Source: Martian Headsets
You’re about to see the mother of all flamewars on internet groups where web developers hang out.
This self-fulfilling prophecy has been brought to you by Google Adsense: funding Slashdot trolls since 2003.
With the web, there’s a bit of a problem: no way to test a web page against the standard, because there’s no reference implementation that guarantees that if it works, all the browsers work. This just doesn’t exist.
I demand documented standards with open reference implementations. That’s why I only develop with Microsoft technologies.
So you have to “test” in your own head, purely as a thought experiment, against a bunch of standards documents which you probably never read and couldn’t completely understand even if you did.
No one has ever written tools to encapsulate the world’s collective understanding of standards; you must start from scratch like everybody else. Freedom 0 buys you nothing unless you can audit the source code yourself. “Faith” in science is no more well-founded than “faith” in religion unless you can personally reproduce the experiments.
Those documents are super confusing.
Hi, I’m Web Developer Barbie. Pull my string and I say, “Standards are tough! Let’s go shopping!”
The precise problem here is that you’re pretending that there’s one standard, but since nobody has a way to test against the standard, it’s not a real standard.
I have never heard of test suites.
98% of the world will install IE8
I am high as a kite.
Secretly? Here’s what I think is going to happen.
Wait… you mean you can… see what I’m typing?
[Microsoft] will say, “look guys, we’re really sorry, we really wanted IE8 standards mode to be the default, but we can’t ship a browser that doesn’t work.”
Microsoft has always had the best interests of the web at heart. Microsoft has never encouraged web developers to use Microsoft-specific technologies. Microsoft has never shipped a browser that rendered pages differently than its predecessor and then let it stagnate for six years while the rest of the world moved on.
Or maybe they won’t… in which case, IE is going to lose a lot of market share.
I know IE is going to continue to lose a lot of market share, and I’m publishing this now so I can blame you dirty fucking hippies for it later.
(as always, with apologies to John Gruber)
Update: this comment about Microsoft’s (lack of) participation in HTML 5 may help you understand why I hold apologists like Joel in such low regard. Or it may not, but I don’t know how to explain it any better than I already have.
As I got about 20% into Spolksy’s line of horse manure, I was thinking, “Man, I want Pilgrim to skewer this.” Thanks, and add one to the tab if we ever meet.
Comment by Geof F. Morris — Tuesday, March 18, 2008 @ 4:44 pm
Do all “Translation to English” pieces have to contain “I am as high as a kite”? I’m not complaining - I actually like that little signature - I’m just trying to see if this is a hard-and-fast rule. And did Gruber come up with it? Should it be Gruber’s Law?
Comment by John Moltz — Tuesday, March 18, 2008 @ 4:57 pm
dirty fucking hippies
Waitaminute… your secret identity is - Atrios?
Comment by EWI — Tuesday, March 18, 2008 @ 5:10 pm
This kind of dissection of stupid commentary was originally created as a response to Robert Fisk diatribes that could not be left to stand as a steaming pile of mental crap without pointing out how absurd it was; hence this is called Fisking. Gruber didn’t invent it, although he prolly thinks he did (in his self-contained Left-brain world).
Comment by Neo — Tuesday, March 18, 2008 @ 5:10 pm
> hence this is called Fisking
Indeed. And view-source for an Easter egg of sorts.
Comment by Mark — Tuesday, March 18, 2008 @ 5:17 pm
Spolsky has a point. A test suite only becomes “official” in my eyes, when it is on the W3C server. Only one of your tests are hosted by them, and it’s for CSS1 :-)
There are several test suites out there, but even ACID3 is not official from what I can understand. It’s always good to have these tests so that you can test your implementation agains what others _think_ the results should be. But unless it is hosted by W3C, Microsoft will always find a way to ignore it.
Comment by Neutral — Tuesday, March 18, 2008 @ 6:28 pm
>Gruber didn’t invent it, although he prolly thinks he did (in his self-contained Left-brain world).
John Moltz was talking about the phrase “I am high as a kite” not the idea of fisking (which was, incidentally, coined following an Andrew Sullivan fisk of Fisk if the Wikipedia is to be believed).
Also, when you’re insulting someone wouldn’t it make more sense to use real words (”prolly”) and not improperly capitalize “left”? I get your poorly phrased point: “Gruber is a Democrat = deluded”, but what does that have to do with tech writing?
A better way to make fun of Gruber might be “he’s a pretentious hack, with delusions of grandeur, and so far up Apple’s a-s it looks like midnight”. I don’t believe that, but it makes more sense than attacking him for (presumably) being on the left side of American politics.
