So development becomes an end in itself, not a means to an end.
There are historical counter-examples which prove the point. When Windows 95 it was a disaster from the point of stability, but it still was stable and standard enough to create the proverbial level playing for developers.
This meant there was an explosion of commercial user oriented products developed by back-bedroom one man band start-ups. This is not a small thing - it's part of what persuaded ordinary people that PCs were worth buying.
That's also how Photoshop and some of the other commercial apps started - from an earlier wave of development which happened once the Mac environment stabilised.
After ten years there has no been no equivalent development in the OS world at the user level. It's happened to some extent at the commercial level, but not on the desktop.
Because OS is a developer free for all, there's no stable environment. There's no guarantee that any piece of software will work with any particular Linux variant. And even if you do your own compilation, you can waste hours in development hell trying to track down dependencies and essential libraries.
So OS doesn't work in user space for two reasons - it's not actually open in any practical sense, and it's primarily designed by developers for developers as a developer exercise, not as a product which end users can work with. This means stability and consistency don't happen, and there's no real interest in keeping non-geek users on board.
I don't think the OS people understand how essential this is for development. Most people would rather pay a credit card fee for a product they know they can use than try to fight their way through a user-hostile developer environment like Source Forge to download it for nothing.
No that's very much the point, because what happens is that the experts tweak it for their own benefit and entertainment.
Unlike a physical craft, where the typical piece of hand made furniture is not much different from the piece of hand made furniture turned out by the typical home workshop hobbyist ... in Open Source programming, all of the most widely used open source programs are very much atypical ... because there is a very strong power law distribution, and the typical program produced is a far cry from the programs most commonly used.
Sometimes they're even condescending and dismissive of users who aren't part of the community of experts who understand how to compile and edit the software for their own use.
For the desktop ready applications described in this diary, this matters no more than the attitude that Microsoft programmers have at work ... the Open Source community of the most commonly used applications is quite different from this, which is a typical Open Source community, and includes a substantial number of contributers that specifically contribute to ease of use. I've been accused of being a Marxist, yet while Harpo's my favourite, it's Groucho I'm always quoting. Odd, that.
There's no peer to peer network at present that can on its own cope with the usability testing demands of the Open Source toolkits at the level of complexity of OpenOffice or Chrome. On the other hand, there is Sun and Google, looking to leverage Open Source contributions by making those toolkits Open Source. A need for corporate contributions to user testing of large scale applications would be a hurdle if there were not the corporations with an incentive to make those contributions. I've been accused of being a Marxist, yet while Harpo's my favourite, it's Groucho I'm always quoting. Odd, that.
No that's very much the point
Aha! Well, I'll be obstinate and say that it's not. You'll notice I said (in principle), and I stand by that. What you are describing is what you claim (rightly or wrongly) to be practical deviations from the principle: "what happens".
One could say as much of a great deal of commercially locked-down software: the principle is that the profit motive leads to the creation of an efficient team of highly-competent developers who must respond to market demand, but "what happens" is that Gates & Co take ordinary end users for cash cows, condescend to them, do not care about their needs, bring out fake "updates", etc.
Your point about the development of a range of specialised applications is taken. But most ordinary end users don't need them. A stable, secure OS, browser, e-mail client, and office apps is enough for most people. Should they pay for (Ford help us!) Vista? Or M$ Office, on which most of the development work in years has gone into chaining users to an inescapable standard?
They of course do not put a lot of effort into the dribs and drabs of end user cash flows, but for the main end user cash flow, the license payments for software installed on a new machine, they are quite vigilant. I've been accused of being a Marxist, yet while Harpo's my favourite, it's Groucho I'm always quoting. Odd, that.
- Jake If you only spend 20 minutes of the rest of your life on economics, go spend them here.
And on new PCs (especially for the mass market), they can and do.
If Linux really made inroads into the end user market, Microsoft would not balk one second as making its OS as free as it was in the 90's... Un roi sans divertissement est un homme plein de misères
Yes, if you buy your PC in the supermarket, it'll come with a paid copy of Windows
That's what I said. This is a discussion here between people who wouldn't buy a Dell or a supermarket PC, but that's no reason to ignore the mass market of those who do.
One could say that, but it would a trite caricature of how the market has developed.
One obvious problem with OS discussions is that they inevitably start from an evangelical and tribal position - i.e. Microsoft is bad, OS is good, and that's all there is to know.
There isn't a great deal room for nuance in an approach like that, and certainly once you start arguing that what happens in practice is less important than principle you may be running headlong into one or two minor consistency issues.
But still - let's try again.
What happened in practice in PC space - as I said - is that even though Microsoft brought out a string of shabby operating system products, they still created enough stability that developers were able to work in that space successfully.
These developers were not big or corporate. In fact the opposite happened - MS ate the corporate developers or forced them out of the user area. The interesting action happened lower down, in areas like music, graphics, photography, and utilities, which MS wasn't much interested in.
