|
Sunday, December 30, 2001
Previous topic
|
Next topic
|
|
 Sunday, December 30, 2001
started 12/30/2001; 1:31:41 AM - last post 12/31/2001; 11:30:28 AM
|
|
Doc Searls - Sunday, December 30, 2001 
12/30/2001; 5:31:41 AM (reads: 9660, responses: 18)
|
|
Wrinkage 
| | This time the primary reason is a drop in the number of registered domain names, as the number of domains not renewed exceeds new registrations. Domains bought during the rampant domain-name speculation of late 1999 are now coming up for two-year renewal, and many are being abandoned. For example over the last three months the number of .com domains has dropped by about 130k, though earlier quarters this year still saw increases. |
X Plosions
| | TIG quotes what I said here, and suggests that this means there is "trouble in Linux-land," because Microsoft is supporting OS X for this strategic purpose: |
| | Pushing any chance of Linux on the desktop is an important aim for them. Why? Because, he who dosent have the desktop today wont have the server tomorrow.. |
| | I think Microsoft supports OS X because Office on the Mac is a huge, steady and successful business for them. but maybe TIG has a point. To drive it a bit farther, he adds, |
| | Today it may seem that apple with its propreitary hardware will suffer the same fate. It wont. Microsoft wont allow it to. Microsoft will allow it to remain ghettoed to the PowerPC and Jobs will play. Port MACoSX to Intel and watch MS yank Office and other products. But as long as Jobs stays at 10% with proprietary hardware, Microsoft will back him fully. |
| | I think Steve Jobs likes running on his own hardware because he runs a high-margin hardware company. Moving to Intel would undermine that. And that choice is up to Steve, not Bill. (Does Microsoft still own all that Apple stock, by the way? Seems relevant.) By the way, Motorola has not suddenly failed to obey Moore's law. Looky here. |
| | In a related conversation, TIG also suggests a number of reasons behind what he calls the DotCom/Linux/OSS implosion. I think they're all correct, but not exclusive to Linux/OSS (open source software) dotcoms, which received a relatively small slice of total dotcom investment spending. I would guess Webvan alone received more than all the Linux and OSS companies combined. B2B and B2C schemes received many billions. |
| | Behind all of TIG's reasons is my favorite line of Stewart Brand's: form follows funding. A lot of dumb funding built a lot of dumb businesses that only looked smart in the mirror, which was momentarily given substance by huge IPOs and subsequent stock valuations. |
| | Yes, there were some good and smart business ideas in there, but they stood the best chances of survival, it turned out, by staying away from the whole thing. I know one software & services company that would still be in business if it had simply been satisfied serving its existing (and growing) customer base. But they took the money, and now they're dead. The software is in limbo, being the only asset the VCs can still leverage. And the customers are out in the cold. (This is one case, by the way, where open-sourcing the software would do a world of good for everybody involved. There's certainly no harm in it.) |
Market relativity
| | "You get what you pay for" is over the top. I don't support that and I'll argue with John about that assessment. It's a cheap shot. I know what it feels like to have someone say that about my software, stuff I worked hard on, and didn't charge money for. I don't put any less effort into making it usable when I am giving it away. |
| | I've heard that 'you get what you pay for' argument so many times, and in so many circumstances. And it's so fucking irrelevant when the subject isn't something you sell or pay for. If the YGWYPF principle applied to everything, the best spouses would be high-priced prostitutes and gigolos. |
| | But in settings where you won't make what you can't charge for, YGWYPF tends to apply. Markets, for example. Here's Dave again: |
| | So to use a Doc-ism, there's no market conversation about open source. The developers don't listen to the users (they're famous for that) but even if they did, the users would be loathe to complain because they're not customers and they know it. |
| | The bazaar Eric Raymond writes about in The Cathedral and the Bazaar is a metaphor for a development style. It has been enarged by Eric and many others (including, and perhaps especially, me) to apply generally in an economic sense. In fact, it did much to inspire the Cluetrain observation that markets are not only converstions, but fundamentally bazaars (rather than demogrpahics, categories, bulls, bears, invisible hands and other abstractions). The problem is, the development style Eric writes about in CATB is rarely motivated by the need to sell the code itself. It is produced by the need to use it, and to share it with others who need to use it. Without that kind of motivation, and that kind of code, we wouldn't have the Net, the Web, XML-RPC or SOAP, to name a very few virtuous outcomes of bazaar-stle development. |
