Our First Weekend Discussion
Friday February 11th, 2000
I've decided to try something new at MozillaZine: a "weekend discussion". Every Friday, we'll have a forum on a topic of interest to Mozilla users and developers. Hopefully it will generate some interesting conversations.
This week's topic is "Web Standards". What are your feelings about the current standards "process"? About Mozilla and its standards support? About what standards Mozilla should support in the future? I'm also interested in getting some opinions from Mozilla developers on this issue. How do they feel about being the "guinea pigs" of the standards process, being the first to implement standards specifications and also being the ones to deal with the vagaries and inconsistencies between revisions in the current specs? How are developers feeling about Mozilla's standards support so far?
Our forums are quite civil, so don't be afraid to speak your mind. We'd love to hear what you have to say. Just click the "responses" link below to get started.
#1 Thoughts on changing standards...
Friday February 11th, 2000 10:39 AM
It it inevitable that standards will change. CSS seems to be the big thing right now, but it may come and go. The best thing that Mozilla can do is remain modular enough that it will be easy to add support for the next standard, whatever that may be.
#2 I love the standards, just want bug-free
Friday February 11th, 2000 11:19 AM
The nice thing about the standards is a write once run anyware ability. I've created a fully HTML4.0 CSS2 page that looks great (minus a few very minor bugs) under Mozilla. Every other browser except for the text-only version make it look ugly (Opera doesn't do too bad.)
The ability for me to set up a page that I can easily change the style layout, whatever is what I want most. I've built half of a database backed web page, and now I am writing the input section. My input page will rely on DOM to do much of the correctness checking so I can submit it straight to the database with no problems. The DOM works great for this, except for one bug in extracting the value from a for input.
Enough rambling. The best part about standards is that you can look in one place for all of the information you need. If I need to know how to change the colors of divs I go to w3.org; for proper syntax in HTML, w3.org; for moving my page to XHTML or XML, w3.org and for DOM methods, again, w3.org. Everything I need to know about writing whatever web application I want is within reach, rather than having to rummage through search engines trying to find the specs for IE's DOM or NS4's DOM or whatever. With a standards body, everything is exactly where I need it.
#3 The problem with standards
Friday February 11th, 2000 11:38 AM
The problem with standards is they take so long to get finished, and they are designed by committee. The particular flaw with the W3 standards is that they come with no reference implementation. (Contrasted with IETF standards which usually have a decent implementation). Since there is no reference implementation, it is up to the browser makers to interpret the standards in edge cases the standards writers didn't even think of. The incompatibilities between CSS1 and CSS2 are an example of this. Particularly with font mapping (see the CSS2 notes for details). The W3C sits up on high and decrees commandments for authors and browsers to follow, but, does not do very much to help the browser makers or authors actually implement them.
This is changing a bit with the latest standards. CSS2 has some detailed discussion and pseudo code, as does parts of the DOM spec. But we need something like Amaya (or a version of Amaya which is stable and full featured) as a reference.
#4 Important missing standards
Friday February 11th, 2000 11:48 AM
While mozilla's standards support is great, there are a few important standards that are still missing. Each of these is, in my opinion, vital to the future of the web, and I hope that serious work is begun on these as soon as possible after NS ship their first branded browser. I'd love to see an NS5.1 within a few months that added support for them.
XHTML (shouldn't be too hard since it's basically HTML4.0 expressed as XML, but I think it's vital to follow it to the letter)
CSS2/CSS3 (CSS2 has been a w3c recommendation for quite some time now... partial support seemed good back when NS first promised it, but as time passes FULL support will be more and more important)
XSLT - Please! This is a biggie. This is a "killer feature" in the latest IEs (even though it doesn't fully support the standard) and it's a big area where IE can still boast better "standards support"
Real PNG alpha - don't know what happened to the bug on this, but it didn't seem to be considered a terribly high priority. I'd like to see this in 5.0.
MNG - the last barrier to getting rid of GIFs on the web. Browsers need to support this *soon*.
MathML - the support is almost completely written already it seems. I hope it will be included in the branded 5.0.
Remember, I love mozilla. I think it is taking a HUGE leap in the right direction. But I really hope that these other standards get incorporated sooner rather than later.
#5 W3C and Standards
Friday February 11th, 2000 12:09 PM
I think the W3C should be more open to outside input, using something like Bugzilla for people to suggest changes or new ideas.
I also think the term "Layer" is more intuitive than "Div" and standards bodies should consider things like that.
I think sdm is right about the reference implementations.
