5 ms·

That does not mean that it is a good thing to put all bells and whistles from different browser vendors under the same umbrella. Let's say you are a) owning an

That does not mean that it is a good thing to put all bells and whistles from different browser vendors under the same umbrella. Let's say you are a) owning and b) managing a project that three competing teams are working on in parallel. If you will not curate the project you will have one team adding <marquee> and another team adding <blink>... In real world you would invite dedicated architect or team of architects to define the spec that all 3 teams will implement. All that WHATWG vs W3C flame is about bazaar vs cathedral management style I think. Origins of the mess: W3C itself has no architects on board - they are just trying to moderate votes of others. Where each vote has an obvious weight (weight(Google) > weight(JohnButSmart)). In contrary WHATWG has professional architect on top of the construct - Ian Hickson. In fact WHATWG was created by him. But Ian is associated with Google and that makes WHATWG legitimacy a bit questionable.

make3 9y ago
the formal complaints of all browser people seem to support the WHATWG
- ttepasse 9y ago
  The browser people are the WHATWG. There are no other stakeholders in that.
  - domenicd 9y ago
    Not sure exactly what your comment is saying, but at least one reading is that "browsers are the only stakeholders in the WHATWG", which is not accurate. The WHATWG is a community organization open to participation by all; see https://whatwg.org/faq#process https://whatwg.org/faq#process for more. We receive a lot of participation from users, web developers, and other companies. There is a formal group, the Steering Group, which represents the browsers that implement WHATWG standards, and serves as the point of final appeal if the community doesn't come to a consensus on its own. But this is similar to the W3C appeals track where if all the paying member companies don't come to consensus, they appeal to "The Director" as the ultimate decider. ("The Director" is nominally Tim Berners-Lee, but recently all Director decisions have been made by W3C management "on behalf of" The Director.)
    - hungryfoolish 9y ago
      Are there any non-browser members as part of the editors or steering committee in the whatwg? To me, this is giving browser makers even more power over HTML than they have otherwise.
      - domenicd 9y ago
        The MIME Sniffing, Streams, Console, and Quirks Mode Standards are all edited by people who are not working for the browser-engine-developer companies. (Streams is co-edited by Googlers as well.) That's 4 out of the 15 standards currently developed at the WHATWG; not so bad, given how few companies are willing to pay people to work full time on web standards. Of course, we have lots of work to do, and if you or anyone else are able to devote a good chunk of time to standards work, we'd love to have more editors---no matter what company they work for. Browser engine developers comprise the steering group, since ultimately we want the specs to reflect what will be implemented in web browsers, so the best way to resolve disputes about what should be in the spec is by asking the people who will be spending their resources to implement it what they think. I think this is probably the right way to go, instead of "The Director" having the final voice.
        hungryfoolish 9y ago
        > That's 4 out of the 15 standards currently developed at the WHATWG; not so bad, given how few companies are willing to pay people to work full time on web standards. Thats actually nice to know, but specifically about HTML (which is one of the most important specs the whatwg works on) it's all browser makers. (That too, majority of them from one particular browser maker). >I think this is probably the right way to go, instead of "The Director" having the final voice. I think having a better mix of people in the decision making process than just browser makers would be the right way to go. I understand that browser makers have to implement the specs, but the web community as a whole has to use the specs to build the web of the future - and as such, people from non-browser companies should have a greater say in the final voice.
