Creating Standards is altruistic?
update: based on the comments, I learned that the original idea "creating standards is altruistic" was wrong. It seems that altruism plays a role, but vendors gain a market advantage when implementing a standard first. Making good products that spread is another way to create a standard.
so, this is wrong for now:
Creating standards, as for example the W3C or the ISO does, is an deeply altruistic behavior and I think a hint that the people involved are altruistic.
A definition of altruistic is "Benefiting others without regard for one’s own needs or safety."
Creating a standard is usually a process of endless and endless discussions, exchanging arguments, giving arguments for every decision, reworking drafts, fixing formulations, agreement, and discourse, and this process can last over years, and usually does last more than a year. The outcome is rather simple: a document describing the decisions in one simple manner. If you look at the HTML, HTTP and URI specs, you can read and understand them in a day. It took years from the first idea to write and improve them to their state, and it took a lot longer time, namely many years, to establish them as a standard.
Now, what is the benefit of creating a standard? For the author, nothing. The author doesn't get paid for it (unless you charge for reading your standard, what many organisations do). You don't get fame nor money, because you invested all your time into the standard, but not into your product, which can now by copied by everyone in the world by implementing the standard.
So, why do people do standards? I cannot say for you, but I work on them because I think that the world would suck very hard if we not had standards for some things. Think of having to switch to a different browser when looking at asian websites, like carrying around an platoon of power adapters so that you hopefully can suck electricity from whatever socket you are confronted with.
The opposite of making a standard is implementing the solution. Then you have it, your great solution, coded in executable binary, nothing can go wrong now. You are quick to do it, you don't have to care much for documentation, and your customer is happy.
But imagine looking at a website hosted by an Internet Information Server (Microsoft) would mean that you have to use a Microsoft Operating System, or a Microsoft TV, or a Microsoft Phone. At first, there is no problem. But after others did the same, I would have to carry around many phones. So, clever people (who are often altruistic and may work for market leaders like Microsoft Corp.) see these problems beforehand and decide to sit down to make a standard, once sitting you notice that there are others seated next to you who feel the same.
Some say "the good things about standards is that there are so many of them to choose from". This is not true, because for anyone in such situation of choosing, "good" is exactly not the way you feel, because you get this slight feeling deep inside, that your decision is shiny today, but sits down in some dark place and comes to bite your behind years later.
So, if you are stuck again while working on a standard, like me at the moment, and life is blackened by endless arguments and paperwork, think about why you started the whole thing.
And on the other hand, if you watch people like the W3C members sitting around in endless discussions about details, dissecting every piece of it and making a lot of seemingly useless paperwork, and you think to yourself "ha, I could do a better standard alone", think twice. First of all, open your cupboard and get out your beloved assortment of travel power socket adapters for a short personal meditation about the stupidity of man. Then, think of the motivation of people writing standards: not money nor fame can be the reason, there is neither of it in it, they do it for you.
p.s. HM, the Queen of Britain, has honored Tim Berners-Lee again by giving him the Order Of Merit, putting him in one league with Florence Nightingale and Mother Theresa. I think this does not only honor him but also everyone wasting his life away for the W3C and technical standards in general. People travelling as much as the royals do, seem to know the irony of power sockets that look a bit unfamiliar compared to the plug you want to put in them, and cars that have their gear switch on the left hand side of you, adding to the irritation of driving them on the "wrong" side of the road. Feel free to comment.
p.p.s. perhaps I don't get it because I am a complicated thinking Java programmer and not a witty Python hacker (like Gromgull), where the standard is set by best practice.
so, this is wrong for now:
A definition of altruistic is "Benefiting others without regard for one’s own needs or safety."
Creating a standard is usually a process of endless and endless discussions, exchanging arguments, giving arguments for every decision, reworking drafts, fixing formulations, agreement, and discourse, and this process can last over years, and usually does last more than a year. The outcome is rather simple: a document describing the decisions in one simple manner. If you look at the HTML, HTTP and URI specs, you can read and understand them in a day. It took years from the first idea to write and improve them to their state, and it took a lot longer time, namely many years, to establish them as a standard.
Now, what is the benefit of creating a standard? For the author, nothing. The author doesn't get paid for it (unless you charge for reading your standard, what many organisations do). You don't get fame nor money, because you invested all your time into the standard, but not into your product, which can now by copied by everyone in the world by implementing the standard.
