It’s just data

Nebulous Recalcitrance

Brendan Eich: The small-is-beautiful generalization alternates with don’t-break-the-web, again without specifics in reply to specific demonstrations of compatibility.

It is interesting how the don’t-break-the-web meme means different things to different organizations: Mozilla, Microsoft.


From the discussion in the HTML WG mailing list, on blogs and elsewhere, my impression is that Microsoft’s “Don’t break the web” vision really means “Don’t force us to fix our buggy implementation” while for Mozilla it means “Don’t make web site authors fix their buggy implementations”.

I don’t think I need to explicitly express which vision I’m most compatible with.

Posted by Asbjørn Ulsberg at

It seems to me that Microsoft has declared war on the web but everyone’s being very softly, softly about it just in case they come round at the last minute. Is there a backup plan if they just stonewall change forever? I’ve read of Java replacements for the video tag (anyone written one in Silverlight?), javascript replacements for HTML forms, the ability to swap out IE’s javascript with a Tamarin-based one etc. but never seen all the strands drawn together to chart a pragmatic way forward without Microsoft’s collaboration or consent.

With the surprise success of Mozilla and Mac OS X/Safari on the desktop, and the coming of age of the real web on handheld devices like iPhone and via Opera, and the neatness of unobtrusive javascript for patching IE there’s probably never been a better time to be aggressive about this and take the battle to them.

Posted by Dave at

[M]y impression is that Microsoft’s “Don’t break the web” vision really means “Don’t force us to fix our buggy implementation” while for Mozilla it means “Don’t make web site authors fix their buggy implementations”.

Well, in fairness, the two kind of go hand in hand, don’t they?  Lots of web site authors' “buggy implementations” are made to work around the “buggy implementations” of browsers...

Posted by Jason Lefkowitz at

the two kind of go hand in hand

Agreed.  The problem is elsewhere.

JS.Net goodSilverlight goodES4 and ScreamingMonkey are bad.

The issue is cultural.

Posted by Sam Ruby at

ES4

I’ve been wondering what was going to be keeping me entertained now that the OOXML stuff has died down a bit. Somehow, Microsoft never fails to disappoint when it comes to causing trouble in the standards community. Of course, in this case, Msft is...

Excerpt from snellspace.com at

We Can't Afford Another Browser War

It was with a sense of foreboding that I read the posts that swam past on Planet Intertwingly today. First came Mozilla’s Brendan Eich’s chastizement of Microsoft’s Chris Wilson, followed in a short while by commentary by Sam Ruby, where he wrote:...

Excerpt from Burningbird at

Browser War? Hardly.

Sorry, Scoble. Even if Microsoft never implements a word of ECMAScript 4, things will not get worse for web developers. That’s because no one wants to change the way ECMAScript 3 (aka JavaScript or JScript) works. Everything that works now...

Excerpt from Rob Sayre's Mozilla Blog at

Browser War? Hardly.

Sorry, Scoble. Even if Microsoft never implements a word of ECMAScript 4, things will not get worse for web developers. That’s because no one wants to change the way ECMAScript 3 (aka JavaScript or JScript) works. Everything that works now...

Excerpt from Get Latest Mozilla Firefox Browsers at

or this one

“Sorry, but most of those JScript deviations are not candidate de facto standards — they are just your bugs to fix. They should not stall ES4 for one second.”

when said about HTML and HTTP, the principle do not break the Web is pulled out. Very flexible principle, it seems depending on the circumstances. I hope I will understand a bit more the position of each people by seeing them face to face next week.

Posted by Karl Dubost, W3C at

Karl, most of the JScript bugs detailed by section 2 in Pratap Lakshman’s document are not in other browsers. For example, IE JScript violates ECMA-262 Edition 3 Section 13, the production:

The production FunctionExpression : function Identifier ( FormalParameterListopt ) { FunctionBody }

whose semantics require the Identifier naming the expressed function to be bound in a new Object scoping the function, by binding Identifier in the variable object of the execution context! In other words, in global code ‘var f = function g(){}’ binds a global property g. This is just flat wrong, a destructive side effect, and polluting. No other browser does it, and no cross-browser web content depends on it. (Pratap’s Safari results would be the same as other non-IE browsers if he used a try-catch, or enabled the Safari debug console).