Comment by Wednesday — Tuesday, March 18, 2008 @ 6:33 pm
Wow, whenever somebody does this, out come the “fisking” trolls.
On the one hand, we have Robert Fisk, a well respected reporter who has worked in the Middle East for many, many years.
On the other hand, we have Andrew Sullivan, who hyped “The bell curve”.
I know who I think is more likely to produce a “steaming pile of mental crap”.
Comment by Erik — Tuesday, March 18, 2008 @ 6:47 pm
Left-brained:
1. Having the left brain dominant.
2. Of or relating to the thought processes, such as logic and calculation, generally associated with the left brain.
3. Of or relating to a person whose behavior is dominated by logic, analytical thinking, and verbal communication, rather than emotion and creativity.
Not a political comment at all.
Comment by Thursday — Tuesday, March 18, 2008 @ 6:47 pm
I got about a third of the way through his rambling, self-indulgent, whining post before I gave up. One of his worst posts ever (purely from a writing standpoint.)
Comment by Bob — Tuesday, March 18, 2008 @ 6:49 pm
Calling it “fisking” is ahistorical, anyway. That particular discursive structure is the hallmark of USENET flamewars.
Comment by Nick Caldwell — Tuesday, March 18, 2008 @ 6:52 pm
Joel often writes pretty insightful stuff. At least, the stuff I’ve read seems that way.
Then, last night, I read that post. I thought he’d been smoking crack.
Then I remembered that most of the rest of the world doesn’t have anything to do with web development, hasn’t been paying attention to how any of the IE 8 debacle came about, has no real idea what Microsoft’s original rationale was (or the first clue about how to determine whether it’s flawed, and from which viewpoint it might or might not be considered flawed), and is used to beta releases being Google-quality instead of “dear Christ, the web development community is screaming blue murder at us, we need to get something out”-quality.
By the time I’d reminisced about all of that, I’d long since lost the inclination to articulate it, and so I was rather glad somebody much better at it than I did instead.
Thank you :)
Comment by Mo — Tuesday, March 18, 2008 @ 6:57 pm
> originally created as a response to Robert Fisk diatribes
It was commonplace on Usenet at least a decade before anyone called it “Fisking”.
Comment by Fossil — Tuesday, March 18, 2008 @ 6:58 pm
I’m getting slow in my old age. Mark, feel free to delete these two comments.
Comment by Fossil — Tuesday, March 18, 2008 @ 7:00 pm
Precisely.
I’m a fan of Joel, I have his book about hiring and it’s brilliant. But the argument that for a company to implement an adequate level of standards compliance in a browser is impossible due to the complexity of the standards is patent rubbish, demonstrably so considering that three major browsers exist that have implemented those standards to a degree that has _already_ revolutionised the internet and moved us seismically towards an open future.
MS has seemed newly willing to implement web standards to a reasonable-ish level recently. This is not, as Joel stipulates, because idealists with no sense of what’s practical have won some kind of war against pragmatists within the halls of Microsoft. It is because that amazing progress towards interoperability through competent implementation of web standards, exemplified by Firefox and Safari, has and continues to *deeply threaten Microsoft’s dominance and ability to generate lock-in using Internet Explorer*.
We are in the wonderful position where MS has little choice but to implement these standards. This would not have become the case without the high-quality compliance we’ve seen in the re-invigorated browser market that so strikes fear into the hearts of the monopolist that stands to gain from the inhibition of such progress.
Comment by Benji XVI — Tuesday, March 18, 2008 @ 7:12 pm
To actually answer John Moltz’s question, since nobody else deigned to: yes. In Gruber’s first or second “Translation of selected portions…” post he translated one section as “I am high as a kite” and it’s been a running gag since.
It’s a funny piece, but I’m not sure I agree with the implicit accusation that Joel Spolsky is a Microsoft apologist. Having said that, I’ve always found the “standards are so tough” arguments bogus, whether in this incarnation, or in the various “XHTML considered harmful” rants from a few years ago (which could mostly be paraphrased as “invalid XHTML is still tag soup and since you’re all too idiotic to use a validator, don’t touch it, you morons”).
Comment by Watts — Tuesday, March 18, 2008 @ 7:13 pm
I loved Dive Into Python and am about to crack open Dive Into Accessibility, but I have to say that this vitriolic post (the first of yours I’ve read) has made a very poor impression on me for Dive Into Mark.
Comment by Thomas Allen — Tuesday, March 18, 2008 @ 7:22 pm
I’d check Joel’s bona fides before calling him out, and heck, I’d also respond to the whole article. Joel’s a very user-centric engineer — one of the few, and one of the best. He’s, correctly, pointing out the damned-if-you-do, damned-if-you-don’t situation that Microsoft is in.