So, why do people do standards? I cannot say for you, but I work on them because I think that the world would suck very hard if we not had standards for some things. Think of having to switch to a different browser when looking at asian websites, like carrying around an platoon of power adapters so that you hopefully can suck electricity from whatever socket you are confronted with.
The opposite of making a standard is implementing the solution. Then you have it, your great solution, coded in executable binary, nothing can go wrong now. You are quick to do it, you don't have to care much for documentation, and your customer is happy.
But imagine looking at a website hosted by an Internet Information Server (Microsoft) would mean that you have to use a Microsoft Operating System, or a Microsoft TV, or a Microsoft Phone. At first, there is no problem. But after others did the same, I would have to carry around many phones. So, clever people (who are often altruistic and may work for market leaders like Microsoft Corp.) see these problems beforehand and decide to sit down to make a standard, once sitting you notice that there are others seated next to you who feel the same.
Some say "the good things about standards is that there are so many of them to choose from". This is not true, because for anyone in such situation of choosing, "good" is exactly not the way you feel, because you get this slight feeling deep inside, that your decision is shiny today, but sits down in some dark place and comes to bite your behind years later.
So, if you are stuck again while working on a standard, like me at the moment, and life is blackened by endless arguments and paperwork, think about why you started the whole thing.
And on the other hand, if you watch people like the W3C members sitting around in endless discussions about details, dissecting every piece of it and making a lot of seemingly useless paperwork, and you think to yourself "ha, I could do a better standard alone", think twice. First of all, open your cupboard and get out your beloved assortment of travel power socket adapters for a short personal meditation about the stupidity of man. Then, think of the motivation of people writing standards: not money nor fame can be the reason, there is neither of it in it, they do it for you.
p.s. HM, the Queen of Britain, has honored Tim Berners-Lee again by giving him the Order Of Merit, putting him in one league with Florence Nightingale and Mother Theresa. I think this does not only honor him but also everyone wasting his life away for the W3C and technical standards in general. People travelling as much as the royals do, seem to know the irony of power sockets that look a bit unfamiliar compared to the plug you want to put in them, and cars that have their gear switch on the left hand side of you, adding to the irritation of driving them on the "wrong" side of the road. Feel free to comment.
p.p.s. perhaps I don't get it because I am a complicated thinking Java programmer and not a witty Python hacker (like Gromgull), where the standard is set by best practice.
|
leobard - 29. Jun, 23:48
|
|
Ora Lassila (guest) - 1. Jul, 02:07
Standards in the Real World
I don't think I quite agree with your view of standards. Even though it sounds nice, altruism plays a much smaller role in the standards definition process than you would think. Take W3C, as an example: Most members of the consortium are commercial companies which typically do very few things because of altruism. Companies participate in the standards process because they have something to gain. Ideally, this is a "pre-competitive" activity, where the emergence of a standard actually creates or enables a market. The standard is a prerequisite for doing business in the first place. Sometimes, however, participation in the process is dictated by the need to keep competitors in check. The beauty of this is that the process itself can be crafted in a way that companies get what they want, and those people who are driven by altruism (like, say, the founders of W3C) also get what they want.
Usually, a standard should be created once the associated technology is sufficiently mature or is understood "well enough". In the case of the Semantic Web, the situation is quite peculiar: The initial standards (like RDF and OWL) were created very early ("too early" some would say) but this was actually necessary for the subsequent research work to be possible.
As for writing the standard by yourself, yes, that of course is possible and in fact the result probably would be better. The benefit of a group of people agonizing over seemingly worthless details over a period of a year or more is that it creates "buy-in" and that is incredibly valuable, because a standard is only worth something if various organizations, companies, etc. are committed to actually deploying it.
Usually, a standard should be created once the associated technology is sufficiently mature or is understood "well enough". In the case of the Semantic Web, the situation is quite peculiar: The initial standards (like RDF and OWL) were created very early ("too early" some would say) but this was actually necessary for the subsequent research work to be possible.
