This article intends to clear up some misconceptions about open source software development practices. It can help developers, IT and business managers transition from a closed development environment to an open one characterized by shorter time-to-market and lower costs. The author, Tom Adelstein — an experienced CPA, code developer, project manager and consultant — makes clear the notion that Open Source Software bears a mark of professionalism. Full Story
How to Misunderstand Open Source
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Excellent read.
That article is dynamite – I’ve rarely read such a thorough, methodical rebuttal of anti-OSS FUD. I think I’ll send the link to Groklaw…
The token analogy is interesting.. But I think of Open Source (note the capitals indicating free as in RMS/LINUX) as more akin to a gift certificate. You’re free to buy whatever you want, but only in that store. And with open source, you’re free to have the source, but if you want to make any practical use (redistribute), you can only do it if you license your resulting code under the Free License. It’s free, with restrictions.
And with open source, you’re free to have the source, but if you want to make any practical use (redistribute), you can only do it if you license your resulting code under the Free License. It’s free, with restrictions.
It’s free as long as you keep it free.
Note that there are plenty of practical uses for open-source code that don’t involve redistributing the code (for example: using the code for in-house projects, or as a prototype, or as a tool to generate useful code/data for your own project).
Gee, could that be the description I was referring to? Just like my analogy. You can buy whatever you want for free, but only at that store. Should I perhaps explain it slower for you?
“It’s free, with restrictions.”
Yes, yes, it is free with restrictions. And?
Where the specific analogy is that Open Source is money, and closed source is a subway token. I as describing that, in comparison with other open source licenses, the GPL is more like the subway token than actual money. Perhaps my detailed explanation will be clear enough for you to follow this simple point.
And as for the Tokens guy next time RTFM
RTFM? Hah, that is another topic. I don’t have time to get into the condescending, patronizing attitude often projected by supported of Open Source. One issue at a time.
(off to search for a token manual)
I’m reasonably certain that this comment will be poorly received here, but I just can’t keep it to myself, so oh well.
Truly, the author does a good job of dispelling one piece of FUD kicking around regarding open source software, specifically the belief that most open source software is written by wild-eyed loners without any concept of planning or design.
Other than that, the article is, umm, not so good. Nearly everything he says about closed source processes describes “big company” closed source processes. I work for a small closed source shop and his description of the open source development process is very *very* close to our process. It sounds as if his only closed source experience was with IBM, which is quite possibly the most extreme example of a process-bound company one could imagine.
Since analogies and similes are so popular on this thread, I would suggest that he what he is saying is like saying that vehicles made in Japan are more responsive than those made in America, using as examples a Japanese sports car and an American diesel locomotive.
Anybody who has worked in or with a smaller, more nimble closed source shop will see his description of the “closed source process” as bullshit. Many of these people will conclude that the author is a crank and proceed to ignore the good point he does make about the professionalism of many open source projects and companies.
The same applies to the stuff about standards. Closed source shops can and do adhere to open standards; I know we make every effort to do so in my shop. Many many open standards were originally developed by closed source shops cooperating to facilitate communication between closed source products and to offer the market choices in how to combine them. I know this is hard for some open source zealots to believe, but many closed source shops know that offering products based on open standards can help improve adoption of new technologies; when the tide comes in, all boats rise, ours and the competition’s both.
Don’t get me wrong, I think open source software is a Good Thing; I use it daily (Mozilla, OpenBSD, Knoppix, blah blah blah) and push it whenever I get the chance. I just don’t think this article is about what the author says it’s about and I don’t think it will appear convincing to anybody that isn’t already convinced.
PS – It also doesn’t help that in an article about professionalism in the open source world he flubs “stock in trade” and “give way” in the opening sections.
Funny that this is the icon for the category is the GNU
logo, since RMS and other GNU fellas makes the point that
they do not develop _open-source software_, but instead
_free_ software. RMS, for one, do not like the concept
of open source.
I think most people by now, even Microsoft in one way or the other can accept open source projects. The problem is really RMS and the likes.
Sun has Open source projects, a whole bunch of them… but what is truly important is that the license chosen for development is a license which favours the project. For instance, any commercial company who relies on GPL should think again… I believe that pretty much all other licenses are better in one way or the other.
