Will Linux advocates be the ones to doom Dells Linux desktop push?
Im watching the progress that Dell is making with their second desktop Linux effort and am increasingly wondering how long before the Linux supporters make it clear to Dell this is a bad idea. While some are being patient, it is becoming increasingly clear to me that many dont fundamentally understand why this is vastly more difficult than it looks, and why Dell will desperately need their support, not their constant criticism, to justify continuing the effort.
This week let me try to explain why it is nearly impossible, but not actually impossible, to do desktop Linux from an OEMs perspective and why, right now, most of Dells competitors are betting Dell will fail in this effort.
Lets be clear: I dont think Microsoft will cause this to fail. I dont think Windows users will cause this to fail. Im damn near convinced that Linux supporters will cause this to fail, and that is the problem that will need to be addressed if this is to succeed.
Linux isnt Windows. Yes, I know you think you know this, but do you really understand the differences from a desktop deployment perspective? With Windows there is a massive amount of support that an OEM gets not only from Microsoft but from a secondary ecosystem that is now decades old and fully tested. This system is integrated into everything from product design through product testing and delivery and has no real analog on the Linux side. And, if anything goes wrong, you can almost generically blame Microsoft (whether or not it is actually their fault).
Expectations, however, often seem to be on both sides that the experience will be similar. It cant be. For Linux to work, it has to depend on the Linux support structure that is embedded in the shop that is deploying Linux.
This means Dell cannot realistically give you the distribution you want on the hardware you want all wrapped up in a bow with full support. Once they know what you want and what drivers are already available, they should be able to give you a few choices that will work with the distribution you have already chosen to apply yourself. They may suggest another, similar, distribution and you should seriously consider their suggestion if you want the result to meet expectations.
They cant fully support Linux as it is likely to be deployed, but they can help you support it. This means that much of the break/fix capability has to remain in your shop and you need to do a better job of initial product testing, because once the system is imaged, Dells capabilities for supporting it will be dramatically reduced.
But, if you think about it, isnt that the Linux way? Linux is supported by a community, not by specific hardware vendors, and the desktop cant break that model. The result is to truly be Linux and not a disappointing Windows want-to-be.
While this may change eventually, initially there is simply too little volume, too little experience, and too many possible software choices to allow this to work profitably any other way.
Dell Needs to Make a Profit
The typical Linux desktop, or server, purchaser has a tendency to buy very little and look for products that have little or no margin. If that becomes the case here then it is in Dells best interest that the Linux effort fails because they wont be able to stand the related margin erosion. The more Linux they sell the worse that erosion will be, and their defensive move will to make it as difficult as possible to get systems.
If you want their support you have to look for ways to ensure Dells margins. Because you will be buying very little software from Dell, you should be looking at mid-priced systems and closer to retail prices if you want them to eventually expand this coverage and need other large vendors to follow Dells lead. Once other vendors enter, you can competitively bid again, but until then if you can find ways to give Dell more of your high margin business and connect it to the Linux purchase youll be creating a sustaining model, and not one that is short-lived and likely to fail out of the box.
If you truly believe Linux will save you a lot of money you have to help make the solution financially attractive to Dell so they continue to provide and enhance it. They are effectively learning by doing this initially, and will be incurring higher costs themselves as a result. If they too can show reasonable profitability even during these relatively inefficient times they are more likely to gain the internal support needed to fully fund the effort and make it strategic.
Typically, OEMs and their customers, particularly with new efforts, are at each others throats because of misunderstandings resulting largely from the fact this is a new effort. Dell is taking a huge risk doing this, they dont have to, and they are doing it because many of you asked them to. To now go back and crucify them because they cant get it right the first time largely because no one yet knows what right is would be stupid. And stupid isnt a word often associated with success.
Cut Dell some slack. And if you know of something that will work better let them know, but if they dont agree, dont shoot them for it. I realize the Linux culture has a tendency to drift solidly into personal attack when things go badly but, you do that here, or let someone else do it, and Dell is likely to back away from this effort sharply and it will be at least 5 years before anyone tries it again if they fail with this effort.
There are a lot of people, both inside and outside Dell, that are betting this thing will fail big and have their I told you so speeches, articles, emails, and columns ready to print. If you want to make these folks really happy, by all means, make this experience for Dell a living hell. If you dont, back the company up.
In fact, even if it doesnt have to do with Linux, if Dell saw Linux folks coming to its defense in other areas when it needs vocal support, they are much more likely to see Linux advocates as friends rather than the necessary evil that often follows them and their Mac counterparts in IT organizations.
What Im suggesting here is if you want Dell to be successful with this desktop Linux effort is to think strategically and accept a little shared pain to make the needed changes happen. You are moving against a set of practices that have been in place for decades and that kind of movement doesnt come without pain. If you accept more than your fair share, it is likely that the effort will, despite strong odds against it, be successful. But if you ask Dell to take on most or all of this burden it will crash and burn like the Hindenburg.
Linux, to work in software or hardware, has to be a cooperative effort between all of the parties. Thats its strength. Where it fails is when it is asked to also showcase the benefits of a proprietary offering. With change comes pain, if you believe the result is worth it, the pain is part of the rite of passage and, if shared, could form a bond between you and your OEM that could be stronger, and more mutually lucrative, than you have ever seen before.
That alone may make it worth doing this right.