        9y ago
  - dragonwriter 9y ago
    Not only are there non-browser-maker WHATWG stakeholders, there are non-browser-maker formal objectors to the W3C DOM move to CR endorsing the same view as the four browser makers (Bloomberg and Disruptive Innovations; I've never heard of the latter previously, AFAIK.)
    - ttepasse 9y ago
      I was curious so I googled a little bit: Bloomberg is know for its Terminal. It seems they have a program where you can extend their terminal with web applications. Presumably Bloomberg didn't built a renderer themselves but there are custom APIs. So they are something between a browser maker and a UI framework like Electron. Disruptive Innovations is a firm by Daniel Glazman, formerly of Netscape and of the CSSWG fame. His main product seem to be continuing the Editor part of the original Mozilla Suite based on the Gecko Layout Engine. There seems to be an NVU Editor and rather new an Editor called BlueGriffon which also does ePub and maybe his custom WebBooks format.
domenicd 9y ago
This information about how the WHATWG works is somewhat outdated; please see https://whatwg.org/faq https://whatwg.org/faq for more.
lambda 9y ago
I'm pretty sure Ian Hickson is no longer involved with the WHATWG. I think he's working on Flutter now (https://news.ycombinator.com/threads?id=Hixie https://news.ycombinator.com/threads?id=Hixie). In fact, I would say that it's not quite like a cathedral and bazaar here. The W3C is like a legislative committee. Lots of things done by committee vote. Formal process that must be gone through to advance beyond committee. Petty politics that screw with actually getting things done. The committee chair having the ability to block work by procedural means. The WHATWG is more like an open forum. The work happens out in the open. Anyone can participate in the discussion (the W3C committees frequently have private meetings, there have been private mailing lists, and so on). There are a few people who have commit rights to the repo, so who actually control what goes in and what doesn't, but they are generally willing to let in changes that have broad support, and are implementable, rather than letting such features get held up by political processes. Besides the difference in structure, there's just a difference in attitudes. The W3C tends to strongly favor certain principles, like accessibility and modularity, but to the exclusion of compromise for technical reasons or real-world reasons. They also seem to have a tendency to get very attached to particular ways of doing things, without being willing to compromise. I think the biggest example of this was longdesc; it was never implemented properly in pretty much any browser, and very few people actually followed the spec and had it point to a URL (many people just copied the alt tag, or provided a longer description in the attribute instead of a URL), so even if browsers or screen readers had implemented it, the content wouldn't be useful. But people in the W3C made a big stink about removing this, and spent a lot of time and effort fighting and litigating over that, rather than actually trying to work on a different feature that could gain wider adoption. The WHATWG tends to take a more pragmatic approach; pave the cowpaths, if there are differences between implementations do it in the most sensible way that preserved compatibility. Now, what the WHATWG has produced hasn't always been the best; there are times when it's made mistakes in its approach. The drag and drop spec, which had been reverse-engineered from IE's drag and drop support, was a pretty bad; not sure if it's gotten better since. But overall, the WHATWG has been a lot more productive in getting standards done that are actually used on the web, because they involve the implementers, and don't override them with tedious, drawn out, political battles over obscure features that no one uses.