For instance, SUN have things open sourced as long as you follow certain guidelines, now this is what it’s all about. Keep source open and make sure the license is chosen for the purpose of the software… I got a really hard time seeing that would be GPL…
If GPL would disappear I’m confident we’d see the term Open Source not be related all the time to something negative… but I guess zealotry will keep Open Source looking like crap still for quite some time =(
Very sad but I’m afraid it’s true…
but if you want to make any practical use (redistribute)
There are plenty of practical uses for OSS that does not require redistribution. Using Samba for a LAN, Apache to serve Web Pages, using OSS in-house for production purposes, or administration, or whatever. In fact, I’d say that most uses in the real world (i.e. not a developer’s world, which is a subset of the real world) do not involve redistribution.
I myself use OSS a lot, and yet I’ve only rarely redistributed it.
I don’t have time to get into the condescending, patronizing attitude often projected by supported of Open Source.
How about the arrogant, aggressive attitude often projected by supporters of Closed Source?
And I think he meant RTFA, as in Read the Friendly Article…
Freedom without responsibility is an oxymoron. Frequently “free” is connotated as meaning unlimited and/or unbound, in uses like “free”=”unlimited usage”. Yet the world of experience teaches us that we are only free in the sense that we MUST take responsibility for the choices and decisions we make-and that we HAVE TO deal with the consequences of these choices and decisions. Being free, means being bound, bound in a relationship to that which secures and maintains this freedom.
Being free entails obligations, failure to fulfill such obligations is tantamount to being bereft of said freedom. For software developers the concept is trully simple: if a developer did not have access to the tools he/she uses to create software, his/her software would never get made; If the tools are NOT free the developer cannot learn from the methods used in the tools themselves and is thus dependent on those who develop the tools to provide him/her with the features the developer needs, this model is reminescent of indebtured servitude; if the tols are free the developer can learn from the methods used in the tools, thus expanding his or her toolset to be a superset of said tools, this model entails the responsibility of revealing all methods used,ie. the code of all the tools being used, because failure to do so would result in methods which, due to obscurity (patents, copyrights, etc.), cannot become part of the expanded domain of avalaible methods. Free software development continually expands the available methods at disposal for development use by developers. Non-free software maintains a hegemonial control over the domain of methods at the disposal of developers, and developers are enslaved to the decisions of those who make the tools which they use.
Non-free software development means you as a developer belong to the company for which you develop. All of your ideas, methods, techniques and knowledge learned through experience potentially belong to the company. If you write stuff at home for yourself, it potentially is something which belongs to the company. I use the world potentially, because there are certain, primarily smaller, non-free software companies and self-employed programmers who are not caught in the legal morass which accompanies the IP restricitions and policies which are mandatory when one works in large non-free software companies. The vast majority of programmers who are employed are employed by large software companies-They are subject to laws and restricitions which transgress basic values of autonomy.
Self-employed software developers are usually against free software, for they see this curtailing the revenue, albeit primarily due to short-sightedness, because much more money can be made through support contracts. Many developers at small software companies have similiar attitudes, because they are not subject to the contractual legal theft of their own creativity, which is par for the course for those who are employed by the big software companies. Because they are not subject to the same draconian contracts, which most developers have experienced, and for which reason Free software came into existence, they fail to see the benefits of free software development.
Open source is only better than non-free software if it means that one cannot take the code, alter it and obscure their changes, depriving the community of the benefits of expanded methods at our disposal. In most cases open source means free software, but some kinds of “open source”, apparently give the developer more “freedom”, due to the false notion that the developers him/herself is not substantially involved in maintaining and securing their own freedom. This type of “open source” software development is profoundly shortsighted.
The fact remains, due to free software development, more developers have more access to more methods(ie.code) that at anytime previous, and this benefit is available to all at not cost($).
as others point out here it doesn’t touch on the major problem with OSS in general – the licensing issue. As
Anonymous (IP: —.austin.rr.com) points out perfectly the “closed source” that the author talks about is referring to the bad old days that are (thankfully) receding outside the largest companies.