The specifications, for setting colors for the body of the document, confused me. It seems that 2-digit hexadecimal values (RRGGBB) are used for text and borders but 1-digit background colors (RGB) are used for the background. How is it possible to specify a 24-bit background color with 1-digit hexadecimals?
#7 Re: W3C and Standards
Friday February 11th, 2000 2:01 PM
I disagree that layer is more intuitive than div. Remember that html markup is (or should be, at least) based on meaning rather than appearance. A div is a logical division in the document that _might_ be layered on top of something else, but then again it might not. It could even be spoken by an aural browser. You misunderstand the purpose of div if you think of it as nothing more than a replacement for layer. What's more, layer belongs to that class of html tags and attributes which includes font and bgcolor, which have quite rightly been deprecated in html 4.
On the other hand, there are plenty of css rules which do appear to be needlessly unintuitive, and I suspect, detrimental to its adoption.
#6 A possible solution to the W3C problem?
Friday February 11th, 2000 1:35 PM
That the W3C is a very closed and secretive organisation is a common complaint, not to mention the costs of joining. I'm not too knowledgable about the actual W3C processes but in general I'd agree that an open system is preferable. But people have been complaining for years about the W3C and it shows no sign of changing, so I'd suggest that the only way to do something about it would be to set up another standards writing body to write standards in the same area. It's just occured to me that Mozilla might be a useful or even necessary part of such an effort.
I'm writing as the thoughts come to me at the moment so I apologise in advance if there are any gaping holes in my logic.
In order to create a new set of standards it would be necessary to overcome an enormous amount of inertia that is keeping people with the W3C's work. To do this there would need to be either some serious backing from one or more large companies, or a very good set of standards with a very good reference implementation. Since Mozilla already has an advanced rendering engine, the source code of which is accessible, it might be very useful for anyone wishing to create or experiment with proposed standards in this area. A proposed standard with a Mozilla based reference implementation would be far more impressive and far more likely to be adopted than just a proposed standard on its own.
Does that make any sense? Is there anything I've missed?
#8 Re: A possible solution to the W3C problem?
Friday February 11th, 2000 2:34 PM
At this point, most of what you typed makes sense to me. The only part I am very skeptical about is the "enormous inertia" of the W3C. It seems to me that very few people paid any attention to the W3C until Mozilla agreed to follow their recommendations. Therefore, maybe AOL has the power, not the W3C. If that is the case then why is an "open" movement like Mozilla following a closed organization like the W3C?
Are you capable of initiating a standards body like the one you mentioned? If you try, I will probably support you. Of course I am not a large company or even a small one. I suspect the movement would be crushed like a Microsoft competitor but it would be valiant. :)
#14 Re: Re: A possible solution to the W3C problem?
Friday February 11th, 2000 5:15 PM
hmm... You have a point there. However, to the extent that they're willing to follow anyone, I think browser vendors are more inclined to follow the W3C than any open source upstarts.
I don't think it's quite true that people haven't followed the W3C though. After all, CSS and XML didn't even have any existing implementations when they were created, yet they have, to some extent, been adopted. The W3C barely managed it and I don't think it'd be possible for anyone with less clout than the W3C to pull that off.
I'm actually quite in favour of Mozilla following the W3C's recommendations at this stage. Though I suspect their methods to be non-optimal, I'd rather have them than have MS defining the standards. If Mozilla didn't follow the standards then neither would MS, and MS would win.
As for setting up a standards body myself, the answer is no. I'm nothing more than an interested observer and I don't have the experience or the influence. Besides which, the first thing necessary in creating a new web technology is an idea, and I haven't got one of those either. (I don't think it would be possible, or even desirable, to take over the development of an already existing W3C standard.)
#30 Re: A possible solution to the W3C problem?
Saturday February 12th, 2000 7:01 AM
NO! The whole reason Mozilla is going standards compliant is because working with two sets of (often incompatable) standards just doesn't work.
Before it was Microsoft & Netscape, what makes you think that W3C & AOL/Moz would be any better?
#32 Re: Re: A possible solution to the W3C problem?
Saturday February 12th, 2000 9:47 AM
The W3C should not be the only "standards body" unless it becomes an open "standards body". If there must be only one then it should be a better group than the W3C.
#33 Re: Re: A possible solution to the W3C problem?
Saturday February 12th, 2000 10:25 AM
I quite agree that a separate set of incompatible standards would be rather unpleasant, which is why I wouldn't advocate another group trying to take over existing standards such as html, css or xml. That would most probably result in forking, which would defeat the purpose of standards. But, where a new technology is created, it would make sense for the creators to make a reference implementation in Mozilla and set up a new body to maintain and promote the standard rather than proposing it to the W3C.
