Slashdot Mirror
Slashdot talks with Red Hat
One of our first and most pressing questions for Donnie was, "What the heck are they going to do with all of this fresh money?" The answer, as one would expect, was some mumbo-jumbo, with some hints of future plans. Barnes said, "We will certainly grow development, marketing, sales, etc." And to whet the appetite of users, "You can certainly expect the 'grow development' part to include some neat new toys for Linux, though we don't preannounce new features. ;-)."
So the company itself will be growing, as we would expect, and the staff will be growing as well. However, what's interesting is the comments on where it will grow.
According to Barnes, "International expansion is a key area for us," which is "very expensive." It is partly due to the sheer expense that they have not really pursued this area before. Some of the shifting in this direction can be seen already, in actions like the severing of the relationship with their former Japanese distributor. In that case Red Hat "basically chose to enter the Japanese market in a bigger way and bring our products and services into that area directly." At the current time, they plan to focus on a European expansion first, to be followed by an Asian expansion soon after.
Waidaminute! Europe? Isn't there already a certain Germanic company established over there? Well, yes. We pressed Barnes about moving into what's considered to be S.u.S.E. Territory, and, on a broader scale, what's going to happen with the expansion of Linux-based companies and competition that will happen between them.
Barnes's response is one that Bob Young has been giving for quite some time; that rather then trying to play a zero-sum game, taking users away from each other, the aim of the Linux companies is to make the entire pie bigger so that /every/ company can grow. Well, every company but Microsoft, I guess, but the point is still there. Both CT and I find this to be a bit idealistic, but given the current atmosphere of change, maybe it will be how things fall out - we'll see how things shape up over the next six months to a year.
So with all this money: International Expansion, more developers, marketing and saleses staff. And, for the record, "...no, there are no big ticket surprises like Lear Jets..."
However, Marc might be purchasing a Porsche dealership. (Note: Joke! Don't sue me, Marc. Please).
The question that everyone had, whether they got the letter, didn't get the letter or just watched the fireworks was the E*Trade/Red Hat stock situation.
In summary, the numbers broke down as following:
5000 open source developers were invited to participate
1300 responded, indicating interest
1150 of those were ultimately able to participate
Looks pretty good, right? It'd be interesting to know how many of those 1150 had to go back and try again, but c'est la vie. When asked about RH's feelings on the stock offering, Barnes said, "We feel that an over 88% participation rate was very good for the community at large." We talked about the people angry about the situation, whether or not a lot of the people originally banned were able to get in. His response: "Most of the angry people were [able to get in], but the bumps we had in the way caused them to be confused as to where they were and what level of participation they could expect. But at the end, most of the angry people were able to participate and I think came away happy."
What I thought was interesting was what precipitated the whole meltdown: Bad Information? Clash of Cultures? Red Hat's answer: "Mostly lack of information, and some bad information from E*TRADE's 'systems.'"
Additionally, "Sure, the open source community meeting big business and high finance was part of it too. Both sides needed education about the other, and I think lots of education happened. :)" I'm sure many of the companes looking at similar situations watched the RH IPO closely, trying to figure out what they did right, and what they can do differently. Barnes recommends examining what happened with them, and also shopping around. The latter part to that comment is the interesting part: Not dissing E*Trade, and Barnes does say they would at least shop them against other brokerage houses, but I think that the lesson learned here is to make sure of the house you use, especially after watching the Red Hat fiasco.
And given the recent hoopla over Unisys's perceived attempt to charge for GIFs that use the LZW patent, and Red Hat's closeness with major patent holders like IBM and Hewlett-Packard it seemed appropriate to talk about software patents. Red Hat's take on software patents: "Well, I think it should be fairly obvious that software patents are a bad idea in general and we are not in favor of them." Additionally, when it comes to big-boys like IBM, HP and others they "...will be working to try to lobby in the proper areas to help this issue. Yes, we do care about that stuff and have already begun talking to folks." (Note: I envisioned Bob Young with a flamethrower in the IBM Corporate Lobby)
Despite this, Barnes allowed for the fact that "Given current regulations you may see RH own patents one day. That may be unavoidable until the current way they are handled is changed." RMS would be displeased, but Barnes said that "What folks do with these patents, otoh, can be vasty different." That's an odd world, but I think it's just a further sign of how messed up the US Patent Office is, regarding software patents.