make3 9y ago

the formal complaints of all browser people seem to support the WHATWG

ttepasse 9y ago
The browser people are the WHATWG. There are no other stakeholders in that.
- domenicd 9y ago
  Not sure exactly what your comment is saying, but at least one reading is that "browsers are the only stakeholders in the WHATWG", which is not accurate. The WHATWG is a community organization open to participation by all; see https://whatwg.org/faq#process https://whatwg.org/faq#process for more. We receive a lot of participation from users, web developers, and other companies. There is a formal group, the Steering Group, which represents the browsers that implement WHATWG standards, and serves as the point of final appeal if the community doesn't come to a consensus on its own. But this is similar to the W3C appeals track where if all the paying member companies don't come to consensus, they appeal to "The Director" as the ultimate decider. ("The Director" is nominally Tim Berners-Lee, but recently all Director decisions have been made by W3C management "on behalf of" The Director.)
  - hungryfoolish 9y ago
    Are there any non-browser members as part of the editors or steering committee in the whatwg? To me, this is giving browser makers even more power over HTML than they have otherwise.
    - domenicd 9y ago
      The MIME Sniffing, Streams, Console, and Quirks Mode Standards are all edited by people who are not working for the browser-engine-developer companies. (Streams is co-edited by Googlers as well.) That's 4 out of the 15 standards currently developed at the WHATWG; not so bad, given how few companies are willing to pay people to work full time on web standards. Of course, we have lots of work to do, and if you or anyone else are able to devote a good chunk of time to standards work, we'd love to have more editors---no matter what company they work for. Browser engine developers comprise the steering group, since ultimately we want the specs to reflect what will be implemented in web browsers, so the best way to resolve disputes about what should be in the spec is by asking the people who will be spending their resources to implement it what they think. I think this is probably the right way to go, instead of "The Director" having the final voice.
      - hungryfoolish 9y ago
        > That's 4 out of the 15 standards currently developed at the WHATWG; not so bad, given how few companies are willing to pay people to work full time on web standards. Thats actually nice to know, but specifically about HTML (which is one of the most important specs the whatwg works on) it's all browser makers. (That too, majority of them from one particular browser maker). >I think this is probably the right way to go, instead of "The Director" having the final voice. I think having a better mix of people in the decision making process than just browser makers would be the right way to go. I understand that browser makers have to implement the specs, but the web community as a whole has to use the specs to build the web of the future - and as such, people from non-browser companies should have a greater say in the final voice.
        9y ago
- dragonwriter 9y ago
  Not only are there non-browser-maker WHATWG stakeholders, there are non-browser-maker formal objectors to the W3C DOM move to CR endorsing the same view as the four browser makers (Bloomberg and Disruptive Innovations; I've never heard of the latter previously, AFAIK.)
  - ttepasse 9y ago
    I was curious so I googled a little bit: Bloomberg is know for its Terminal. It seems they have a program where you can extend their terminal with web applications. Presumably Bloomberg didn't built a renderer themselves but there are custom APIs. So they are something between a browser maker and a UI framework like Electron. Disruptive Innovations is a firm by Daniel Glazman, formerly of Netscape and of the CSSWG fame. His main product seem to be continuing the Editor part of the original Mozilla Suite based on the Gecko Layout Engine. There seems to be an NVU Editor and rather new an Editor called BlueGriffon which also does ePub and maybe his custom WebBooks format.

domenicd 9y ago

This information about how the WHATWG works is somewhat outdated; please see https://whatwg.org/faq https://whatwg.org/faq for more.

lambda 9y ago

I'm pretty sure Ian Hickson is no longer involved with the WHATWG. I think he's working on Flutter now (https://news.ycombinator.com/threads?id=Hixie https://news.ycombinator.com/threads?id=Hixie). In fact, I would say that it's not quite like a cathedral and bazaar here. The W3C is like a legislative committee. Lots of things done by committee vote. Formal process that must be gone through to advance beyond committee. Petty politics that screw with actually getting things done. The committee chair having the ability to block work by procedural means. The WHATWG is more like an open forum. The work happens out in the open. Anyone can participate in the discussion (the W3C committees frequently have private meetings, there have been private mailing lists, and so on). There are a few people who have commit rights to the repo, so who actually control what goes in and what doesn't, but they are generally willing to let in changes that have broad support, and are implementable, rather than letting such features get held up by political processes. Besides the difference in structure, there's just a difference in attitudes. The W3C tends to strongly favor certain principles, like accessibility and modularity, but to the exclusion of compromise for technical reasons or real-world reasons. They also seem to have a tendency to get very attached to particular ways of doing things, without being willing to compromise. I think the biggest example of this was longdesc; it was never implemented properly in pretty much any browser, and very few people actually followed the spec and had it point to a URL (many people just copied the alt tag, or provided a longer description in the attribute instead of a URL), so even if browsers or screen readers had implemented it, the content wouldn't be useful. But people in the W3C made a big stink about removing this, and spent a lot of time and effort fighting and litigating over that, rather than actually trying to work on a different feature that could gain wider adoption. The WHATWG tends to take a more pragmatic approach; pave the cowpaths, if there are differences between implementations do it in the most sensible way that preserved compatibility. Now, what the WHATWG has produced hasn't always been the best; there are times when it's made mistakes in its approach. The drag and drop spec, which had been reverse-engineered from IE's drag and drop support, was a pretty bad; not sure if it's gotten better since. But overall, the WHATWG has been a lot more productive in getting standards done that are actually used on the web, because they involve the implementers, and don't override them with tedious, drawn out, political battles over obscure features that no one uses.