As for writing the standard by yourself, yes, that of course is possible and in fact the result probably would be better. The benefit of a group of people agonizing over seemingly worthless details over a period of a year or more is that it creates "buy-in" and that is incredibly valuable, because a standard is only worth something if various organizations, companies, etc. are committed to actually deploying it.
leobard - 11. Jul, 08:55
creating standards II
(meta: I assume that above comman was written by Ora Lassila, but without openid, you never know.)
ix, the argument that by promoting SPARQL, MQL or Versa have been held back, is true. The same will probably happen with RDF VS microformats, whoever gets popular first, will be used.
My definition of standards is rather vague, the blogpost sums up some personal assumptions I have. I am working now in one EU project were we create standard documents, but I have never worked in a W3C committee (I contribute to an interest group for marketing).
The binary statement: "The opposite of making a standard is implementing the solution.", can also be read as two ends of an Axis, where a standard-making-process can be placed. The W3C require reference implementations at some stages of the process. Measuring the "invested work" against the visible outcome counted in documents, implementing a solution as a vendor and then documenting it is cheaper than implementing the reference implementation and then discussing it within a working group. This difference is also part of the binary sentence. Well, the sentence was blunt. But it caused you to comment, which helps me ,and the interested reader, to learn.
If I understood it right, ix wrote that creating a product that is prevalent because of its usefulness is altruistic. I would not connect that as a causal connection with altruism, because some of these products were made to make profit. But as Ora Lassila has put it, creating W3C standards is also driven both by altruistic and by commercial interests.
So probably the whole title of the blogpost is wrong, "creating standards is altruistic". I will fix that.
about the status of RDF: as Ora Lassila put it, RDF and OWL were created very early, they have some "rough edges". I also had trouble with XML serializations, but often they were caused by the "latest SVN-alpha" versions of libraries I used.
ix, the argument that by promoting SPARQL, MQL or Versa have been held back, is true. The same will probably happen with RDF VS microformats, whoever gets popular first, will be used.
My definition of standards is rather vague, the blogpost sums up some personal assumptions I have. I am working now in one EU project were we create standard documents, but I have never worked in a W3C committee (I contribute to an interest group for marketing).
The binary statement: "The opposite of making a standard is implementing the solution.", can also be read as two ends of an Axis, where a standard-making-process can be placed. The W3C require reference implementations at some stages of the process. Measuring the "invested work" against the visible outcome counted in documents, implementing a solution as a vendor and then documenting it is cheaper than implementing the reference implementation and then discussing it within a working group. This difference is also part of the binary sentence. Well, the sentence was blunt. But it caused you to comment, which helps me ,and the interested reader, to learn.
If I understood it right, ix wrote that creating a product that is prevalent because of its usefulness is altruistic. I would not connect that as a causal connection with altruism, because some of these products were made to make profit. But as Ora Lassila has put it, creating W3C standards is also driven both by altruistic and by commercial interests.
So probably the whole title of the blogpost is wrong, "creating standards is altruistic". I will fix that.
about the status of RDF: as Ora Lassila put it, RDF and OWL were created very early, they have some "rough edges". I also had trouble with XML serializations, but often they were caused by the "latest SVN-alpha" versions of libraries I used.
- add comment - 0 trackbacks
Trackback URL:
https://leobard.twoday.net/stories/4017979/modTrackback
but creating 'standards' out of the blue based on academic discussion has a much tougher argument for altruism. i'm having trouble thinking how something like the RDF/XML 'standard' helped me when i couldn't even successfully crossgrade data between redland, virtuoso, and jena without parsing/import errors in one or both directions (or exporting errors for that matter), requiring me to invent my own JSON format that actually worked for what the 'standards-compliant' apps couldn't. 2 interoperable implementations does not make a standard. 80% market concensus and little change in that concensus for 2 years and 3-6 'major' implementations and many minor ones, does.
i could also argue that making up a standard like SPARQL has held back implementations of Versa and development of MQL-like languages, by virtue of how those who blindly follow 'standard's bodies allocate their development resources.
so i agree that standards are altruistic, i disagree on your definition of standards. i also especially take issue with "The opposite of making a standard is implementing the solution." then equating these with binary/proprietary. many who created what became de facto standards knew the 'rules' before they broke them.. and often offer their results as open source. standards should never 'prescribe' solutions, and it sounds like thats what youre arguing for here (along with the good feeling of being the 'good doctor'.) i don't buy it.