The issue of software patents led to talk about the other software licenses that have come out, like the APSL and Sun Community Source License. As you would expect, Red Hat would like to see fewer open source licenses, but "...at the end of the day, as long as they are truly open source and companies are *using* them, I don't really care." Barnes is pleased to see that Sun is taking steps to making good software more widely availible. I asked if Sun was viewed as being a competitor. Barnes says that perhaps, but that RH is trying to make fundamental changes to the way software is built. "Perhaps Sun will join. They have taken steps in that direction so far, so they could be an ally."
Oh, and again for the record, Hurricane Dennis got Barnes "personally quite wet."
The last question sparked a lot of back-and-forth, regarding the whole Red Hat trademark issue.
(Note: To get background on this, read the original article, and the subsequent follow-up with a letter from Bob Young.)
The following, according to Barnes, is what's going on: "We have to protect our trademark. People were selling hand-burned CDs and using our name, box image, and logo to do it. We had customers complain vehemently to us directly about that as they felt burned. We had to take action to keep that from happening." This was what sparked the original issue with Amazon, and how it appeared that Red Hat was banning anyone from using their name in any way unlicensed by them. The issue, then, is, how do people know what they're getting? For example, people want to know what they're getting when they buy Cheapbytes/LinuxMall/Bob's1$LinuxEmporium/LinuxCentral Linux. Are they getting Debian, Suse, Red Hat, Caldera etc.? Recognizing that Red Hat has a valid brand they must protect, where does that leave the manufacturers? RH's answer: "If they ship "Red Hat Linux" as their own "Cheapbytes Linux", then it must *be* Red Hat Linux. We can verify that with PGP signatures and the like. They can say it contains Red Hat Linux, but they can't explicitely *call* it that." For example, under this scheme, they could "call it `Foo Linux which contains Red Hat Linux 6.0' or something similar." What they can't do, however, would be to release a CD with a PGP signature that checks out, but call it Red Hat Linux. Only Red Hat can explicity title their distribution "Red Hat Linux". And if a company puts out a CD saying it "Contains Red Hat" and the PGP doesn't check out (e.g. modified from original or something, then "they can't say that and we'd have cause to have the product sales stopped."
Phew. So, how do you handle something like Mandrake-Linux? Can you call say "Derived from Red Hat Linux?" Barnes points out that he's not a trademark lawyer, so don't hold him to be a legal litmus test, but that, yes, that will be possible. (Note:More guidelines on this will be coming out soon. Keep an eye on Red Hat's site for details.)
But when it comes to recognition, of course Red Hat appreciates being recognized for having contributed, e.g. Mandrake putting "Derived from Red Hat" on the CD. As Barnes says, "Acknowledgment of one's hard work is always welcome. We just want it done properly... they can take Internet bits and build their own name. You don't see us claiming the name of every other distribution that contributed some code to RH, do you? But they can take our code and use it. That's what this is all about."
Wrap-up:: Interesting to talk with Donnie, and you can see that the quiet period has been hard on them - being unable to respond to comments would be, I suppose. It seems that the basic direction of Red Hat remains unchanged; that is, commitment to the GPL, working with the industry on software patent issues, and trying to get people to join the "movement." The greatest concern that CT and I is how all of this money will affect them as an organization, but I suppose that's an issue for the greater Linux universe as well. Only time will tell. And, on that platitudianal note..
disclaimer:Hemos owns shares in Red Hat
Just some points:
- Where is the sticker on RedHat's distro paying tribute to all the people who contributed to their distribution?
- Mandrake isn't a RedHat clone any more. It started back with 5.1 with some adaptations but has now aquired a reputation of its own. 'Derived from RedHat' may have been true for versions up to 6.0, but bringing up this topic now - after the Axa investment - has some foul taste.