| | Among open source developers — guys who write code to "scratch their own itch" — there has always been a corrollary to Skoop Nisker's closing line on his old newscasts: "If you don't like the news, go out and make some of your own." To bazaar-style developers that corrollary is, "If you don't like the code, go out and write some of your own. Or better yet, stop complaining and help us out with this thing." Again, this fuels the development of some terrific software, but the main motivation is to use that software. Not to sell it. |
| | We're still looking for the formula that brings bazaar-style open source development together with the bazaar we call the marketplace. We're also looking for the answer to this one: If code has customers, can its source be open? The answer is clearly "yes" if we're talking about small ticket items like Linux distributions. But for large-ticket items like Caldera's Volution Manager and VA's SourceForge, the answer is "no." I think in the long run that will change, but for now that's what the jury is deciding. That jury is the marketplace. |
| | Some of the best wisdom I've ever read on this matter just appeared on Dave's blog: |
| | Commercial vs open source is like a hot and cold water valve on a faucet. Where you want competition, give away the technology. Where you want to be competitive, keep it to yourself. |
Handbasket weaving
| | Buck, the investment banker, is torn apart by an enraged bear on pay-per-view. No exit strategy? Poor Buck, how you might wish to exercise your options now! |
Snarky new year
| | On the Internet no one knows you¹re a dog. Wait, actually, we all do know that about you, it¹s just that we¹re too polite to mention it. |
Flaps about flops
| | As I mention here, Eric Raymond the other day said open source had "won the technical argument" but not the business one. Context: we were talking about operating systems in general and Apple's open source kernel'd OS X in particular. Not about all forms of software products. Still, it was an interesting statement. |
| | As I've said many times, I think the software industry is turning into something rather like the construction industry from which it borrows much of its vocabulary (architect, design, build, tools, etc.). Construction is a $2 trillion industry worldwide. It's either the biggest business there is, or the second (after drugs). Much of it is DIY — do it yourself. Much of it is the work of small independent contractors and subcontractors. Much of it is artful. Much of it isn't. Much of it is the work of large companies, often employing smaller ones. All of it involves expertise and craft. |
| | There aren't many secrets about how to build a house, a stadium, a dam or anything else. Builders, architects and everybody else share their knowhow freely. There are lots of $billion companies, some with revenues exceeding Microsoft's, but no Microsoft. |
| | When the software industry is mature, it will look a lot like the construction industry. But we're a long way from there. The overfunding of fantasies has made the journey a lot longer, and not just for the software industry. |
| | There were fantasies about the Net as a new form of every medium you can name: TV, radio, film, publishing. There were fantasies about a whole new advertising business, and about all kinds of things that would be funded by advertising. There were fantasies about retailing. About online markets. About "push" and "portals" and many other forms of nonsense. About the "new" economy, which was never anything more than a scheme for scoring big by selling specious futures to the largest possible pool of gullible investors. |
| | There were also fantasies about Linux and open source. Investors saw them as the natural successors to Windows and everything Microsoft stood for. |
| | There were tragedies everywhere. AtHome was a brilliant idea, mostly well-executed, that delivered on the broadband promise at a remarkably low cost. I'm benefitting from it right now, paying about $50/month for performance that shames the T-1 at my old office. But instead of making money the old fashioned way, AtHome went and mooshed itself into Excite, buying into the fantasy that advertising was going to pay for everything. Stupidly, ExciteAtHome sold its fundamental service essentially below cost to the cable companies and eventually died an awful death when the advertising fantasy didn't pan out. |
| | As Dave has often pointed out, the business promises of open source and Linux sold excessively well to the VCs and then to Wall Street. One effect of that was to undermine investor faith in all commercial software developers other than Microsoft. As a class, independent software developers were shunned and struggled to survive. Some didn't. While the stock values of Red Hat and VA Linux collapsed, Microsoft toughed it out. Now Linux and open source are getting the shun treatment. Sure as night follows day, bust follows boom. |