Look, every web developer trying to make a buck is complicit in Microsoft’s lousy standards support. Because we all wanted some serious marketshare, we put in the hacks to make web pages work right on every browser. (at least every one we cared about) Now we’re pissed at Microsoft because of why? Because they want to support all the hard work we put into supporting their browser?
Sure, MS screwed up. But we let them, and we profited from their users, each and every one of which we wooed with our hacks and duct tape to ensure they got a Good Experience.
Well, if we want to serve that market, we still need to provide that Good Experience. If we can push standards a little harder, too, that’s just dandy. But in the meantime, I want the next website I build to work for everyone who wants to visit.
Comment by Nik — Tuesday, March 18, 2008 @ 7:23 pm
Thomas Allen, you must be new here. (apologies to Mark himself)
Comment by Justin Watt — Tuesday, March 18, 2008 @ 7:36 pm
Thomas Allen: Get used to it. While Mark is a really smart and articulate guy, it seems that anyone who doesn’t agree with him — about anything — is labeled a “fan boy”, or made fun of, or worse. There’s no room in Mark’s world for personal preference: you either do it his way, and like it, or you’re stupid (or at the very least, ill informed).
Mark will likely write this off as the pitiful whining of an ignorant loser, rather than consider — even for a moment — that maybe there’s something to this. Doesn’t matter, though. It needs to be said.
Comment by Anonymous Coward — Tuesday, March 18, 2008 @ 7:45 pm
I realized I put my email address in the Website input, which made me feel silly. So this is a correction.
Comment by Thomas Allen — Tuesday, March 18, 2008 @ 7:48 pm
> But in the meantime, I want the next website I build to work for everyone who wants to visit.
I think this is the flip side of what Stefano was talking about. Underlying your statement, I believe, is an assumption that there is One Way To Build A Website That Will Work Forever, and all you have to do is Find It and then you’ll be done. This assumption is false; it has always been false and it will remain false until the web is dead and gone. Standards are updated, browsers are updated, new standards are invented, entirely new browsers spring into being. (*cough* Safari 1.0 beta broke my site. *cough* It was entirely my fault. IIRC it was because I assumed that any browser with the word “Gecko” in the user-agent could handle application/xhtml+xml. Safari came out and proclaimed to be “like Gecko” but didn’t yet support the XHTML content-type. I switched to Accept-based negotiation and then, nanofamously, dropped XHTML altogether.) New authoring techniques are invented and popularized (XMLHttpRequest) and then browsers are updated to support them, or to support them better or more interoperably. And so on. Everything requires maintenance.
Comment by Mark — Tuesday, March 18, 2008 @ 7:56 pm
Ironically, I’m now echoing Joel’s main point (MANY to MANY, independent innovation). The only part he really got wrong was his conclusion — that because IE stagnated so long, we should carve its quirks into granite. Underlying *this* statement is the assumption that Microsoft’s position is somehow special, and, perhaps more subtly, that IE will be the majority browser forever. Neither assumption is true.
Comment by Mark — Tuesday, March 18, 2008 @ 8:00 pm
I love it when these supposed “programmers” like Joel Spolsky complain about how hard software development is. I managed to develop my website by cobbling together some javascript and a few PHP database queries. It was easy! How much harder can web browser development be? So what if the massive spec is ambiguous and filled with sentences that are basically incomprehensible without hours of study. Learn to cut and paste from example code, lazy browser developers!
And what’s up with stupid Microsoft? Why do they care about all this backwards compatibility testing and quality assurance? Hey stupid Microsoft, nobody cares if 90% of the websites that normal people want to use are broken! All that matters is that it will be easier for us web-developers to change the designs of our blogs. Isn’t it obvious that making *our* web-developer lives easier is waaaaay more important than making our *users* lives easier.
Just make IE8 work with our ACID test and then release it. Sure, that’s just a simple test suite, and doesn’t guarantee much about the spec actually being implemented properly, but all the other browsers can pass it so they are obviously more standards-compliant than you. It’s really all that matters, as far as complaining about IE8 on our web-developer blogs goes. We’ll totally stop and be your BFF once you break the internet for the 90% of your users. Promise.
Comment by web-developer — Tuesday, March 18, 2008 @ 8:15 pm
Dont you work at Google?
Comment by Uncle Arnie — Tuesday, March 18, 2008 @ 8:18 pm
I’m a big fan of both Mark and Spolsky, but I can’t stand how he says
then says
and then, TEN WORDS LATER, says
Spolsky just lost several points and the top spot on my personal scoreboard of People Who Aren’t Full Of Crap.