- Where are the CPU optimizations in RedHat? Starting with 6.0 Mandrake has recompiled every single package for optimum speed. I mean, if you compile everything on your own, make customized configuration and boot files that would qualify for being called a new distribution.
- Mandrake keeps compability with RedHats file hierarchy because we think compability is a Good Thing(TM). But this doesn't make us a RH clone. Should we fragmentarize further by making up a standard of our own, just to avoid being forced to advertise for RedHat?
Regardsan one time AC ;-) (I love this new feature!)
I think that the most profound lesson that any of us can take from the past three months is that we can no longer implicitly trust Red Hat, either as a product or as a corporate entity.
I guess things had to go this way, but it is disturbing as more and more of my career is based upon Red Hat. I would have liked to keep the trust up at least a little longer.
We've seen two main problems recently: the IPO, and Raster's departure. Raster asserts that E will no longer be Red Hat's choice of a window manager, to which Red Hat has not replied. More GUI confusion is the last thing that Red Hat needs at this time.
I think that Bob Young should have wished Raster well in a public way, then confirmed or denied if E will be ejected from future distributions.
These problems illustrate bad management. I only hope that the ominous trend does not continue.
As far as CDs go, I am shortly publishing material that will include Red Hat Linux. The legal staff is having difficulty in getting permission to merge the updates into the distribution. If I can't do this, there won't be room for all these updates.
IMHO, whenever the updates exceed 50MB, there should be an automatic new version with the updates applied.
Red Hat needs to get its act together in many ways, but this would be a good start.
Heh, as long as Linux or UNIX don't become as fragmented as Windows, we should all be fine. The Macintosh managed to switch architectures with less trouble than Microsoft has with its (many, user level?) libraries every day.
:) ...and (b) I started out before I went to NCSU trying to use Debian, and I didn't like the way I had to install the packages. However, we all owe a lot to the earlier distributions, like the nifty color-text Slackware installation program with all the package descriptions! Yay! (don't you wish Windows could do that?)
I'm glad that Red Hat and all other linux distributions, and any good UN*X have standards on how to handle shared libraries correctly, and don't have incompatible versions of their programs for different versions of X and whatnot.
Use whatever operating system / distribution you want, no one is forcing you do do anything. I use Red Hat because (a) I go to the school where it all started, so I guess there's some local pride there.
pb Reply or e-mail; don't vaguely moderate.
Dividends are one way of getting a handle on the "real" value of a company and hence it's stock. They are most important in industries which have stable profits and slow growth rates, such as electrical utilities.
They are not the only valuation method that is valid. Clearly when a company generates profits that money goes somewhere. If it is payed out in dividends it gets taxed twice (once as corporate profit, and once as income), and is often reinvested in any case. If instead the company reinvests that money either in it's own operations or by acquisitions, the assets (book value) and purchase price (market value) both naturally increase. The stockholder can get the effect of dividends by selling a small portion of his stock holdings if he wishes and, if he has held the stock for a sufficient length of time (1 year?), he pays only capital gains tax at 15% instead of income tax.
In either the dividend case or growth case, the key is profits, and more importantly the expected growth in future profits. The only other reasonable basis for valuation is the expected value of a acquisition of the company by a larger one.
That being said, I agree that Red Hat, and many internet companies can't justify their collosal valuations. A few, like eBay, are profitable and have the potential to become much more profitable without much more costs by leveraging the network effects of their clients and customers.
I don't see Red Hat being able to make Microsoft's profit margins without Microsoft's market control and proprietaryness. Their support model works well but the costs seem to scale fairly linearly with revenues. Boxed software sales are fairly high margin, but are hard to support without proprietary components (which the current boxed sets _do_ in fact have, just not much).
So the bottom line is that Red Hat, and all other recent IPO's, don't have to have dividends to justify their price. But they will have to have profits, and soon.
--
"L'IT c'est moi!"
Redhat made an announcment yesterday that didn't even get posted on slashdot.
h _vone_2.html
This announcement seems to have started the upward trend.