| | But take away all the stupid investing, all the politics, all the adversarial talk, all the distrust and blame, what do you have with open source? |
| | I think you have a building method — a way to build stuff. A choice you can make. Yes, it has technical advantages. In some cases it may have business advantages as well, but those are highly conditional, and far from being fully tested. |
| | But there's a third variable, which is market demand. Craig Burton says the challenge is to drive global ubiquity while fostering shareholder value. Open source is good for the former. About the latter, it depends. We won't really know until we're way out of the current bust phase. Meanwhile we do know about closed source. It's a fundamental means for securing value of software built to sell. Which is a big reason why companies from the open source community such as Caldera and VA Software (no longer VA Linux) are selling closed source software products. |
| | My vote for Flop of the Year is the whole investment bubble. Every company has two markets: one for its goods and services and one for itself. In the dot-com bubble, interest in the latter completely overcame the former. Rather than selling their goods and services, companies sold themselves, over ane over, in round after round, to investors. Markets for many kinds of goods and services were seriously hurt. Software especially. |
| | I'd like to say it won't happen again, but I'm sure it will. |
discuss
|
|
Dave Liebreich - Re: Sunday, December 30, 2001
12/30/2001; 5:04:19 PM (reads: 628, responses: 2)
|
|
|
Software industry as construction industry - I like the analogy. But I'm wondering how software testing and qa would fit in to it (since that's what I do fo a living). I've got some ideas, but most of them are based on breaking the analogy.
(things like the fact that software projects are usually many orders of magnitude more complex than construction projects, so simple inspection does not really work. Also, the internal concept of liability is different between a builder and a programmer - but that could change)
-Dave
discuss
|
|
rusty - Doc, you maniac
12/30/2001; 5:54:47 PM (reads: 647, responses: 6)
|
|
|
When the software industry is mature, it will look a lot like the construction industry.
Misled by the language, man! The software industry might use the same words as construction, but it will never, ever look like it as a business. Here's why:
Every building is a unique object. They may be mostly put together from lots of the same basic components, but they are all discrete, unique objects. You can't build one high-rise and then copy it off 200 million times. You have to build each one separately.
Software, on the other hand, is a "build-once, sell often" business. The process of building one piece of software is similar to building one house, but when you're done building it once, you can go out and sell it to everybody, without ever having to build it again.
This is why a Microsoft can exist in software and not in construction. No construction company, no matter how big, can specialize in everything and take on every construction contract at once. Microsoft, on the other hand, can build one of each kind of software and then sell it to everyone. In construction terms, it's like if a company could build "House 2000" and "Skyscraper 6.0" and sell identical copies off instantly to anyone who needed a house or a skyscraper.
If you want a real-world analogy, try the cola industry. Coca-Cola comes up with a formula for Coke, and one for Diet Coke, and one for Sprite, and so on. Then they license the formula and some branding symbols to bottlers all over the world, who create trillions of cans of the exact same thing, and sell them to everyone. One product for each "application", and each copy of the product is identical. The only difference is the "bottling" process for software is so cheap that MS doesn't even need to license someone else to do it.
discuss
|
|
Dave Winer - Re: Doc, you maniac
12/30/2001; 6:01:36 PM (reads: 732, responses: 2)
|
|
|
Rusty you might be surprised. Software is always heading in new directions, getting closer to the work process, esp in engineering and marketing, and when it does, companies differentiate themselves based on how work flows and their software systems need lots of customization. At some level we suck more and more of what used to be experimental into the operating system (we all do it, not just the recognized OS vendors) but there's always a frontier where the users do the customization and R&D for the suckers. If you view software at a single moment in time you miss this effect. There's always another upheaval waiting to happen, as developers figure out how to commercialize what the users have been doing at a one-off level. Or when users get tired of waiting for the gorillas to move and do it themselves.
discuss
|
|
Dave Liebreich - Re: Doc, you maniac
12/30/2001; 6:08:54 PM (reads: 681, responses: 2)
|
|
|
Rusty,
You are breaking the analogy by concentrating on the distribution end of the "business" rather than the front-end "making stuff" part.