Comment by Lanny Heidbreder — Tuesday, March 18, 2008 @ 11:38 pm
> I love it when these supposed “programmers” like Joel Spolsky complain about how hard software development is.
If your company’s flagship web-based product was written in a homebrew metaprogramming language that resembles Visual Basic with functional programming lodged in its rectum and compiles down to VBScript, PHP, and JavaScript, you’d probably whine about development being hard, too…
Comment by Anonymous Hero — Tuesday, March 18, 2008 @ 11:45 pm
Joel’s post is mostly pointless because all those sites that will be broken in IE8 (including Google Maps) can also include one line of code to fix it. Freeze it in IE7 mode forever or at least until they update their site.
Comment by Jeff Adams — Tuesday, March 18, 2008 @ 11:50 pm
I knew either you or Gruber would tackle that one. Nice work!
Comment by Collin — Wednesday, March 19, 2008 @ 12:12 am
Congratulations, you have just won your troll merit badge.
Comment by John Doe — Wednesday, March 19, 2008 @ 12:13 am
IE8? They’re still working on that browser? 75% of market share and falling quickly
Comment by Rob Cooper — Wednesday, March 19, 2008 @ 1:24 am
Spolsky gave up on comments on his site because he didn’t want to hear from the trolls anymore.
I think he just got tired of people criticizing him on his own site. I stopped reading him after that.
Turns out, I just went there for the critics.
Thanks for this one, Mark. I saw that post linked from some other site and thought I’d give it a go. A third of the way through the whining, I gave up.
What happened to Joel? He used to be somewhat interesting even when he name dropped Microsoft in every other sentence.
Comment by Jim — Wednesday, March 19, 2008 @ 1:58 am
To those supporting Spolsky’s stance as pro-user or whatever. Look — ’standards’, strangely enough, means standards. If you want to write specifically for IE5’s, or IE6’s, or IE7’s ’standards’ mode BUT make your pages depend on certain broken or ‘extra’ bits as implemented my Microsoft… go ahead. Do it. But don’t specify standards mode without specifically browser-sniffing and detecting that version of IE. Or if you do, don’t complain. Because its not standards. It will always break, should always break, if you are idiotic enough to do that. To do that (or to encourage it) is just totally 100% devaluing the concept of a standard.
Look at it like this. Company X makes cars, and says ‘our fittings conform (mostly) to international standard ISOxxx’. Company Y makes parts for those cars, but takes special advantage of some minor non-standard differences/optimizations in company X’s fittings. Their advertisements and manuals say, these parts will fit on all cars that conform to ISOxxx. Company X updates the cars, makes its fittings slightly more conformant. Company Y’s parts break the engine due to the new changes. Repeat several times. That’s crazy. Well its even crazier on the web because Company Y has lots more options: (1) work within the published standard (2) work with modifications for each browser (3) do either of the above depending on the situation. There are even special tools to do that…
Comment by Darryl R. — Wednesday, March 19, 2008 @ 2:24 am
>If your company’s flagship web-based product was written in a homebrew metaprogramming language that resembles Visual Basic with functional programming lodged in its rectum and compiles down to VBScript, PHP, and JavaScript, you’d probably whine about development being hard, too…
Anonymous Hero, do you work for one of the telephone switch manufacturers? That’s the kind of crappy web-apps we’ve been having to deal with for our new phones.
Comment by Just a Reader — Wednesday, March 19, 2008 @ 3:20 am
Again, a lot of people are missing the point. His article is not about the validity of web standards, but more about the issues of creating an interoperable solution in the real world. In a way he is a Microsoft apologist, but his point is valid: the world is not perfect, and standards are only a solution to some degree. There are always several factors to consider: company strategy at any point in time (for instance: market dominance is the goal of any sane company, and Microsoft is not the only company that tries to do this). Another important factor is the fact that standards *are* flawed (trust me, I’ve read them, and most of them are ambiguous) and are created by and interpreted by humans.
Comment by Ben — Wednesday, March 19, 2008 @ 4:45 am
It’s also worth pointing out that W3C hosts automatic validators for HTML and CSS:
http://jigsaw.w3.org/css-validator/
http://validator.w3.org/
They may not be perfect, but there’s really no excuse for web developers not to use them.
And that’s not to mention the fact that many web development tools have their own built-in validators, or that you can use third-party tools like James Clark’s SP for off-line validation.
Comment by alastair — Wednesday, March 19, 2008 @ 7:14 am
It’s funny how Joel rants about pages breaking because of IE6/7-specific “hacks”, and then suggests that the way to fix them is to add an additional IE8-specific hack.