Basically they're teaming up with another company to make extranet server software with encryption.
http://dailynews.yahoo.com/h/nm/19990907/wr/tec
Easy.
"joebob fixed a nasty/esoteric bug dealing with 3c59x driver".
Does it need to highlight *exactly* what you contributed? No.
Do academic papers clearly delineate where one's work stops and another's starts? No.
For example, if I'm working off of some student's graduate thesis for my own work, unless I'm quoting directly, I talk to the person first, let him/her know what I'm doing, and go from there, in exchange for me clearly saying, "this idea of mine is clearly based on an idea first espoused by Dr. Soand So". Credit given. Pointer to the exact source listed in the references/bibliography. If you want to see the original, you get to look it up yourself.
I don't see where the problems are.
No, no. I meant how does it keep track of each person's contributions. For example, if I were to find a bug in wuftpd, and the authors accept it for inclusion into the main source tree - according to some licenses they need to give me credit. The problem is that certain licenses (*cough* GPL *cough*) require you to note the contributions of everybody if you redistribute. Therein lies the problem - how can redhat guarantee compliance with those licenses? And the bigger question - should it?
--
It would be interesting if Red Hat (or some other free software related company) started registering software patents and allowing free use of them for free software development.
They'd then be able to enter into the kind of "patent swap" deals that big companies do with one another on behalf of the free software world.
The 1,150 number given by Donnie has to come from E*TRADE. There's no way Red Hat can know the exact number, except from E*TRADE themselves.
Now, take a look at this article. If you read further on, you'll see that many people reported that E*TRADE was giving them only up to 400 shares.
According to my abacus, 400*1150 does not come out anywhere near 800,000 shares which were allocated to this directed share program. People should've been able to purchase at least 700 shares, using these numbers, assuming that everyone applied for at least that amount. There were many people, it seems, that only applied for 100 or 200 shares, so the excess amount had to go back into the pot for people who applied for more shares.
Also, I seem to recall that earlier E*TRADE claimed that there were 2,000 participants left, but I can't find that reference anywhere. If anyone kept the link to that story, please forward it to me via the feedback form on my web page.
Now, this 1,150 number also comes from E*TRADE, and I'm suspicious of this number as well. Let's refresh our collective memories, and go back to the previous Slashdot article. The day before, if you recall, E*TRADE basically threw everyone out of the directed share program. Their excuse, as far as anyone could tell, was that everyone must reconfirm their application because the IPO has been repriced at $14. Furthermore, people claimed that they had ridiculously small time window where they were supposed to reconfirm.
The extension announced in that Slashdot article came about as a result of this latest move by E*TRADE. Now, take a careful look at that telephone number where people were directed to call. Go back to that Slashdot announcement, and read it.
Now, that doesn't look like the number of E*TRADE's toll-free call center, does it? It looks to me like a telephone number for one of E*TRADE's offices. Reviewing my files, I believe that this telephone number really belongs to E*TRADE's Legal and Compliance division. Apparently, this whole mess was kicked up to them to handle, after being horribly botched by their lower-level staff. Or, it might've been Jason Saxon's extension, but in either case, this is a single telephone line.
Now, the extension of the reconfirmation interval was for one business day only. Can anyone please convince me that E*TRADE fully expected to handle almost 1,200 calls on that lone telephone number within the space of one business day??? Can someone make that case for me, please?
Now, as far as I can tell, E*TRADE was legally required to distribute that 800,000 shares allocated to the directed share program, unless there was insufficient number of applications. By all accounts, E*TRADE put an upper limit on the maximum number of shares anyone could've purchased, and this number of 1,150 participants just doesn't add up. Just doesn't add up at all.
--
The GPL has no such credits requirement. Atleast I don't remember ever seeing it. Is it possible you are thinking of the advertising clause in the BSD license (which has been removed). AFAIK, the reason BSD and GPL were not compatible was the advertising clause, which seems to indicate that the GPL cannot have a credits requirement of any sort.
-matt
Competition is a good thing if it is kept fair.