I don't know what such a new technology would look like. If it were too similar to a pre-existing standard then that might cause problems if they came into competition. However, if such a thing were to happen I wouldn't place any bets on the newcomer winning.
#9 Make Mozilla the "Reference Implementation&qu
Friday February 11th, 2000 3:07 PM
Mozilla is the most open browser project currently in development. Its newsgroups, website, daily builds, and bug reporting system make it the perfect candidate to be the "reference implementation" that the W3C needs sorely needs. Currently, developers have to battle with conflicting standards and vague wording in specifications. The CSS3 spec is currently in development, but there is no complete implementation of *CSS1* yet. The process chugs along, unwilling to stop to let the industry catch up.
Since the Mozilla development occurs in the open and crosses platform and operating system lines, it seems like it would be the ideal solution to this issue of the lack of a reference implementation. What do you think?
#10 Re: Make Mozilla the "Reference Implementatio
Friday February 11th, 2000 3:54 PM
Mozilla - Standards In Action. I like it. One thing that Mozilla has already done is to bring standards to the table. Other companies, even Microsoft, have taken note that being standards compliant is a marketable trait (see thier mac ie port). This is good. Now to take it one step further as Chris suggested would be grand. If Mozilla can get some kind of stamp of approval from the W3C as a reference implimentation then others will also seek this stamp and before you know it we have a happy world where all kinds of different platforms running all sorts of browsers can present content to users in a consistent and powerful way. I am excited by this potential and praise those who have worked hard to help realize it.
-Asa (posted with today's build which has links in the sidebar working again)
#11 Re: Make Mozilla the "Reference Implementatio
Friday February 11th, 2000 3:57 PM
I think, because Mozilla was formed by Netscape to create Netscape 5, making it the "reference implementation" for Web standards is like making Windows 2000 the "reference implementation" for operating systems.
If the W3C recommendations become actual standards and the implementation of these standards is based on Mozilla then people would be forced to comply with those standards, and therefore forced to comply with Mozilla. It would be Microsoft part II.
#12 Why exactly is that?
Friday February 11th, 2000 4:20 PM
Just because Netscape uses the Mozilla code for Netscape 5 does not make Mozilla less worthy of becoming the reference implementation. It is, in fact, the only browser that could *possibly* become a reference implementation at this point, because of its well-supported open development process.
Netscape unleashed the source code. But the source code that we have today shares very little in common with Communicator 4.x. Mozilla has grown in an open environment, and it is ideally suited to the task of becoming the reference implementation.
The only other option would be to take the Mozilla code and create an entirely new development tree, which would be a huge waste of effort. Or, they could see if they could get public support for the Amaya browser. I don't see that happening. It certainly hasn't so far.
No one is forced to comply with the standards (see today's current situation for proof of that). It could not possibly become "Microsoft II", because the public would be allowed to comment on and correct the Mozilla implementation. If there were problems with Mozilla's code, those problems would have to be fixed (just as they would in any other browser).
It might give Mozilla a home court advantage, so to speak, but they already have that, IMO. They're the only group out there putting their code on the line. If MS opened up their development process, then we would have two possibilities for "reference implementation". I don't see that happening, either.
#13 Re: Why exactly is that?
Friday February 11th, 2000 5:02 PM
I believe mozilla would be a good browser to use to test the standards and show how they should be implemented. I did not say it was "less worthy". I feel misquoted.
My thought was that Mozilla is going to be more associated with Netscape than anything else, which would promote Netscape due to the "credibility" of the standards organization. That would give Netscape "de facto" credibility by association even if it was not the only standards-compliant browser.
Amaya? It could be a good "reference implementation" if they would get it to work. Has there been any progress? For instance, does it properly display Webpages made after 1990 yet?
I think a standards organization should be neutral, not aligned with a particular commercial product. Windows is the operating system, which, in an ideal situation, would be neutral and not aligned with a particular brand of software. Instead, Windows promotes the use of Internet Explorer specifically. If Mozilla were to suggest that people test their Webpages using Mozilla then they would be promoting the use of a particular COMMERCIAL product. That is why I associated this situation with the Microsoft situation, and referred to that situation as "Microsoft part II".
<defensive>My MozillaZine posts were innocent statements of my views on the MozillaZine topics, which is what I believe this forum calls for. I believe you offered a suggestion and asked people for comments on that suggestion. I believe I respectfully answered.