Which is fine - stuff is meant to be broken (after all, that's what I do for a living :-)
-Dave
discuss
|
|
rusty - But, will each product be unique?
12/30/2001; 6:46:09 PM (reads: 811, responses: 1)
|
|
|
Ok, my analogy was a little far toward the other end, perhaps. But I think the basic point is still valid. No matter what you can pre-fab, nearly every building must be finally assembled on-site. Yes, there are delivery houses, but they're such a small part of "construction" that I'm not really counting them. The point is, construction, by it's nature, requires individual attention to each and every end product.
Software, on the other hand, doesn't. I know what you mean by software systems being more and more customized. I wrote contract code for a long time. Most programmers never work on "consumer product" code. But the nature of software does not demand that. In fact, software, by nature, tends to the general. Every application needs some customization for each copy, usually called configuration. When the customization ges too complex to be done by the end-user, something usually happens to the software. Either it becomes more than one product, partly pre-configured for a particular use, or someone fixes the configuration so it can be self-serve by the end-user again. I think the case of the software industry as lots of people writing one-off code for one client is a huge waste of people and effort, and is a degenerate result of software companies sucking. Probably, we can trace the present situation straight back to Microsoft, who have made it nearly impossible for the kind of small-scale software shop that makes one or two products for a niche market to exist. So instead, we get 1,000 contract hackers producing the same basic thing for each client in that niche.
But that's just how it seems to be right now. I'm not really trying to look at the software business at a particular moment, as much as the essential natures of software and construction. I think irreproduceability and tangibility is the heart of construction. That's why it's a craft, and that's why it's practitioners share their knowledge so freely. Techniques get you to the end of the job, but they aren't the job itself. I can tell you how I made something without giving you the thing.
In software, on the other hand, the code is both a description of the technique, and the actual product itself. Software, no matter how customizable, is inherently reproduceable. In both businesses there are boundary cases that overlap (prefab houses, NASA software), but in broad terms, the two businesses just aren't the same.
By the way, I have worked in both industries. I could be wrong, but I'm not just making this up. :-)
discuss
|
|
rusty - Business comparison
12/30/2001; 6:49:24 PM (reads: 759, responses: 0)
|
|
|
Hmm. I thought I was pointing out how the analogy was broken because Doc was concentrating too much on the front-end "making stuff" part. ;-)
I do agree that the activities of building software and building "stuff" are very similar. Extremely similar. No doubt about it. But, I was trying to point out why an analogy that predicts that the business of software will come to look like the business of construction is mortally flawed. The distribution ends are too different.
discuss
|
|
Doc Searls - Re: Doc, you maniac
12/30/2001; 10:53:04 PM (reads: 758, responses: 0)
|
|
|
I'm talking analogies and metaphors here. Not identicalities.
The center of gravity in the construction business is in services, not manufacture and distribution. Early in its history (where we are now), the center of gravity in the software business is in manufacture and perhaps distribution as well. I think this will change as *both* get bigger.
To Rusty: I see what you mean about Coke, but Coke is the Microsoft of its business. Not the analogy I'd like. More in another post...
discuss
|
|
Doc Searls - Re: But, will each product be unique?
12/30/2001; 11:00:39 PM (reads: 894, responses: 0)
|
|
This is an interesting point:
Probably, we can trace the present situation straight back to Microsoft, who have made it nearly impossible for the kind of small-scale software shop that makes one or two products for a niche market to exist. So instead, we get 1,000 contract hackers producing the same basic thing for each client in that niche.
I wonder if there are any studies on that...
As for the reproduceability issue, you're right. Construction deals in tangibles, and that's a huge difference. The commercial software business seeks to reduce that difference by selling binaries, and by key-lock systems like Microsoft is building into XP.