As long as RH does not resort to the same tactics that a certain other company uses like buying out and threatening its competiton, it will be good. It will force all these companies to produce better products, that will have to be more stable.
This in turn should mean that some of these distributions should be super easy to install for newbies, and super easy to configure for newbies, and hopefuly more 'english' like doocumentation other than man pages. Yes I have used man pages, and often I understand them, but some newbies may not. Gnome has a good start to an interface to man, but there are some features lacking.
Don't get me wrong I am very happy with my configuration, and distribution. There are a few things I wish I had and if I get the time will write myself.
I hope that this competition in Europe will make SuSE open up a little in there Licensing of there YaST and there distribution too. Currently SuSE has 'evaluation' distributions, which IMHO I think is bogus.
I wonder whow this will affect the Easy Linux distribution. Maybe well see cdroms of there show up on linuxcentral, and linuxmall. I'd love to test them out.
Only 'flamers' flame!
Four, actually... Burlington Coat Factory also signed a fairly good sized support agreement with RedHat yesterday.
I couldn't get my broker on the day it opened, and the *cough* idiot decided to buy it on the second day at 86, so I was a little grumpy, but hey! I just wanted to own a piece of the OS that I run :) so I was in it for the long haul regardless. /., hmmmmmm mayhaps Rob and the gang have some vested interest?????? (j/k), but it has been shown to be true
So I've been VERY happy the past day or so watching it generate at 120+ (Thanks Mom)
but now I'm wondering if someone could do a little interpolation to see this phenomenon.....
It seems like the stock climbed everytime that there was a story posted on
-- Life: Hate the Game... Love the cereal
You make a pretty strong claim about how the community can no longer implicitly trust Red Hat. Your evidence? The IPO and Rasterman's departure. Yet you don't state why the IPO was a problem, nor do you seem to recall that Raster attributed his difficulties at Red Hat to a particular manager. I agree with another poster who said that a public response to an individual leaving a company is highly inappropriate. I would go further: it is inappropriate for Red Hat to respond to claims made by a former employee about the direction the company is taking.
In other words, I disagree that the events illustrate bad management, and I question the use of the powerful term "ominous." Furthermore, I question how professional it is for you to make these public claims when you're having private troubles with the company in question. I've always considered it illegitimate to air one's dirty laundry in public, particularly before there's been a resolution of the situation.
In short, whether you intended it or not, this post smacks suspiciously of Fear, Uncertainty, and Doubt. I would hope you would either substantiate or retract these claims, preferably the latter. For some reason you've been moderated pretty high, and I fear that you're spreading ill will without sufficient justification.
Hemos wrote:
"Waidaminute! Europe? Isn't there already a certain Germanic company established over there? Well, yes. We pressed Barnes about moving into what's considered to be S.u.S.E. Territory..."
This strongly implies that Hemos doesn't expect to see, or doesn't want to see, Linux companies battling each other. I've seen this from other people as well. What's up with this? Everybody wants competition with Microsoft but no competition with each other? Let Linux distros duke it out and the best man win. I agree with Donnie Barnes, that while some companies may make more money than others, there should be plenty to go around in the post-MSFT world.
Linux is already being proclaimed as the alternative to Microsoft. What I see happening is that once a person has decided to go to linux, they then have a number of choices. The companies that succeed will be the ones that distinguish themselves from the others. I do not see the competition arising in order to capture each others market share, but from gaining the attention from those that are sick of Microsoft.
Right now, that company is Redhat... and they are smart to further the Wall Street iterpretation that they are indeed, the embodiment of linux. They will always have the advantage of being the first. But the first often fail.
It is like all the E-Trades out there.... their *real* competition is not the full service brokerage firms that they attack in their ads, it is the other countless on-line brokerage firms that have sprung up. Those ads are not as much to entice people from the Merril Lynch's out there... but to differentiate themselves as the best alternative to that.
I believe this market will soon grow to mirror that.
Sure, the Linux distribution vendors are competing with each other. JoeBuck gave you that in his posting .