Almost exclusively, viable products for those spaces were produced by small developers working with tiny budgets. Because developers had to sell their products, they put a lot of effort into meeting actual user requirements. And people really do use them. Almost everyone takes photos now and does some basic editing, almost everyone has a music collection, more and more people are making videos or doing basic web editing.
Aside from the fact that MS's need to create a monopoly created this space, this has nothing to do with MS itself.
The point is that standardisation, even on a crappy standard, and an interest in meeting real user needs both turned out to be a very good thing.
It's a shame it couldn't have been standardisation on a much more robust standard, but in the end it almost didn't matter - developers made products that people wanted, and peopled paid for them.
That hasn't happened with open source. MS has certainly tried to lock down Office, but this isn't just about Office - it's about the fact that when you buy a Windows PC, you can buy or download a truckload of software for it, and most of that software will meet a direct need after a very simple installation process.
OS actually offers little or no benefit to users who want their computers to work like that.
It's the Monster Truck racing of software design - just because you can fit giant wheels to your SUV, doesn't mean that most people are going to want to.
For that minority of people who enjoy assembling things from components, it's fine. For people who want something that Just Works in a standardised way, the Windows experience isn't perfect, but it still saves so much time and overall effort that the 'free' part of open source becomes irrelevant.
The corporate experience is something different, and there's more to be said for creating clean installations which meet limited but specific needs. If I ran a large business I'd certainly look at OS seriously as an alternative to Windows.
But what OS people don't seem to get is that 'Use it! It's Open Source!' is a non-argument. While there are a lot of sheep in corporate IT, anyone who does IT properly is going to be breaking down a big installation into support and training, legacy data access, compatibility and interoperability, workflow, stability, and so on.
Access to source code comes low down on that list. If you don't have in-house developers, it's useless to you.
If Open Source wanted to be more successful it would spend more time touting an integrated approach. Some OS companies do it, but from the outside it can look like 'It's Open Source! You can change it!' is as far as most developers seem to go.
And if open source isn't more popular yet - that's why. It's just a different bunch of people who aren't any more interested in listening to ordinary users than MS is.
You know, I didn't start this one. Every time I've seen you write about this, it has been to say OS is bad, and sometimes in stronger terms than here (iirc). Now, I'm not saying OS is all good (I didn't say that what happens in practice was less important than the principle, I simply made clear the difference between the two), and do read me again when I said that I took your point about specialised software. M$ is bad? After twenty-five years of "ordinary user" acquaintance with them, I'm not satisfied with their products. Your point about the stable standard permitting development is Microsoft's own; it's possible to look at it in another way as, once there's a single standard installed on the vast majority of the world's machines, there's not much choice for developers but to work with it. And I can remember the complaints of said developers at different times in the past about the difficulty of working with M$'s standards (though I'm not up to date on that).
A lot of what you say may be true (a good deal I'm no judge of), but I did speak of ordinary users' desktop needs, and I get the feeling your notion of what that means is coloured by your own needs and experience.
I was absolutely in favour of MS being dismembered ten years ago because their monopoly wasn't a win for anyone. But that doesn't mean I don't think in hindsight that there wasn't a historical user benefit in a de facto standard, even if it was a broken one.
It's not a black and white issue. If you happen to be interested in the history of computing, not many issues are black and white.
Historically the point stands - whether or not developers enjoyed the experience, the fact is that over the last decade there's been far more useful desktop-level software of higher quality produced for Windows than for Linux.
There are complex reasons for this, and they go a little beyond 'Yay Microsoft!' But apparently it's not possible to make this point without being shouted down by angry people.
Fair enough.
But that's very much the negative user experience I was talking about. 'You're either with us or against us!' isn't usually a position that makes me feel positive about a project.
So this exchange and everyone's comments have proved the point - OS culture seems as sealed to outsiders as large corporations are. You're either in it, as a card carrying evangelist and paid up member, or you're outside it, as one of the evil hordes who want to see MS stamping its jackboot on the face of users forever.
At worst you're 'Just a user' who can't even use a compiler or file a bug report properly.
I don't know about anyone else, but that kind of polarisation doesn't seem like a good thing to me.
What happened to 'On the one hand... on the other hand'?
I don't belong to any tribe, I'm just a basic user who doesn't need the specialist stuff you're familiar with. I use both OS and M$ software. OS may well have failed to take sufficient market share. I find that regrettable, because I'm not happy with M$.
So You're either in it, as a card carrying evangelist and paid up member, or you're outside it, as one of the evil hordes who want to see MS stamping its jackboot on the face of users forever feels to me, you know, irrelevant and, frankly, way overstated.