The licensing issue hangs over OSS and will continue as far as I can tell. The interesting OSS has to be released under LGPL or MozillaPL or BSD license and say NOTHING about why they do that. One word about the truth of the pure GPL will result in DoS attacks, bad mouthing, and hacks to the server.
Java Developer’s Journal had a good article about “coexistence” (I believe in Sept. or Oct. edition). Until the zealots are self-policed by the others in the OSS the licensing questions will continue to hold it back.
IMHO – if you truly like OSS models and want them to flourish:
1) stand up to the GPL bullies and demand pragmatic attitudes (like Linus, Miguel and others). The “biggies” like HP and IBM are leading the way.
2) release your source under MozillaPL or others that allow for extensions by open and closed source projects
3) design for cross-platform (GTK, Java, wxWindows, etc) and WELCOME the WIN32 users. This is the REAL way to get switchers to non-WIN32 OS – when the software “left behind” by the switch doesn’t matter.
Peace.
Can you explain what it is about the GPL that you find so objectionable? As far as I can tell, it is one of the things responsible for the most successful free software project to date (Linux kernel or Gnu tools, pick one). If the GPL disappeared, you’d certainly see companies taking the work of developers, doing whatever they wanted with it, and giving nothing back; but I’m not sure why you think that would be a good thing.
Not everyone who thinks that the GPL is a Good Thing is a bully or a zealot. They also aren’t necessarily wrong. There’s a reason that major companies are contributing to the development of Linux rather than just taking the code, developing their own versions independently, and doing whatever they want with them. The GPL created an environment in which something of this scale was possible, and in which there was no legal alternative.
Then I think we’re on the same side. I don’t have a problem with GPL under all circumstances. But the bullies exist and often dominate the discussion, to the detriment of us all.
“”As far as I can tell, it is one of the things responsible for the most successful free software project to date (Linux kernel or Gnu tools, pick one).””
What about Tron? AFAIk that has its own licensing, is open-source and has a far larger installed base than Linux or the GNU tools.
The author spends a great deal of the article railing against propietary development. Fine and dandy, I tend to agree there. He also spends some time on the infrastructure of ‘Open Source’ development. The CVS and the interaction between groups. Fine again.
But a closed-source company can be leader in standards and have a great development structure.
The core of open source is just that. The idea that your work is freely available to anyone for use and extension. You do not own the fruits of your own labour. Nor is it thought that you should be paid for your labour directly. Tt is also GENERALLY thought that money should be made through servicing, rather than the work you have actually done. It is this the author basically skips over.
While a nice article, I leave it with the same impressions of Open Source I had before.
Fair enough. I suppose one can measure success in different ways, and if one’s criteria is “most devices running,” then perhaps TRON takes the #1 spot (I haven’t seen reliable figures, but that obviously doesn’t mean there aren’t any, and it certainly seems plausible). I think that various GPL’d projects have been more influential and will be so in the long term, but it doesn’t really matter. My point is that the GPL isn’t somehow inherently evil, and its use can and does lead to amazing successes in the real world.
I agree that it is to the detriment of us all. Free rational discussion can’t take place when people aren’t able to say what they believe, or just get shouted down. While my sympathies lie with free software and the GPL, I find it ironic that many people who talk about freedom spend so much time trying to limit what can be said.
“You do not own the fruits of your own labour.”
Programmers for firms producing proprietary software don’t own the fruits of their labor. Also, many free software licences (including the GPL) do give the programmer control over what happens to their programs — they insure that the code cannot be stolen from the community. This is a _choice_ on the part of programmers, and it protects their code in a particular way. That doesn’t mean the code isn’t protected at all.
“Nor is it thought that you should be paid for your labour directly.”
It isn’t? Who is it that doesn’t think you should be paid for your labor directly? Companies releasing free software seem to be paying their employees. Even volunteer projects give money to developers when they’re able to. I challenge you to find one major free software proponent or organization that publically states that programmers shouldn’t be payed. The fact that there are volunteers who willingly give their time and code doesn’t mean they don’t want to be payed or don’t think they should be payed.