If any people or <loud tone=evil>gerbils</loud> feel that I am somehow being irrational, please explain this to me.</defensive>
#17 Re: Re: Why exactly is that?
Friday February 11th, 2000 6:14 PM
--If Mozilla were to suggest that people test their Webpages using Mozilla then they would be promoting the use of a particular COMMERCIAL product.--
I disagree. Mozilla is not a COMMERCIAL product. Mozilla is not a COMMERCIAL product. Mozilla is not a COMMERCIAL product. If people test their webpages with mozilla it advances Mozilla and _any_ browser based on Mozilla. Netscape is not Mozilla. Many other commercial entities have and will continue to adopt Mozilla as a browsing solution. Testing web pages against Mozilla will advance Mozilla and ALL of the projects that use mozilla or that adhere to the standards.
--I think a standards organization should be neutral, not aligned with a particular commercial product.--
Mozilla is not a COMMERCIAL product. Mozilla is not a COMMERCIAL product. Mozilla is not a COMMERCIAL product. Standards organizations should be neutral. But they _should_ also be alligned with commercial products, lots of them. The more commercial products that work in a standards based way, the better. Isn't that the point. So the more commercial (or non-commercial)products based on Mozilla or based on the same standard, the better off we all are. If Microsoft comes along and offers up a browser that is stndards compliant then it will be a good model for other products, a good reference. The thing that makes Mozilla a better reference platform than even a standards compliant Microsoft browser is that Mozilla is open. Everything is open to scrutiny and debate. This is the definition of neutral.
Maybe I'm misunderstanding. If so please explain more about why an open source non-commercial project being developed by people from all over the world should not be considered as a good spec on which to base other standards complaint products.
#20 Re: Re: Re: Why exactly is that?
Friday February 11th, 2000 6:36 PM
Open Source is not mutually exclusive with commercial. Free Software is not mutually exclusive with commercial. AOL will be a hunk of cash off of mozilla. They make tons of money from advertising based on the browser and searches people do with it. If they were not standing to make money off of this product, they would not be funding it. That's FINE. The point is, the W3C does not endorse ANY reference implementations of ANY standard. That's not what they do. So mozilla can not be the reference.
#21 Exactly the problem
Friday February 11th, 2000 7:24 PM
They don't have a reference implementation. And that's made all the difference in the world, IMO, on the advancement of standards compliance. It's time for them to deal with whatever issues they have regarding reference implementations. If they don't we're doomed to continue on as we have been.
#24 Re: Re: Re: Why exactly is that?
Friday February 11th, 2000 9:35 PM
Saturday February 12th, 2000 8:37 AM
That means that both Internet Explorer and Opera in addition to Mozilla and Netscape are promoting Netscape, because they've ALL been reporting userAgent strings of Mozilla x.x for years.
#28 Netscapes is still the boss
Friday February 11th, 2000 10:27 PM
These there simple questions will lead us to why Mozilla should not be the defacto standard...
1. Who released the source? 2. Who still controls the mozilla development process? 3. Who will gain from the W3C being aligned with Mozilla.
And the awnser is Netscape!
Netscape would adverise and push that fact to death that THEY are the defacto standards implementation and that everyone should base their code off of the Mozilla project. Netscape could not get that kind of advertising power even if they had Bill Gates pocket book. Because while mozilla is still an open source project Netscape is still the one pulling all the strings and I have no dought in my mind that Netscape would take all the credit if this happened. Thats not to say that mozilla so far isn't the best standards implemtation out there, its just that even though it is not a commercal product it is still very, very, heavily related and tied to a commercal company.
#29 Re: Netscapes is still the boss
Saturday February 12th, 2000 6:51 AM
I totally agree with this. Yes, Mozilla is the most standards compliant browser out there. This is what I think needs to happen:
1) The W3C opened up, with the public discussing and debating issues with HTML, XML and related things. 2) The W3C creating an independant reference implementation. I can't see why it can't use Mozilla / Gecko as a base to start from. Or parts of it at least. And keeping this reference up to date. 3) Easier to understand specs. The recommendations are very detailed, geared more towards browser programmers than web page developers. The W3C should take a look at Microsoft MSDN or Netscape DevEdge and create easy to understand versions of the specs. 4) A W3C-compliant logo. Like Intel Inside, it can be placed on products which pass a W3C compliancy test. 5) This contradicts number 1, but if a company joins the W3C, they should be forced to follow the standards. It should be in the contract.