The question is, is this a good thing? Will they attempt to differentiate on value or try to lock you in with incompatibilities, as the Unix vendors have done?
There's been a lot of talk on /. recently about ethics. I've always thought that the practice of going out of your way to decrease the value of a product to increase revenue is unethical. Hardware vendors used to do this all the time. The fact that the first 486 SX chips actually went through an extra process step to disable the floating point processor is an example of this. Intel later developed a 486 core without the floating point processor, but doing that initially was wrong. I could list off a half-dozen less famous examples in hardware, but you get the idea.
The same thing happens in software. A company will implement unnecessary source level incompatibilities just to make porting large systems difficult. POSIX helped some, but it still happens.
Perhaps a summit of the Linux distribution vendors is in order. Maybe someone could draft a "Treaty" that all will be asked to sign. Signatories would be allowed to display some Trademark in advertising. Something like "Linux Common Compatibility Assured - Version 1", or something that captures this idea.
At a minimum, it would assure that all Linux distributions support some set of FSF compilers/libraries/headers, that some "standard" set of X Windows headers/libraries/Server be available. Maybe a "standard" Perl, Python, Tcl/Tk, GTk, etc. etc. will install and work directly on the system with a provided installation/configuration procedure.
The vendors would be free to provide newer releases of any of these things, but the distribution should work with some agreed upon set.
The vendors product would be "out of compliance" if it is shown that they are engaging in a bad faith effort to break any of the agreed upon software interfaces. It's too much to ask the vendors that all of these software interfaces work to some level of compliance. The validation would be way too expensive and lead to endless debate and problems, especially with regard to validation on different architectures.
What do people think? Would this address a problem? Or would this kind of compliance be just another check list kind of thing that didn't mean very much?
Lord, how do I wish for such a situation as what you have us imagine...
:)
:)
This is the same reason why I don't fear about redhat - in fact, it's why I think Red Hat is probably the _greatest_ thing to happen to Linux in a long time. Not for what they may bring to the project as a provider of source and whatnot (which I don't know all that much about anyways), but for putting a corprorate lookalike face on Linux. Which is obviously not what's REALLY going on (Red Hat != Microsoft for all time), but the sort of endlusers who normally buy Microsoft 'cause it's Microsoft can't see any different.
RH's duping the endluser, like Microsoft, but it's in a _good_ way
I can see it now... "Nobody was ever fired for buying Red Hat" - at least that there's a company who can't dictate the important parts of policy if it ever gets taken over by some troll and gets corrupted...
(Now all I have to do is hope that RH doesn't start getting into anti-MS FUD like so many Linux supporters seem to go for...)
"I don't believe that there is one, single, perfect spiritual way and, in realizing that, obviously you become a lot more open."
If we do not change our direction we are likely to end up where we are headed.
I haven't heard of anybody getting more than 400 shares. 1150*400=460,000. Redhat set aside 880,000 shares for people in affinity program. I guess there are manyh people who apllied for 100 or 200 shares. It obviously means some people got in 1000s. Is it possible for Redhat or Etrade to give us some statistics on who got how many shares perhaps? While I am happy to have got 400, I am curious to know what the criteria was to allot more shares.
The Unix companies all thought that their main competition was each other, so they messed things up and did their best to break source-compatibility between the different Unix flavors. Meanwhile people were tossing out their Unix systems and putting in NT.
It would be a disaster for Linux if the vendors start thinking of each other as the primary competition -- because it means they will be motivated to break compatibility.
SuSE is certainly competing with Red Hat, but each company can do far better by taking away users from Microsoft than either can by trying to take users away from each other.
I think you should ask the press where they got
that story from, not assume it was ever closed..
I'm afraid I could never understand the big noise concerning Red Hat's protection of their TradeMark.
.... *CRUNCH*)
If you remember, the whole point of the GPL vs BSD licence was that people *know* who did it, i.e. you put your mark on it!
Red Hat only gets money if people are willing to pay more for their distro 'cause they know what they're getting.