“Tt is also GENERALLY thought that money should be made through servicing, rather than the work you have actually done.”
To the extent that this is true (that is, a little bit), it is a matter of practicality; and what will matter in the long run is whether or not it works as a business model. Furthermore, it is only partially true. Companies sell free software every day. Go down to your local mega-ultra-huge-computer-store of choice; chances are you’ll see free software for sale.
IMHO,
the GPL is really only good in 2 areas.
1. people who don’t want to make money on the software they write. This would include people like university researchers, people who just want to code to get something done or for fun.
2. Large corporations who don’t sell software but sell a complete system. This is like the old analogy of the telephone exchange system. This includes IBM, Sun…
The GPL is probably one of the poorest licences for the small business person or for anyone interested in selling software (game developers, image editing packages…)
The “Nor is it thought that you should be paid for your labour directly.”
and
“Tt is also GENERALLY thought that money should be made through servicing, rather than the work you have actually done.”
are one in the same.
I have been down to the local computer store. No, I don’t see a whole lot of ‘free’ software. I see redhat linux, which last I checked gave up on users buying their software and focussed on enterprise. The games I buy all tend to make thier money from the non Open Source community.
As to who says money should be made off servicing in the Open Source community not on the sale of software itself or the labour in creating that software..its been said 1 000 000 times over in places like Osnews and slashdot. I don’t think you’d have to look far for an example of that.
I personally belive that all code should be viewable, but you should be able to modify it and sell it given proper royalties of course.
Open source is only better than non-free software if it means that one cannot take the code, alter it and obscure their changes, depriving the community of the benefits of expanded methods at our disposal. In most cases open source means free software, but some kinds of “open source”, apparently give the developer more “freedom”, due to the false notion that the developers him/herself is not substantially involved in maintaining and securing their own freedom. This type of “open source” software development is profoundly shortsighted.
That was about the most crappy thing I’ve ever read. The problem is for both smaller and bigger companies that GPL, which you seem to mean is what free software is all about, seriously hurts the ability for small/medium sized companies to make software and actually live on it.
I’ve read all this stuff about GPL companies trying to get by etc… but frankly, most people who support and write code for them are students and there are only a few employed. Sure that does work, to some extent, but in the long run it’s a dead end.
GPL’d software or Free as in under obligation resulting in Imprisoned is a serious threat to small innovative people who want to make a living writing software.
Now MIT’d or BSD’d software… or MPL’d or APL’d software is a way to successfully contribute to the community as well offer all the methods you talk about WITHOUT limiting the opportunity for people to make money.
I believe that GPL is a destructive license made to limit options and opportunities for people to make a descent living on what they love to do (assuming many devs like what they do).
And to prove the very thing that other licenses ARE successful AND can survive I’ll give you some examples..
XFree86
FreeBSD
OpenBSD
NetBSD
Apache
Mozilla
Just a few examples…. and Apache, which is NOT using GPL obviously have more than 50% of the global webserver market. Even microsoft at some point were going to give some small support for it.
Is there any GPLed product which, in terms of market share, can compete with Apache? Ehhhrrrr nope…
Still the GPL as in Imprisoned code is the better choice?
Yah right…
Can you explain what it is about the GPL that you find so objectionable? As far as I can tell, it is one of the things responsible for the most successful free software project to date (Linux kernel or Gnu tools, pick one). If the GPL disappeared, you’d certainly see companies taking the work of developers, doing whatever they wanted with it
*cough* Apache *cough*…
Since Mozilla is a pretty interesting License, I see Sun just stealing it for their Java Desktop? Whatabout Ximian?
don’t you see… being able to build on top without issues stopping them from putting some sugar on top is a good thing. They will still push in things which are called necessitys and compete with luxury…
Drivers is one of those things where companies benefit from helping eachother… geeez, why not let people post closed source drivers… even better… as long as it works….
Anyway… what I was saying is basically for BSD or Linux or whatever…. companies are interested in making it work for all platforms etc… so they offer time to write drivers… while sugar on top issues are closed source… Isn’t this a good thing?
Is there any GPLed product which, in terms of market share, can compete with Apache? Ehhhrrrr nope…¨
How about Samba? That is a very successful GPL’d project, and the people who work on it are all well-paid engineers.