You have some points but you are also missing some stuff. Let me just point two of them:
When an open source application is released it is usually done so in an immature state (version 0.x). And that's when the testing starts. This is a big difference between open and commercial software: while the later has to be released after testing by a selected group of individuals (one more additional cost), the former is released untested and the testing team is the whole world. That's why things like BugZilla exist, an interface between testing and developing teams. This is something I feel you are missing: the power of Community in the open software process. Beyond having a much larger testing team, there's a further advantage of the communal development/testing process: debugging and improving goes on even after the application has reached maturity. This last characteristic creates a dynamic between user and developer that commercial vendors can't replicate; in my own experience this at the moment is a major advantage of open GIS over commercial GIS: user queries are dealt with much faster.
The word Community leads us to the ideological background that was behind the genesis of Open Source with Richard Stallman's GNU Manifesto. It is no secret that I and most folk here at ET identify ourselves with this kind of e-Socialism. But commercial software is posing problems today that are beyond any ideological stance. Several companies have acquired monopolistic market positions of a sort possibly never seen before that create user dependency and manage to guarantee it prevails through time. Instead of fighting them (even for the sake of Free Market if for nothing else) Governments have instead joined the lot of dependent users. That's why Public Administration are having fora like that I attended: they understand they are compromising the State by willingly becoming hostages of foreign private corporations. In the long run no ideology will stand that.
But of course, if you're happy with the commercial products you use, let it be, this is a Free Europe after all. Vencit omnia veritas.
I've also promoted the concept of transparency to corporate clients, with some success. Business transparency (part of what I would call 'reputation management' which is replacing 'branding') is related to OS.
However, like TBG, and as stated above, for my own work, I cannot find OS software that does what my proprietary software does, as seamlessly, efficiently, user-friendly and so well integrated with other programs, and makes to possible to exchange projects with collaborators. You can't be me, I'm taken
And that's when the testing starts.
No, this is only how OS people think the process works. What happens in the real world is more complicated.
Generally bugs are only fixed sometimes, because fixing bugs is tedious and uninteresting, and a lot less fun than adding Cool New Stuff. So in practice most desktop projects and some of the core Linux variants end up in a kind of homeostatic limbo, caught between half-hearted bug fixing and the constant developer need to hobby-code.
Beyond having a much larger testing team, there's a further advantage of the communal development/testing process: debugging and improving goes on even after the application has reached maturity.
Except that on the desktop, very few projects ever do reach maturity. There is a lot of abandonware where developers simply get bored with maintenance, and a lot of sort-of-not-quite-there coding which works in a limited way.
I'm on a couple of developer lists so I'm not unaware of how the process works.
The word Community leads us to the ideological background that was behind the genesis of Open Source with Richard Stallman's GNU Manifesto.
Stallman doesn't actually support Open Source - like all good authoritarians he wants other people's freedom to be on his own (virally GPL'd) terms.
It is no secret that I and most folk here at ET identify ourselves with this kind of e-Socialism.
Open Source is not e-Socialism - except in a very limited and dilettante-ish sense.
OS is more like mechanics getting together and telling everyone that because it's possible for anyone to build their own car everyone should, and hey, did we mention that we're not Chrysler and GM but look at our cool car designs anyway because they're at least as good, really they are.
Mechanics can be good and useful, but not everyone needs to be a mechanic.
OS may be anarchist, in a limited way, but it's certainly not socialist - I don't think anyone who understands what socialism is would make that claim.
And in fact, not a few developers seem to be libertarian capitalist types.
I'd be more convinced by this kind of e-socialism if the OS community as a whole spent more time teaching computing skills in low income areas or organising charity PC donations - practical things to help people from other economic demographics.
I'm sure some developers do that, but it doesn't quite seem to be a key feature of the open source project.
I appreciate your comments but I think you are taking this debate to much to the idealogical side.
My own experience with debugging of OS software is in total contrast to your simplistic characterization. I had interaction with folks developing Squirrel, Eclipse and Firebird and it is way beyond anything commercial vendors provide. For instance, with ESRI help comes mostly from other users and rarely from the developing teams.
Stallman an authoritarian ... what's authoritative in developing software for free? And an authoritarian an anarchist? I don't see how do you square Anarchy with Communitarian Development in a process where actor roles are clearly stratified.
But then again, to me this is not at its heart an ideological issue. Vencit omnia veritas.
And how annoying is it not, that every time there is a discussion on GMOs or financial banking it always degenerates into a tribal fight over the nastiness of Monsanto and Goldman-Sachs. Why can't people see that the important thing is not such trite considerations as power concentration?
As long as the consumers get what they want, it is all good. Food and credit, and in nice packages too.
ThatBritGuy:
Yes, windows users never get errors that they do not understand which hinders the simple processes. That would be unheard of. And of course there are no programs for Linux users. It is not even commonly distributed with the OS or distributed by way of the internet. And Monsanto products just taste better. A vote for PES is a vote for EPP! A vote for EPP is a vote for PES! Support the coalition, vote EPP-PES in 2009!
So it's not that standards don't exist outside MS products. It's just that MS - by virtue of being a (near) monopoly - is able to enforce standards, and for various reasons governments have been unwilling or unable to enforce the original public standards (standards that in many cases made more sense than MS' ones do).
