or three spot -- likewise for Mandrake.

The point is that Caldera's Stowell did not show us any data to back up McBride's questionable market data claims. Nevertheless we looked high and low for some data that would support McBride's claims and we could not find any. The only data that did not show Red Hat on top was the popularity poll data. However, there it is Mandrake that comes out on top -- which is not a UnitedLinux company either.

Of course there are many different market-share and user-base metrics and data available. Perhaps McBride managed to pull one out someplace to support his claims when he was attacking Red Hat and slighting Mandrake in San Francisco. If so, then why was Stowell not able to provide that data -- without delay and without dilatory tactics?

Even if the entire UnitedLinux consortium has greater Linux market share than Red Hat, it's likely not because of any market share contribution by McBride's Caldera. According to Caldera's fiscal third-quarter financials, Linux revenues were only a meager $300,000 worldwide, or about 2% of total Caldera $15.4-million in revenues for the period. Compare that to Red Hat's $2.6-million in just U.S. retail sales and SuSE's $1.6-million in just U.S. retail sales -- oh, and let's not forget Mandrake's $1.0-million in U.S. retails sales for the first six-months of 2002.

(Caldera financials from: Caldera's Linux Revenues Are Pitiful, Linux Business Week Presents "Maureen O'Gara's LinuxGram", Volume: 1 Issue: 4 - Monday, September 2, 2002. Incidentally, that article and some other recent Caldera related articles by Maureen O'Gara in her LinuxGram are a must read! Other market data from: NPD Intelect US Retail Software Report for July 2002)

That leads us to conclude that McBride's San Francisco statements are merely VaporHype and nothing more. Have we not had enough of that sort of stuff with Enron, WorldCom, AOL, and so forth? Does the Linux community really want or need VaporHype in the Linux landscape?

Here is more of our discussions with Gaël Duval about McBride's divisive statements.

MozillaQuest Magazine: Caldera's new CEO Darl McBride has been making statements to the effect that there are, or will be when UnitedLinux is released, (only) two advanced/enterprise level Linux servers UnitedLinux and Red Hat. Is not Mandrake a player in the advanced/enterprise level servers realm also?

Gaël Duval: Mandrake is getting bigger and bigger as a server solution, whatever competitors say. Readers might be interested in having a look at http://www.mandrakebizcases.com where hundreds of companies explain why they use Mandrake as a server.

MozillaQuest Magazine: McBride also has been saying that the four UnitedLinux companies taken together are bigger than Red Hat. Specifically: "If we combined our market share, we have more than Red Hat," And "If we combine our developer count, we have more than Red Hat."

Gaël Duval: It's easy to say things. Now we need proof.

MozillaQuest Magazine: Once again, McBride does not even acknowledge the existence of Mandrake.

Gaël Duval: Surprised?

MozillaQuest Magazine: How does Mandrake weigh in here regarding market share?

Gaël Duval: It depends on what you call "market share". If you consider the number of users, Mandrake Linux is clearly number 2. Regarding the income, we're smaller. But our strategy is different: we have built a huge userbase in four years. Now we're going to transform this userbase into business.

MozillaQuest Magazine: Is UnitedLinux bigger than Mandrake?

Gaël Duval: Regarding the number of users, no.

MozillaQuest Magazine: Is UnitedLinux bigger than Mandrake and Red Hat taken together?

Gaël Duval: No. It's not my feeling, it's based on what I see in surveys. I think it's clear for everybody that the only big player in UL is SuSE. Others have very little market share and very few users. As a result, it doesn't change the ranking. But who cares?

MozillaQuest Magazine: How does Mandrake's developer count match up against UnitedLinux's developer count?

Gaël Duval: UL clearly has more employees than MandrakeSoft. But Mandrake has a very big community of developers/contributors.

MozillaQuest Magazine: McBride also said "There's increased demand from vendors who want more than a single distribution of Linux. Red Hat is one who will be the other?". Has he not again ignored Mandrake and its role in the Linux community?

Gaël Duval: Surprised?

DesktopLinux conducted an interesting, on-line, reader poll over a seven-month period. Readers were asked which distribution(s) they use (or plan to use) for desktop purposes. The results through 26 July 2002 show Mandrake a clear leader with about 32% followed by SuSE with about 16% and Red Hat with about 12% of the total vote. Neither Caldera, Conectiva, nor Turbolinux rated high enough to make it into the listings -- other than perhaps in the other category, which gleaned only 2% of the votes. Please see Figure 1.

Jill Ratkevic, DesktopLinux Producer, told MozillaQuest Magazine that the poll uses a combination of IPs and cookies to protect against multiple voting.