What about the Linux Kernel? This is a hugely successful project, released under the GPL.
OpenOffice.org, another successful software project, is dual-licensed under the GPL and the SISSL (Sun Industry Standards Source License).
Lots of smaller successful projects are GPLed as well.
It is false to say that the GPL takes away from the developer’s freedom, since he/she can freely choose to release the software under the GPL or not. If he wants to re-use GPL software, he must release under the GPL, but again he’s under no obligation to re-use GPL code at all!
If the GPL was such a destructive license, why would it be used so widely?
Troll Tech has done very well with the gpl verion of QT. They have also have a lisence available (for $$$) for closed source commercial development, which to me works out great. Free for those who are developing for free, and costing money for those who are making money off their develoment software.
MySQL also follows this model very well. If you can dual license your work (gpl + commercial), you can do rather well for yourself.
Yep,
No argument there. That’s one reason why I love QT. If you want to develop free software, they’re there. If you want to sell your software, QT is there as well. Of course, now that u want to sell your software, you have to pay them for using their code for profit.
Makes sense to me. Sadly, their license is not the norm in the open source evironment. Nor do they show the same willingness to be free towards windows users who may wish to develop open source software.
you state that every other license out there is pretty much better then the GPL…in your humble opinion, i’m sure.
well here’s the thing.
GPL is not just about being open source. there are plenty of other licensing schemes out there that allow source to be open, without the GPL’s requirements.
THE POINT of the GPL is to ACTIVELY prevent mega-corporations from owning the IP rights to everything everywhere.
Without GPL, there is absolutely nothing to stop companies from snapping up opensource/public domain software and then doing an embrace/extend number on them, making the originals irrelevant.
so AC take your head out of your arse. we have plenty of licensing schemes that will accomplish the same thing, be it closed source, bsd, or public domain style licensing…the end result is the same…big money can buy it out, or make it irrelevant via embrace/extend type strategies.
isn’t there enough of that already? how about something to help (in some tiny way) to balance the scale. that’s the point of the GPL. to stop the eventual ownership of all IP, and all standards by mega-corps.
the part of capitalism that most people like is competition. we _love_ competition. makes for great products, good prices, and a fertile economy.
the part that people forget is that eventually, in capitalism, there is a WINNER. and that winner eventuallysome humongous gazillion dollar mega-corp…leaving the losers like burnt dead carcasses.
when small business (the part that really drives this economy) is wiped out by the walmarts, microsoft, and taco bells..of the world….and you find yourself an employee with a badge stating “empl.# 4,506,234” .you’ll be wondering what the hell went wrong.
i run my own business, i believe in the free market, i believe in the right to bear arms (kick my door in during the middle of the night and find out), i believe in choice.
but i’m not so stupid as to not recognize that just as we can be ruled like sheep by “big government” we can similarly be ruled like sheep by “big corporations” and their big money.
to not recognize that possibility, is to be a complete fool.
First of all… OOo is LGPLed not GPLd… OOo isn’t that successful yet to be honest. Compared to MSO it’s definitely not the competitive product many have wished for even though I like it a lot and think it’s improving.
QT is dual licensed rather than simply GPL which is a HUGE difference. They can do whatever they want with their software pretty much and also offer the commercial end of it, unlike strictly GPL, same goes for MySQL.
the part of capitalism that most people like is competition. we _love_ competition. makes for great products, good prices, and a fertile economy.
the part that people forget is that eventually, in capitalism, there is a WINNER. and that winner eventuallysome humongous gazillion dollar mega-corp…leaving the losers like burnt dead carcasses.
when small business (the part that really drives this economy) is wiped out by the walmarts, microsoft, and taco bells..of the world….and you find yourself an employee with a badge stating “empl.# 4,506,234” .you’ll be wondering what the hell went wrong.
Now here’s the real issue. You see a big company like Microsoft or IBM has the money and the resources to build a product from scratch and outrun any GPL product if they like. A small company can not. The small company HAS to depend on getting some stuff for free.
So the philosophy you’re talking about is exactly why the less strict licenses really offers a capitalistic view on software to work.