But I think the construction metaphor is still useful one, if only to show a *similar* business in a mature state. The fact that construction has plenty of room for all kinds of manufacture (prefab and "open source"), distribution, service, etc., and is an $X trillion business worldwide, is a helpful example for software, which is still a very young business.
discuss
|
|
Patman - Re: Snarky New Year
12/31/2001; 12:52:44 AM (reads: 598, responses: 0)
|
|
|
|
jt - Re: Flaps about flops
12/31/2001; 5:20:28 AM (reads: 662, responses: 4)
|
|
|
Mr. Searls,
Thank you for this post and the one on Market Relativity. It sure seems rare to hear folks tell it like it is. Thought your analogy of the software and contruction industries was very interesting, as well as the discussion that followed. Have to think about that some more.
What I REALLY liked was how you pointed out the fantasies that have driven the OSS movement. And the tragedies that ensued.
~~~
But you stumped me when you said "But take away all the stupid investing, all the politics, all the adversarial talk, all the distrust and blame, what do you have with open source?"
IMNSHO, that was the biggest fantasy of all.
You can't take away all the stupid investing, politics, and adversarial talk.. the distrust and blame.
That's what all organizations become, once they reach a certain size. You aren't going to defeat human nature with the OSS movement... Sorry.
The OSS movement worked well on a very small scale. But the success of OSS soon allowed it (forced it) to grow past the size where it worked well. The theory was that, if OSS was based on altruism, the old business-as-usual wouldn't apply...
That fantasy has clearly been proven false.
~~~
I s'pose I should say "IMHO", (although I don't think it's a matter of opinion at all), but the idea that OSS provides "a building method — a way to build stuff." Well.. that's patently false, as well.
As Mr. Winer basically said, what the OSS movement has accomplished is taking the customer out of the equation. That's the point I take from your Market relativity post: the OSS movement works well when the "customers" are the guys who both use and code the stuff...
Point being, that's a very limited market... Such an extremely small segment of the total market, that it's currently the choke-point of the OSS movement. (That isn't a trend.. that's what has happened...)
jt
SiliCow Valley
(Columbus, in the dim state of Ohio...;-)
discuss
|
|
Doc Searls - Re: Flaps about flops
12/31/2001; 5:48:46 AM (reads: 800, responses: 3)
|
|
|
Thanks.
I think the problem OSS had from the beginning wasn't that it took the customer out of the equation, but that it had a hard time bringing the customer into the equation. This was mostly a cultural matter, and a legacy more of the free software movement than of the open source movement (think of two overlapping circles in a Venn diagram, the latter overlapping the commercial marketplace far more than the former). With Jabber, for example, there was an early effort made to isolate the .org open source organization from the .com commercial organization (Jabber, Inc.). I'm not exactly sure why that was done, but I suppose it was to keep the open source (public, nonprofit, etc.) side free from the contaminating influences of real business, with real customers. This was a big mistake that both sides have been working to correct ever since.
discuss
|
|
jt - Re: Flaps about flops
12/31/2001; 6:20:14 AM (reads: 881, responses: 2)
|
|
|
Mr. Searls,
I normally read up on people before I write them. Sorry I didn't in this case. MAN...! Probably best I didn't, or I might not have written... That's my way of saying you honor me with a reply. (Do you do this with everyone and/or did you somehow hear of me?!?)
I've been meaning to read the Cluetrain Manifesto, and didn't recall who the authors were... I've been told a few times I'm clueless.. but I wanted to see for m'self, and get it straight from the source...;-)
I'm a newbie on a lot of this. Just read a little of RMS and ESR, and seen some (what I consider to be) flaws in the logic. Didn't know the history of Jabber, but it makes a lot of sense.
It appears that the legacy (or horror, depending on POV) of the FSF is that they didn't understand that you don't eliminate the role of the customer, even if you give the product away for free. The developers of all the OSS were the customers, IMV (in my view of the world).
If they didn't start with that understanding, then they won't understand that different customers have different needs from the products.
I'll have to think on that some more (as it's 2:15 here, and way past my bedtime).