#15 Re: Make Mozilla the "Reference Implementatio
Friday February 11th, 2000 5:33 PM
I don't think that Mozilla has to be *the* reference implementation, and I'd be slightly uncomfortable if it was. Html in particular, and css to some extent, are standards which allow a multitude of different implementations. In the case of html this is definately a Good Thing.
I would be very happy for the Mozilla code to be used as a testbed for new technologies, and I'm happy for people to point to Mozilla when they want to give an example of something specific that it does right. But a single reference implementation, whichever browser it was based on, for many of the W3C's standards would be undesirable.
#18 Off topic: Netscape 6 ?!?!
Friday February 11th, 2000 6:20 PM
Our friends at CNET have this lovely article about Time/Warner/AOL/Netscape/we-own-everything's plans for netscape 6. Yes 6. They have decided to take the Micro$oft number game to the next level. Anyway, my two favorite lines from the article were 1.) "... the most notable difference is the way AOL said it will distribute the product. The company plans to allow Web sites to launch their own branded versions of the browser..." (no mention of the completely new code base, standards compliance, open source, etc.) 2.) "Until now, Netscape has relied partly on its open-source development organization, Mozilla.org, to speed development of its Communicator browser, with volunteers and paid developers refining the product. But despite Mozilla's efforts, AOL has acknowledged that it has been late in updating its browser technology." What do you mean, "until now"?!?! "partly" ?!?! <a href="http://technews.netscape.com/news/0-1005-200-1548169.html">read it for yourself</a>
#19 screwed up the link
Friday February 11th, 2000 6:21 PM
http://technews.netscape.com/news/0-1005-200-1548169.html read it for yourself
#22 Re: Off topic: Netscape 6 ?!?!
Friday February 11th, 2000 7:52 PM
My guess would be that the author got his facts wrong, and is /really/ talking about version 6.0 of AOL's client software (which is probably still based on IE).
#23 Re: Re: Off topic: Netscape 6 ?!?!
Friday February 11th, 2000 7:57 PM
A lot of stuff in the article is vague, so I won't be surprised if the facts in it are completely screwed.
#25 Re: Re: Off topic: Netscape 6 ?!?!
Friday February 11th, 2000 9:43 PM
It took me hours to recover from installing AOL 5. Why must there be another one?
#27 Standards: The webs blackhole of Cofusion
Friday February 11th, 2000 9:59 PM
Well as far as the standards "process" goes, I think its horrible. Now this isn't to say that the people that define the standards are unintelligent, but its that only x number of people are allowed to define them if they pay a huge sum of money and then the standards making process is entirely secretive. I mean come on, you'd think they were designing secret military weapons of mass destruction. And then the fact that almost ONLY big businesses have any outside influence. Now you can't tell me that all those businesses don't have all their secret and alterer motives to have the standards conform to whatever new product, etc, they are coming out with. And then I saw that they TradeMarked the word XHTML and I almost spit my pop all over my monitor. Now when the heck did the W3C think it became a big business... I can imagine 5 years down the road Bill Gates giving an anonymous sum of money to the W3C for ths new standard MSHTML.
But they define a standard behind closed doors then one day decide its a spec and then they throw it out to the public with few example and almost no reference for the implementation. And then phew Browser's makers and developers are expected to pick it up and use it like it was the best thing since sliced bread....
Anyway I'm rambling..., -DHTML Fiend
#34 W3C better than nothing
Saturday February 12th, 2000 1:53 PM
I'm encouraged by the helpful observations I've read here, and I find I agree with many of them, such as: W3C is overly secretive and puts up too high a barrier for membership to be considered truly open. But perhaps that reflects the excessively political and adversarial nature of the struggle for control over standards. The DOM situation in the version-5 browsers, while hardly perfect, still represents a significant advance in cross-browser scriptability.
Mozilla is clearly way ahead of Microsoft in implementing some of the DOM Level 2 Event stuff. My experience in writing browser-forks for DOM 2 events to compensate for Microsoft's neglect of Events shows me we're not out of the woods with our dueling DHTML woes.
#35 This Should B The W3C's Standard!
Sunday March 12th, 2000 2:03 AM
The Webpage Author's "Bill of Rights": 1."The Webpage Will B Rendered by The Renderer, The Way The Webpage Was Orginally Written by The Webpage Author!"
An Compliance Should B Judged by The Ability of The Renderer (Mozilla?) 2 Render The Webpage Author's "Original" Mark-Up/Code as *IT* was Orginally Written by The Webpage Author, w/o Change!