All this talk about IP etc… I think softwares previously mentioned are perfect examples of things not getting locked in… let me repeat them once again so I can hear your view on how they get completely exploited and left to rotten
Xfree86
NetBSD
FreeBSD
OpenBSD
Apache
Mozilla
For instance BSD has helped OSX to become something, it helped MS to improve their netstack, it helped so many people without fainting away.
Whatabout Xfree86? I bet that’s what you’re depending on for your Linux desktop isn’t it? Has SUN demolished XFree86? I think not…
I can explain why the GPL is so widely used…
A) It’s because that’s the license people been imprinted to use in lack of knowledge… the so called RMS disease.
B) Socialistic/Communistic views on society of today…
If you fall into the b) part I can accept it, even though I dislike it. The a) part is just complete madness and is really horrible to see.
Oh… and I also strongly believe the License issues is why so many start with Linux, get educated and then oftenly move on to BSD for instance…
“Is there any GPLed product which, in terms of market share, can compete with Apache? Ehhhrrrr nope…”
For one, market share doesn’t say everything. For two, yes, such exist. I’m not gonna do the dirty work for you but yes, there are GPLed and _fast_ webservers. Apache doesn’t scale for some tasks.
For example on embedded devices there’s Fnord, which is GPL. Very small binary. It scales very well and is ideal for HTML or image sharing purposes. PHP doesn’t scale anyway…
Both the author of this article and a lot of the posts here are wrong IMHO, software licences are completely beside the point when it comes to development methodology, you could use the open source methodology he describes for a closed source project and vice versa (unless of course i misunderstood what he meant by standards) and both methods have their pros and cons.
This can be easily seen with when he says “In the closed model above, software development begins with an analysis of a user’s needs without consulting existing standards.“, nothing wrong in exchanging standarts compliance with user satisfaction if you can afford the increase in development time and investment capital.
Whereas on the other model you’ll have a faster development and better standarts compliance with the downside of the team members lacking specialization in some areas.
(Article URL doesn’t work here)
“A) It’s because that’s the license people been imprinted to use in lack of knowledge… the so called RMS disease.
B) Socialistic/Communistic views on society of today…”
Please state proof. You don’t state _any_. I’m not gonna state arguments for why-not when there are no arguments states why. I never heard about that disease either, perhaps it is so-called by you only? Thanks,
“Oh… and I also strongly believe the License issues is why so many start with Linux, get educated and then oftenly move on to BSD for instance…”
You believe? Cool, i believe in dogs which fly to Mars. Please state proof, like above. Thanks,
Like or not, the reason that Gnome is growing is because of LGPL. The GPL prevents extensions that are not given away for free and locks one into the “services” model of software that no one (repeat NOONE) has made into a successful business model. Even RH is no longer trying, because the software _is_ the investment and giving it away can never be profitable.
Linux is coming of age, which is all to the good. But the idea that GPL software can make a large-scale profitable business model based on services simply has an awful track record. In five years there will be increased Linux usage. But if you think for a minute that Solaris/Java Desktop combo, or Novell/Suse desktop, or RH on IBM/RH supported desktop will cost LESS THAN .NET/Longhorn, you are deluded my friend. It will be different, no doubt. It will not be less expensive.
As the GPL OS companies come of age the profits will have to come from somewhere. I suggest you look in the mirror for the source.
Hi Anonymous from bitstream.com. I don’t care about GNOME, nor i understand why you addresss your post to me. However beyond the fact that i don’t give a rat wether Sun can exploit/use a desktop i have a few points for you which you haven’t addressed: you claim GNOME grows so fast because of it is LGPL and state no proof. I take it you also believe KDE http://www.kde.org/whatiskde/project.php#factsandfigures>grows because of it is LGPL/GPL? For the record: i don’t use either of these 2, i’m not a ”fan” of either.