I'll be curious to see how Jabber makes out in the future. 'Cause IMV, two organizations would be the ideal... But rather than worry about "contamination", I'd view cross-pollination as a very beneficial thing.
See.. that's my theory, but I don't believe in much of anything that hasn't been thoroughly tested (other than my code, that is...;-). My view is that the market is THE best place to test these kinds of theories.
Beats the crap out of both computer-simulations and people boasting what could happen.
Again.. thanks for the reply, Mr. Searls,
jt
discuss
|
|
Doc Searls - Re: Flaps about flops
12/31/2001; 6:31:11 AM (reads: 909, responses: 1)
|
|
|
I'm honored that your honored, although I'm not sure what makes me, like, huge or something. But, whatever. It's cool.
Anyway, to set things a bit straighter, Jabber is definitely back in a converged state. Lots of cross-polination. But again, it's still early. Even if it's late for you (I'm heading for bed myself).
ds
discuss
|
|
jt - Re: Flaps about flops
12/31/2001; 6:50:55 AM (reads: 974, responses: 0)
|
|
|
ds
(I lied, I'm off after this.)
Yeah... After reading About Doc, you're definitely huge...!
I'll look into Jabber, to see if converged means one organization.
jt
discuss
|
|
Michel Benevento - You lay what you pay for.
12/31/2001; 1:27:23 PM (reads: 607, responses: 0)
|
|
|
If the YGWYPF principle applied to everything, the best spouses would be high-priced prostitutes and gigolos.
No, wives and whores are not in the same market.
Expensive prostitutes are nicer to have sex with than cheap ones.
I hear.
happy new year!
discuss
|
|
Doc Searls - Re: Sunday, December 30, 2001
12/31/2001; 2:01:40 PM (reads: 726, responses: 1)
|
|
|
Well, like I said in response to another post, metaphors, similes and analogies work because they are not identicalities. It seems significant to me that software borrows some of the languguage of constrution. It also seems significant that software also borrows the language of other domains. For example, we "write" software as well as "build" it. It has "syntax" and "language."
We mix metaphors when try to protect software as both a form of property and a form of speech. The cognitive linguistics can get pretty complicated.
On the qa issue, by the way, I think there are plenty of similarities between a bug list and a punch list. Differences too.
discuss
|
|
Dave Liebreich - Re: Sunday, December 30, 2001
12/31/2001; 3:20:30 PM (reads: 843, responses: 0)
|
|
|
I'm the kind of person who looks at something and immediately starts thinking about how it doesn't work, or how I can break it. That's what I enjoy doing, so that's where I usually go first.
Maybe the construction metaphor works for any structured creative effort? We build teams, organizations, support, ...
The end goal, I think, is to frame the discussion - in this case, how is programming like or unlike construction - so that we can discover more about programming. It's a technique to facilitate understanding, or to teach.
So I think the analogy of the business structure of construction works better in the IT programming world, but not as well in the COTS or embedded-system world. Likewise, the analogy of the construction process.
And I'm not any closer to figuring out how the construction analogy fits with the process of testing . . .
Thanks for the discussion.
-Dave
discuss
|
|
Bernie Dunham - Re: Sunday, December 30, 2001
12/31/2001; 3:30:28 PM (reads: 694, responses: 0)
|
|
|
If construction is going to be the metaphor for open source, then perhaps the people writing the code are more like people associated with Habitat for Humanity (I'm on the local Board of Directors, so I'm a volunteer). Some are volunteers, some make donations, some sell products or services (sometimes even at a discount). The intention is always to do something useful and well intentioned to improve the quality of life, even if it is only one family at a time.
The Object Oriented seminars, such as those for SmallTalk, liked to describe the process a few years ago as reuseable components, using a manufacturing methaphor, rather than a construction metaphor, whereby programming was like working on an assembly line. Hopefully, open source and "free" software will not become, or be percieved as, the manufactured housing of the software industry: cheap and low quality double-wides on wheels, following minimum standards.
discuss
|
|
|
Copyright 2009 The Doc Searls Weblog
|