This is just one example.

The same goes for many over-specifying descriptions in section 3 of Pratap’s document (which is quite informative; we’re grateful to Pratap for doing the work to investigate and write it, although as you can see from the JScript blog, web developers bitterly note that it’s years late and many dollars short). ECMA-262 must specify much more than the C or Scheme standards do, to support the scale of interoperation that JS has achieved (unique among programming languages, as far as I know). However, certain intentional under-specifications are desirable (e.g., sort algorithm and stability). Complete specification is not required for interoperation in all cases.

Please tell me why you apply the “don’t break the web” principle to IE JScript bugs that occur in no other JS engine? That is not the definition those of us building interoperable browsers, and working in the WHAT-WG and Ecma, use. Our definition applies only to shared quirks that cross-browser web content depends upon.

I think scoring my use of the “don’t break the web” principle off as “very flexible” is a cheap shot, given this vital distinction.

HTML and HTTP quirks reverse-engineered in the same way across browsers are likewise a different matter from IE-only HTML and HTTP quirks that have not mattered enough to force matching reverse-engineered behavior in other browsers, generally speaking. The challenge of evolving the web compatibly is impossible if one has to match every single behavior of IE, since only the IE source code compiled appropriately can do that.

I take it you are attending the TG1 face-to-face next week. I do not see you among the attendees yet. Could you please mail me about which Ecma member you are representing, and I can help get you wiki access? Thanks.

/be

Posted by Brendan Eich at

I no doubt misunderstood Karl’s “face to face” reference to apply to the Ecma TC39-TG1 meeting next week, when he almost certainly was referring to the W3C steering committee meeting (Mozilla has reps at both meetings). Sorry about the confusion.

Karl, I will refer David Baron to Sam’s blog so he knows about our little exchange :-/.

/be

Posted by Brendan Eich at

Brendan,

Thanks. I have already discussed in the past with David Baron, which I prefer to interact face to face with. He is a gentleman in physical meeting. I was talking about F2F in Boston during the W3C Technical Plenary. I guess then you will not be there being at TG1. Sorry to have not been clear.

Posted by karl dubost, w3c at

Primeiros tiros nas segunda guerra mundial dos navegadores?

E parece que a coisa está esquentando mesmo. Depois dos primeiros dois textos pelo Chris Wilson (Microsoft) é a vez de Brendan Eich responder com uma carta aberta consideravelmente pesada onde ele acusa Wilson de estar propagando falsidades sobre os...

Excerpt from Superfície Reflexiva at

Primeiros tiros nas segunda guerra mundial dos navegadores?

E parece que a coisa está esquentando mesmo. Depois dos primeiros dois textos pelo Chris Wilson (Microsoft) é a vez de Brendan Eich responder com uma carta aberta consideravelmente pesada onde ele acusa Wilson de estar propagando falsidades sobre os...

Excerpt from Superfície Reflexiva at

Free Culture: Why buy the Amazon Kindle when you can give

and get an OLPC XO-1 for the same price? I just discovered Kindle: Amazon’s New Wireless Reading Device . About $10 per e-book sounds ok, but $0.10 to put my own files on it?!?! It can read blogs like Slashdot and boingboing for as little as $.99 per month over the $399 purchase price. It...

Excerpt from Planet RDF at

We Can't Afford Another Browser War

It was with a sense of foreboding that I read the posts that swam past on Planet Intertwingly today. First came Mozilla’s Brendan Eich’s chastisement of Microsoft’s Chris Wilson , followed in a short while by commentary by Sam Ruby , where he...

Excerpt from Bb RealTech at

Brendan's Roadmap Updates: Open letter to Chris Wilson

Brendan’s Roadmap Updates: Open letter to Chris Wilson by François Hodierne & 1 other (via) microsoft javascript mozilla [link]...

Excerpt from Public marks with tags javascript & microsoft at

Add your comment