Joel's Strategy
Joel Spolsky: What’s going to happen? The winners are going to do what worked at Bell Labs in 1978: build a programming language, like C, that’s portable and efficient. It should compile down to “native” code (native code being JavaScript and DOMs) with different backends for different target platforms, where the compiler writers obsess about performance so you don’t have to. It’ll have all the same performance as native JavaScript with full access to the DOM in a consistent fashion, and it’ll compile down to IE native and Firefox native portably and automatically. And, yes, it’ll go into your CSS and muck around with it in some frightening but provably-correct way so you never have to think about CSS incompatibilities ever again. Ever. Oh joyous day that will be.
I can’t help but wonder why Joel and others didn't make greater use of Dean Edward’s IE7 at the time. Possible reasons:
- Joel and others hadn’t heard of Dean’s efforts at the time, for whatever reason.
- Browser incompatibility wasn’t so much of an issue in 2004-2005. One way to look at this is that IE’s dominance at the time was substantial enough that developers could get away with requiring IE and the situation has changed enough that this merits revisiting.
- Dean’s efforts were in the wrong direction... people prefer the APIs as Microsoft implements them.
- Dean’s efforts were incomplete, or performed poorly, or was otherwise inadequate.
- Dean’s IE7 was an all or nothing proposition. It would modify the behavior of existing apps, which was unacceptable.
- Applications worked nearly enough without the download that enough people wouldn’t make the effort, and the net result would be one additional platform to support.
- Dean’s efforts, while interesting, didn’t have enough corporate backing to justify a company to depend on it.
Possible solutions:
- Microsoft, Mozilla, and others work to eliminate their differences. Joel discounts this possibility, and given the progress at WHATWG and the W3C, it is hard to argue with Joel’s assessment. Microsoft has to walk a fine line between throwing its weight around and not showing up, and at the moment they seem to be consistently erring towards the latter. Having a solid beta / tech preview would address this, but there is no indication that this is likely any time soon.
- Something like Silverlight comes along (coupled with Moonlight). Such a solution doesn’t fix your CSS, it replaces many of the technologies that you are used to with new ones. There are no indications that such a solution will ever be deployed on an iPhone. Some have expressed concerns that once Silverlight gains enough tractions, the cooperation that Novell is currently seeing will dry up.
- Something like ScreamingMonkey comes along and solves the JavaScript incompatibilities. I see no indication that this plans to address DOM incompatibilities or CSS incompatibilities, either of which alone can be worse than JavaScript incompatibilities.
Overall, I can’t see anything in any of these solutions that is clearly better than reviving Dean’s efforts, but I would be interested in hearing Joel’s take on this.
Joel’s Strategy
I wonder if Joel is hedging Fog Creek as a potential winner? The thing I don’t get is, doesn’t Joel realize Google has already done what he described?Posted by Justin Watt at
Joel’s Strategy
Yeah I agree with you Justin, GWT is also a solution. I guess maybe Joel is saying we don’t know who will be the dominant framework yet (consider all the others like Dojo, Prototype, YUI, etc), but that’s kind of obvious.Posted by Jeff Schiller at
Joel’s Strategy
“where the compiler writers obsess about performance so you don’t have to”
Are you kidding me? Pre- and post- operators, weird pointer stuff out the wazzoo, macros. C was a portable assembly language. Heck, writing a compiler for it was almost like writing a super-high-end macro assembler. Of course you had to obsess over performance to use C. And the compiler writers barely had to obsess at all because C surfaced all the funky stuff the machines of the day had in their assembly languages.
I do see things like Silverlight and Flex redefining semantics within their own world. The Flex guys already tout that they deal with browser incompatibilities.
Cheers,
BW
Posted by Bob Warfield atJoel’s Strategy
Something like Silverlight comes along (coupled with Moonlight).
Hardly.
Any non-free software developed without standards-track specification (Silverlight, Moonlight, Flash etc) will never be the future of the web. Not my web, anyway.
It’s hard to imagine any serious web technologists suggesting these applications as an answer. If they are doing, I’m not sure I want to know the question.
