Slashdot Mirror
A Bold Essay From Tim O'Reilly
skydryedblue writes "On XML.Com, there is an interview in which Tim says, that The Linux community is far too focused on the battle with Microsoft's current operating system. Some see that the big goal is to develop a competing desktop and compatible desktop applications. And while I think that's a worthy goal and Linux is doing pretty well at it, I see Microsoft much more clearly and strategically focused on what kind of software will be needed to support that next generation of computer applications, and that worries me. "
I believe with atleast Netscape Mozilla 5.0, the open source is showing an adoption of standards, which is a good thing for any market.
But linux, seems to be in a "wholy ware" That is dediced based on how you conform. Whats so different from being a Microsoft Biggot to a Linux Biggot? Bot OS's have there advantages and disadvantages, but someone has a direction.. and i agree, microsoft holds it.
While Redhat, and other distributions have a release map that seems to be 6 months (from previous discussions, i still find that too short of a release cycle) there is *no* roadmap other then directly related to the kernel. There is no beta system to show a developer road map, no enduser solution map, and no training map.. training seems to be specific to distro, which doesn't mean doodoo to an admin (since all distros are familiar) but to a corporation/business specifics do matter.. so with all these certificaitons going, there is no specific cover all certification roadmap.
Like i've express my OPINION before, i believe for linux to be stable, for personall and business use, it needs a long term plan, and short term updates. It doesn't need short term plans and long term updates :) By what i mean, slow down the release schedules, Business like 3 year product life cycles.. that means from 6-7 there should be a good few years to get your monies worth (yeah, yeah, its free. but installation/support in a business environment isn far from free). .1 releases should be minor upgrades or patch releases.. kernel upgrades should have some form of controlled distribution. and distributors should have long term plans in beta. Like redhat 7.0 for example should be in a long term beta.. throw in Xfree86 4.0, kernel 2.4 beta's, the newest kde, the newest gnome, the newest office apps, debug the system as a whole, give endusers/developers something to work with and work from, but most of all, it shows a roadmap of whats to come, and provides ample time for business to ramp up to that product
I love linux, i'm not dissing it, i love open source, its agreat concept. but for business, it needs something i can gurantee my job and and the company can gurantee its data on. not just something i get for free or something i can look at the source at..
I think a lack of focus is inherent in the Linux movement. There are no managers running around telling people what to do. I think this is, competitively, a much better strategy for large base software design because so many people want so many different things....
M$ is very focused, they are a company. I might argue that the individual companies within the linux movement (RH, etc) each have goals and management strategies of the same feel as M$'s. I think that all of the companies together make up this so-called lack of focus. I strongly feel this is a good thing and the reason that other OS's have failed in light of M$ (such as OS/2) is because they were very similar to M$. M$ has a hold on the market, the only thing that will change that (IMO) is a radically new philosophy on software development.
-- Moondog
But, and it's a big but, ignoring Microsoft can't work until they no longer unfairly control the hardware and software markets. If merely being better than Microsoft was sufficient, BeOS would far more successful than it is today. We can and should ignore Microsoft on the day the following can be asserted with truth:
1) Hardware vendors are just as likely to create drivers for Linux as Microsoft.
2) System vendors can't be pressured successfully by Microsoft to avoid using competing products like Linux or Netscape.
3) Microsoft no longer dominates standards thru controlling the OS platform used by nearly everyone.
It would be nice if Microsoft just started playing fair. But I don't expect it. Rather I expect them to lie, cheat, and steal as necessary in an attempt to ensure dominance. Desperation is rarely pretty.
His point about web applications being the future is worth consideration. The advantages in making large databases like Amazon and Yahoo available that way are quite clear. I don't quite see the clarity of that view when it comes to editors, compilers, or games however -- though I could be wrong, I much prefer local programs on my own computer for those.
The other point he makes that is well worth considering is about the open nature of web development. Clay Shirkey did an excellent paper on this subject a while back. You may find many of the other papers on his page of interest as well.
Yes, thank you. Exactly.
Let me see if I can concretize this a little, at least from my point of view.
Right now, if I want to buy books over the Net, I go to Amazon. Now, I dislike Amazon's web site. It's carefully designed to feed me all sorts of data that I don't want. Amazon doesn't want me to buy the single book I decided to buy; if I do that, they've lost. They want me to buy dozens of books, *and* sign up for Amazon mailing lists, *and* move my book conversations from Usenet over to Amazon-managed chat boards, *and* put buy-from-Amazon links on my web pages.
I despise that. (Other people don't, but I'm talking about me right now.) I dislike that for many of the same reasons I dislike using Microsoft software. They're ignoring my goal, and in some cases deliberately making it harder, because their goals are totally different.
So. What if Amazon (or whoever) builds their entire web presence on Linux and Perl, and I use Lynx to shop there? Is that a victory for open-source software?
Yes and no. Yes, because they're using OSS, and that has benefits (stability, low costs, interoperability, choice of browser.) But no, because no matter what software is involved, I'm still having this crappy time buying a book.
This is O'Reilly's point: that the "no" part of that is going to get very important compared to the "yes" part. I agree; that crystallizes a whole bunch of my misgivings about the way the Net is evolving. If the OSS movement is about choice, I want choice about what I do. Five years ago, that was to run applications on a desktop machine. I still do that, but now I also buy books on the Web. Times a-change.
Here's the question, I guess: how can the principles of open software development be applied to, well, whatever the buzzword is for the Next Thing?
Obvious answer: Have Amazon offer shopping data, in a standard interchange format, so people can use it without going through Amazon's idea of a book portal. (That's what O'Reilly had in mind, and why he was talking about it on XML.com, hint hint.)
Obvious followup: Why should Amazon bother? They've got a proprietary lock on what they do, and they're making money on it. (Sound familiar?)
We need to think about what advantages an open approach offers to that -- analogous to the advantages that open-source development offers to software. Then, of course, we have to convince the web sites.
-- Andrew Plotkin (erkyrath@netcom.com)
Summarizing Tim in a single sentence: Chase the dream, not the competition..