Reconsidering Linux (News.com) [LWN.net]

Reconsidering Linux (News.com)

Posted Jul 30, 2003 1:50 UTC (Wed) by iabervon (subscriber, #722)
Parent article: Reconsidering Linux (News.com)

I think he is mistaken in thinking that Linux distributors intend Linux to be a product; rather, it is more the case that some Linux users (businesses in particular) want to get a product, and distributors exist to fill this need. The process produces the software, and the distributors make it available to people who want the product.

There's actually been rather little product differentiation in the Linux space; mostly it's been areas where a tool was needed by distributors that wasn't needed by the community at the time (package management, in particular) or work that wasn't generally interesting to the community. Beyond that, the distributors generally participate in the process behind the scenes, or at least make their code available to the community.

Linux is really in the same position as the Internet; it's a open, standard, platform. But it also has a lot of proprietary protocols in use for things which aren't covered by the standards. As time goes on, those that are worth standardizing become supplanted by standard versions.

(Log in to post comments)

Reconsidering Linux (News.com)

Posted Jul 30, 2003 11:06 UTC (Wed) by haraldt (guest, #961) [Link]

Asking me, it's both ways.

If you look at the packages being proposed for business customers, "advanced server" and the likes, it isn't hard to spot traces of customer lock-in. Not as far as you can see from certain other vendors, but there's a definite smell.

Now on the other hand, you're right. Customers who don't know any better do expect to be treated that way. Everything has its ups and downs, customer lock-in too. With customers raised to see the positive sides of customer lock-in, how do you tell them to think otherwise?
They don't want to be told how stupid they've been. And business is about making money.

But again; is there any hurt pointing out, in a positive manner, the basis of the process and how they can gain from it?

GNU/Linux software has shown to be more cost efficient by numerous independent studies. As I see it, mainly by reduced sales effort and license handling, closer bonds between user and developer, and less room for making a few people famously rich.
If IT budgets have basis on the usefulness of IT deployments, it's easy to conclude that the efficiency of GNU/Linux helps the IT market. Or at least, gives it the strong edge needed to keep it stable.
Money saved on IT deployment gives more room for educating the users, and for better, faster IT deployments. Spending money on support and development is still needed, and free software can help the market accelerate by focusing into these areas.
Thinner. Faster. Reorganization? Yes. Loss? No.

It's a valid comment, that one above here, well worth. But I can't see the need of product fixation either. If users can learn to appreciate the process and make the most out of it.

Reconsidering Linux (News.com)

Posted Jul 30, 2003 15:48 UTC (Wed) by iabervon (subscriber, #722) [Link]

Even if distributors do their best to make it easy, there's a substantial cost is changing systems, simply because the standards, in order to be flexible, must leave some things undefined, and you are likely to have become accustomed to the way those things are defined by your particular vendor. The goal in openness cannot really be "no lock-in", but should be "no more lock-in than the user's habits".

In the "advanced server" area, there is a substantial amount of lock-in where a vendor has added support for something that other people haven't added. You're stuck with that particular vendor until someone else puts together the support you need (or you do it yourself), but that's not the fault of your vendor.

Using the process is great if you're interested in the particular thing you're using, but using a product is better if you only want to use the software to do something else. My home desktop installation is entirely put together from source released by the various projects. But I'm still only using the release tarballs from these projects, not code checked out of CVS. I'm using gcc productized (to a very slight extent) by the gcc team, because I'm not interested in debugging and developing a C compiler, just using it to compile other code. I've participated in the process with jFreeChart (contributed a chart I needed), and slightly with make (fixed an obscure case), but it's impossible to use the process with every piece of software you have and engage the entire development community.

It's important that the process exists, and that you can escape lock-in if you are actually unhappy with what you're using. But an important portion of end user choice is the ability to ignore the choices until there is a need to change.

Reconsidering Linux (News.com)

Posted Jul 30, 2003 15:59 UTC (Wed) by allesfresser (subscriber, #216) [Link]

If someone wants to stay with a platform or process for a given amount of time, that should be their choice. In other words, if they want to lock themselves in, that is their option. Vendor lock-in takes the key away--in essence you are forced to keep paying to stay in prison. And I'd rather a developer spend time on making the software smoother than figuring out new ways to keep me locked in. I don't consider a vendor who tries to make my life difficult very friendly.

Reconsidering Linux (News.com)

Posted Jul 30, 2003 18:02 UTC (Wed) by haraldt (guest, #961) [Link]

It should be important to keep a perspective here. And a note about the author and where he comes from.

The article does not diss the need of stabilized, customized distributions. To quote:
"Sure, Linux distributions--integrated collections of software combined with the Linux kernel to form complete operating systems--have been the primary drivers of the commercialization of Linux over the past several years. And the distributions do play a vital role in the Linux ecosystem. They collect "best of class" open-source technologies and integrate them into complete offerings. This allows Linux to be distributed to end users as an integrated solution as opposed to just a collection of disparate technologies."

But our author is also the "Ian" of "Debian", the founder of that general-purpose GNU/Linux distribution. By Debian, you're not required to compile your own binaries to have a wealth of choices. Rather, one of Debians strengths is not to make more choices than you need.
That's not much help for a beginner, but a savior for those who'd otherwise compile their own.

Debian is a platform for custom installations, for those who need to customize some and keep the rest at minimum hassle. Soft lock-in, by your own choices. Not because someone told you "this is what you bought".
There are custom selections and customized Debian-based distributions. All you need to do something different is to open up the archives and make your choices.

And from that perspective, I still can't see the wrong of it.

Lastly, you noted:
"You're stuck with that particular vendor until someone else puts together the support you need (or you do it yourself), but that's not the fault of your vendor."
That's highly dependant on who you ask and the requirements they have. Agree there's hardly a base to drag the vendor to court. But for someone who don't like it that way, that is a fault. Should the customers be blamed?

Reconsidering Linux (News.com)

Posted Jul 30, 2003 21:43 UTC (Wed) by MathFox (subscriber, #6104) [Link]

When I was at a "Linux in business" symposium I realised that a lot of the "decision makers" still think about software in terms of suppliers and users as totally seperate entities. They are interested in becoming users of open source systems for mainly economical reasons: create a more efficient IT environment for their own organisation. They don't realise the additional benifits that open source software can provide for an organisation.
I have the impression that the software world will look significantly different in 5 years (at least in the Netherlands; the US will be slower than Europe). A significant foundation for large scale open source deployment is under construction. That could be the reason that juridical actions against Linux have been started...
Anyway I think a distribution as Debian has its place to keep the "commercial Linux" distributors honest. It remains important to keep an critical eye on them, but as long as there's a symbiosis between them and the open source movement we should respect their business