Should i state that this combined multi-license model is actually both company as well as freedom-friendly? Do i have to proof that GPL companies still exist? Perhaps you can state companies which moved away from GPL? Perhaps you can enlighten your anti-GPL zealotry by looking to the OGG Vorbis licenses (LGPL, GPL, rBSDL). It would benefit against your current black vs. white thinking. GPL is compatible with various licenses, you know…
First of all… OOo is LGPLed not GPLd…
That doesn’t make much of a difference in this context. So, what you’re saying is that the GPL is evil and the LGPL is good?
OOo isn’t that successful yet to be honest. Compared to MSO it’s definitely not the competitive product many have wished for even though I like it a lot and think it’s improving.
OOo is very successful considering its young age and the market penetration of MS Office. The fact that we use OOo at my company (a game developer, and thus a proprietary company) is proof positive of its success, in growth if not in market share (yet).
While reading the author’s comments about the differences between OS and CS development cycles, something sorta stuck out.
I am NOT a developer, unless one counts PHP, HTML, etc as developing (I don’t). But I do consider myself an expert end-user. In reading his description of the development process one sees that the closed source has a guy consult the user, someone else write the specs and so on. According to the author’s narrative, the user is consulted before, during, and after the project. The OS shop consults the specs, other OS projects, and so on. At no point in his description of the OS process is the user mentioned.
At one point he does mention that the people who originally created the standards consulted the user at some point in time. Consulting the user is lumped in with “wasted time.”
In my perspective the relative inferiority of OS software as compared to proprietary software in the useability dept DOES reflect this low regard for the end user. It does do the job, but rarely does it do it elegantly. Backend stuff is fine (as in PHP, Apache, MySQL, etc). Front end stuff – never.
As far as standards bodies consulting the user let’s look at web standards. I’m not talking about the person viewing the web page as the end user, I’m speaking of the web authors. These are the real beneficiaries of web standards. We don’t have to use Netscrap’s or MS’s proprietary tags for presentation effects. But look how LONG it takes standards bodies to act. We like standards so we don’t have to code for 40 different browsers (like OpenGL or DirectX APIs for graphics card makers).
Tables are the big deal for web authors. Is CSS adequate for replacing tables? Any author will resoundingly say “No” to the current implementation. Unless you’re ugly and boring and mediocre. Tables for layout was a work-around – a “trick” – to enable better positioning due to a lack of presentation capabilities in the original HTML standards. CSS still does not offer the same ability as tables (rewrite a vBulletin board without tables with the same customizability). CSS – designed specifically for presentation – still cannot beat tables – a workaround. To get anywhere near the same functionality, we have to resort to CSS tricks which never display uniformly across browsers – mostly because these tricks aren’t supported. As CSS develops, it’s gaining in complexity without much gain in useability. We won’t even mention Flash.
Let’s look at the underlying basis behind this connundrum. All WC3 cares about is: “Is the information structured correctly?” It never even occured to them that the presentation of the information is just as important as the info itself, or in this case, the structure of the information. Anyone who disagrees with that can just close the windows on their ivory tower. So now presentation is being tacked on as an afterthought, mostly at the urging of the CS corps and the web creation community. We were doing it with or without WC3 and so they were losing control.
If we limited ourselves to using open standards exclusively, the web would be a much uglier place. Yes, the info would still be there, but it would be pretty mediocre.
In the overall scheme of things, my experience using the applications says your description of the process accurately reflects the end products – OS is less user-oriented than CS but more interoperable.
“We focus on functionality first and polish second.”
Other comments in the article back this statement up. It sounds nice on the surface, but it is the same mindset as the W3C. Regarding the presentation of the application as “polish” (and it is obvious this author considers it as such) demonstrates the engineer/ programmer mindset – not an end-user/ re: customer/ re: the person who’s commissioning this work/ re: the whole reason why we are doing this -focused approach.
Functionality IS important, but equally important is the presentation of that functionality.
There are three things that stand out of your argument:
At no point in his description of the OS process is the user mentioned.
The idea is that the user is rarely if ever consulted (say, by focus groups or polls) in OSS, while closed-source IT shops consult users all the time.
In reality, though, things are quite different. Successful OSS products get a lot of feedback from a certain type of users: those who receive the mailing lists and participate in beta testing. This skews things a bit towards the technical users, however the input can be priceless for developers.