If RedHat doesn't protect that, then what do they get from Open Source? They're a company, not a cool individual who does neat stuff fro fun. They must have a bottom line or they're out of the game.
No, I can't spell!
-"Run to that wall until I tell you to stop"
(tagadum,tagadum,tagadum
-"stop...."
What will happen in the event that Red Hat does decide to file patents? Will they be donated to the FSF?
Beer recipe: free! #Source
Cold pints: $2 #Product
The current system appears to be either give people credit inline with the contributed source code, put their names under the authors/contributors section, or create a dedicated CREDITS file to list contributions in more detail. None of these systems is ideal.
How is redhat (and obviously the larger free software community) dealing with this problem?
--
It will be interesting to see how Red Hat fare in Europe. As the article points out, SuSe have a very strong base in Europe, although in my experience Red Hat still have very strong recognition here. It's not difficult to find Red Hat here in Sweden for instance, as long as you know where to buy good quality software.
;-).
I would be interested to know more about how they plan to reach the market here (Europe). Obviously getting RH into the stores is the main thing, but are we likely to see country customised websites, advertising campaigns and the like over here? I guess we'll just have to wait and see.
In my experience, European marketing is often very different to US marketing. I have frequently criticised Netscape in the past for concentrating on the end-user server market here, when leased line costs are so expensive, that the vast majority of companies host with ISP's. If Netscape had focussed on selling to ISP's first and cooperating in helping them deliver new services, they could have a way bigger market share. As it is, most ISP's here offer IIS for database driven sites.
My point is that Red Hat will need people with local experience and be prepared to listen to them if they want to crack the local markets (you guys listening? Drop me a mail
A little planning goes a long way...
Do the math to see why RedHat's strategy to grow the market for Linux is the right one. It boils down to if you want a potential market of half a million or half a billion.
Here are some made-up numbers: Say Linux has 10% of the total OS market. Of that RedHat has 30%, or 3% of total, and the other distributions have 70%.
Who has the other 90% of the OS market? Microsoft of course! So RedHat can either fight the other Linux distributions to get at the other 7% or they can target the remaining 90% that Microsoft has.
Instead of a big slice of a small pie, give me a smaller slice of a much bigger pie.
Like it or not, to the average CEO/CIO/CFO it's Linux versus Microsoft. I just got my official Idea shot down at the place I work: to use old boxen for Linux test DB and test Web servers for the developers. Why? Because "at this time, the platform is still immature and there are few stable applications that have been ported to it." Oh, and the misconception that using Linux "will not reduce the failure rates nor make them Y2K compliant."
So, yeah, around 2002-2005 we'll see the Linux Wars (TM) between Red Hat, S.u.S.e., and whatever distros are doing ok then, but for the next few years the big boy on the block is Microsoft.
If I was Red Hat (wait, I own shares - guess I am), I would offer an O2K disaster relief program. For anyone who installed W2K and O2K, offer to give them a $24.95 upgrade to RH 6.2 with StarOffice with full docs and support, so they can get their servers back running in production mode.
Cheap, effective, and gets MSFT where it hurts.
Will in Seattle
What RH is doing is successfully bringing Linux into the corporate world, by showing it can be neatly supported, that it can be marketed successfully and that it is a viable business solution with a company front. Unfortunately, no amount of good-willed hacker PR can achieve that. The market is a place for suits and ties.
Even if RH turns into a Microsoft for the year 2000, they still won't dictate the core, only affect the public's perception of Linux by altering its distribution. What you end up with, once the fluff is over, is still the same old GPL apps.
Imagine if DOS has been Open Sourced from the get-go, and Microsoft then built their own distribution of it, including a custom-built GUI called Windows. You could download the DOS kernel for free, try other GUI's if you wanted, and Microsoft would be busy trying to enhance the user-friendliness of Windows, not trying to pass it off as a decent OS.
And as long as I could hack the OS and decide which GUI I want with it, I'd be fine if Microsoft was in that position. So I don't fear about Red Hat.
"There is no surer way to ruin a good discussion than to contaminate it with the facts."