Posted by Noah Slater atFlex/Flash Solve A Lot of Problems: They May Be Disintermediating the Browser
Joel Spolsky gave out a big yada yada post over at his blog that wound up saying the browser incompatibility has got to stop. He draws parallels with how portability won the day in programming languages. A lot of what he says is right, some of it...Excerpt from SmoothSpan Blog at
Joel’s Strategy
I’m partnering up with Noah. Neither of the mentioned solutions are good enough. No, not even GWT or OpenLaszlo. They are all either too proprietary, too closed or too inaccessible to be able to replace all the goodness that comes out of simple HTML+CSS+JavaScript based applications.
Both OpenLaszlo and GWT serializes to JavaScript mayhem, completely inaccessible by Google’s own search engine and anyone else who has a problem comprehending anything but simple HTML. The other ones require plugins and are very un-webby. That’s so un-cool.
I’m not saying that Joel’s dream web framework will never exist, I’m just saying it doesn’t today and none of the contenders are even remotely close to being it.
Posted by Asbjørn Ulsberg atMore reactions to Joel's strategy
In my last post about Joel’s ultimate javascript runtime predictions, I linked to Nikhil Kothari’s Script# project, which distills C# into javascript. This article comes now to my attention, about how ExtJS is thrown into the brew to extend the...Excerpt from Web-Program-athon at
Joel’s Strategy
When I read Joel’s piece, I can’t help but be left thinking his whole premise is wrong. Joel says,
The NewSDK will be the second coming of Microsoft Windows; this is exactly how Lotus lost control of the spreadsheet market.
To me, this seems to imply that “he who creates NewSDK will 'own the web'”, which is patently absurd.
Can software development and distribution from 1990 really be compared to today? Hasn’t the entire dynamic of competition and software choice changed?
I think we need to worry more about data silos and software patents then we do about some new framework or API getting a monopoly in the web-application space.
Posted by Kevin H atJoel’s Strategy
Why would Silverlight be any more feasible than Firefox with the WHATWG specifications implemented? (Silverlight, I’m sure, is more convenient than the fairly modest WHATWG specifications, but that cuts both ways.)
Silverlight isn’t more widely deployed than Firefox. I find it quite unlikely it will become more widely deployed. And the WHATWG specifications may be ignored by MS for a while (maybe a long while), but that might not even be a big deal; people can replicate the functionality in IE without MS’s help.
When placing bets, I’ll bet on the web every time. Most of the people working on the web don’t even know what we’re talking about here, and probably don’t care. The inertia is incredible, which strongly favors incremental improvements.
Posted by Ian Bicking atJoel’s Strategy
I started to read Joel’s piece yesterday. At first I was reading, but then he started talking about “NewSDK” as if SDKs actually mattered in the grand scheme of things. So then I switched to skimming, and the further along I got, the more hand-waving he did, and the less sense he made. So I stopped reading and went back to trying to be productive.Posted by Bob Aman at
Joel’s Strategy
I guess I would have saved a lot of time if I had simply pointed to this piece.Posted by Sam Ruby at
Joel Spolsky : The winners are going ... to compile to javascript and DOM
Joel Spolsky : What’s going to happen? The winners are going to do what worked at Bell Labs in 1978: build a programming language, like C, that’s portable and efficient. It should compile down to “native” code (native code being JavaScript and...Excerpt from onGWT - Tracking News on GWT at
links for 2007-09-21
WordPress database error: [You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'can’t see anything in any of these solutions that is clearly b] SELECT ad FROM...Excerpt from All in a days work... at
Sam Ruby: Joel's Strategy
[link] [more]...Excerpt from reddit.com: programming - newest submissions at
Joel Spolsky : What’s going to happen? The winners are going to do what worked at Bell Labs in 1978: build a programming language, like C, that’s portable and efficient. It should compile down to “native” code (native code being JavaScript and...
Excerpt from Tailrank: Posts for 'http://planetatom.net' at
[from connolly] Sam Ruby: Joel's Strategy
“Joel discounts this possibility, and given the progress at WHATWG and the W3C, it is hard to argue with Joel’s assessment.”...Excerpt from del.icio.us/network/myakura at
Rebuttals on Joel's Strategy - another: http://tinyurl.com/37gdup
I can’t help but wonder why Joel and others didn’t make greater use of Dean Edward’s IE7 at the time. Overall, I can’t see anything in any of these solutions that is clearly better than reviving Dean’s efforts, but I’d be interested in hearing......Excerpt from del.icio.us/billyg/screamingmonkey at
Joel’s Strategy
When I read this last night, I was wondering if Joel had ever heard of OpenLaszlo.Posted by Jeff Schiller at