The second point you make is that OSS apps are deficient in usability. You seem to forget that a lot of proprietary apps have a crappy UI as well! Elegance is certainly not unheard of in OSS as well; in fact, one could convincingly argue that the KDE and Gnome desktop, along with the new generation of QT and GTK2 apps, are as “elegant” in usability as OS X, surpassing WinXP.
Some great apps, from a UI point of view:
http://k3b.sourceforge.net/uploads/images/7/audioproject.png
http://www.trolltech.com/images/screenshots/designer/mac.png
http://apollon.sourceforge.net/pictures.html
http://www.methylblue.com/filelight/#screenshots
http://www.frozen-bubble.org/
And I could go on and on. I’ll add that, as far as eye candy goes, Linux had one of the most beautiful icon sets ever made first: Everaldo’s Crystal:
http://www.everaldo.com/crystal/
Finally, you make some weird opposition between tables and CSS, though it is very possible to use both within the layout of a web page, and rail a bit about the need to adhere to non-MS standards.
Really, it’s kind of a moot point, since Mozilla can render practically any web page IE can. Konqueror is pretty good as well (it’ll be even better in 3.2, with the Safari enhancements) – just be sure to masquerade Konqueror as “IE 6 on WinXP” sometimes, so that pages that might refuse to display (or display correctly) because you don’t have a “supported browser” suddenly pop up fine!
There’s so much that I find wrong in the article that I want to write about. But maybe when I have time.
Anyway, a person has, by chance, neatly summarized all my thoughts on the commentary section of the authors website.
====
…
What’s worse, the article does not make sense. The author consistently confuses the ideas of “Open Source” and “Standards Based”, as though the two were tightly linked. He uses examples from his own experience to make sweeping (and incorrect) generalizations, such as the idea that Closed Source development implies a specific (and in his description, poor) process. These as well as other inaccuracies make me (and I would risk a guess, other professional developers) doubt his technical grasp on the topic.
…
====
There is another discussion on Anti-Slash about this same topic, but attached to a more interesting article.
http://www.anti-slash.org
Sorry about the tables/ cs – the illustration wasn’t done very well. Anyway, it is meant as an illustration of why relying on standards bodies for user testing is not practical. The author stated that OSS shops relied on standards bodies as well as work by other developers (by reutilizing existing code) in useability testing.
Standards bodies are fine for interoperability, but not useability.
Using your statement – techies and such on mailing lists provide user testing. Techies look at the world with different eyes than users. The original W3C guys had plenty of user testing with other techies. An infinite # of mailers went out “Can you read my web page?” “Yep” Can you read web page?” “Yep”… And everyone was happy with their infinitely scrolling textedit/ notepad-looking whitepapers on esoterica. And then some guy like me said “How do you change the font face?” And some techie said “Why would you want to?” And that’s why presentation of the data is an afterthought and standards bodies are inadequate to the task of user testing.
Yes, you can use tables/ CSS on the same page if you are trying to design a site to best suit your clients’ needs. On the other hand, if you are relying on a standards body (W3C) to provide you with guidance, the use of tables as a layout device is highly discouraged. CSS/ XSL/ XSLT should be your sole presentation markup.
Using your statement – techies and such on mailing lists provide user testing.
That’s why I say that results are skewed. This is changing though – comments by less technical users are becoming more frequent on mailing lists as Open Source’s popularity grows. Sure, it’s a relatively slow growth, but it is happening. Also, OSS developers are more and more aware of UI issues, and as such you have lots of more recent software that shows a marked increase in useability. The various open-source apps I gave links to are good examples of this.
Also, one should not overestimate focus group testing. I work at a company which sometimes use focus groups, and while they are helpful, the fact is that users are quite a varied bunch, and not everyone has the same ideas and/or expectations about how a UI should be designed…It is an imperfect science. Bad UI designs obviously survive focus groups and other forms of user input in proprietary apps as well! The latest realplayer is a good example of poor UI design (and the latest Windows Media Player as well, for that matter). Meanwhile, an Open Source media player solution – Konqueror with embedded Kmplayer – has a great UI…
What you say may have been true at a certain time, but things are evolving quite fast in this department.