Comment by OMGHACKER — Wednesday, March 19, 2008 @ 8:21 am
Joel went pretty far out of his way to make it clear that 1) he doesn’t like the situation any more than anyone else does 2) he doesn’t know how to fix it. It’s really just background and history; it’s the story of how we got to where we are.
Despite that, much of the reaction (like this piece) takes it as given that Joel is advocating some particular course of action.
Comment by Western Infidels — Wednesday, March 19, 2008 @ 10:10 am
Joel was and still is binary. Sometimes he is spot on (leaking abstraction, iceberg principle, hiring), sometimes he is totally clueless.
Peace
-stephan
Comment by Stephan Schmidt — Wednesday, March 19, 2008 @ 10:46 am
> takes it as given that Joel is advocating some particular course of action
You are right. You can read nothing into the fact that he labels one group ‘pragmatists’ and the other group ‘idealists’. And you can ignore the fact that he said “as usual, the pragmatists are right”. Please ignore the man behind the curtain and all that.
Comment by Sam Ruby — Wednesday, March 19, 2008 @ 10:58 am
>for instance: market dominance is the goal of any sane company, and Microsoft is not the only company that tries to do this
Few disagree with this — it’s obvious to anyone that isn’t naive or in denial. Microsoft wants to control the world. Microsoft wants to control all standards. Microsoft wants to be in a situation where they can squeeze out a “standard” on the same day as a complete product, and foist it on the world, always three steps ahead.
I don’t fault them for it at all, and indeed they are doing the sort of thing that most any normal business does.
Who I do fault are people who stuff their head in the sand and declare it all just anti-Microsoft conspiracy theories. It’s bad enough when it comes from a Microsoft employee, but at least they’re coerced by the grossly manipulating effect of a paycheque, but it’s 100x worse when it’s just a fanboy hoping that one day the Eye of Microsoft might turn their way, rewarding all of their great astroturfing with a job or something.
Comment by Dennis Forbes — Wednesday, March 19, 2008 @ 11:00 am
> It’s bad enough when it comes from a Microsoft employee
For the benefit of anyone who doesn’t know, Joel used to work for Microsoft but now runs his own software company.
Comment by Mark — Wednesday, March 19, 2008 @ 11:24 am
>> If your company’s flagship web-based product was written in a homebrew metaprogramming language that resembles Visual Basic with functional programming lodged in its rectum and compiles down to VBScript, PHP, and JavaScript, you’d probably whine about development being hard, too…
> Anonymous Hero, do you work for one of the telephone switch manufacturers?
I believe this is a reference to Wasabi.
Comment by Mark — Wednesday, March 19, 2008 @ 11:26 am
> Dont you work at Google?
I do, but I don’t speak for them on this blog. Perhaps I should add a cat picture to make this clearer?
Comment by Mark — Wednesday, March 19, 2008 @ 11:28 am
I don’t get all the “sometimes Joel is just clueless” posts. It is entirely possible to be both entirely clueful and entirely full of shit at once. Hanlon’s Razor really deserves a further corollary: “never ascribe to stupidity that which can be adequately explained by a desire to make a buck or sound clever“.
90% of the tech punditosphere can be explained by this. Certainly both Gruber and Spolsky are imminently full of clue on a variety of subjects upon which they nonetheless post execrable nonsense in their blogs.
- Chris
Comment by Chris Cunningham — Wednesday, March 19, 2008 @ 11:33 am
> standards are only a solution to some degree
They would be more of a solution if Microsoft were actively participating in the development of HTML 5. Yes, I know Chris is the co-chair. Go count the number of posts he’s made to either the WHATWG list or the W3C HTML Working Group list. Go ahead, I’ll wait.
Oh look, here’s Chris 8 days after joining the group: Versioning and html[5]. It was rightly met with strong opposition and never made it into the spec.
And then… nothing. For months. A few administrative bits about teleconferences, a promise for a thorough spec review that never came, a one-liner here or there about accessibility, and then… nothing.
And then… IE8! Which, shocker of shockers, implements EXACTLY THE KIND OF FRAGMENT-THE-WEB VERSIONING SCHEME that the working group shot down last April. Apparently, they spent 7 months figuring out how to route around the HTML working group, and finally settled on using existing de facto extensibility mechanisms like an X- HTTP header and a META tag. And, apparently, getting some friendly stooges like Zeldman to launder it through A List Apart.
Let me make this clear: this is Microsoft’s SOLE CONTRIBUTION to the ongoing HTML 5 standards process. You can read the mailing list archives yourself if you don’t believe me. They completely ignored it for YEARS while it was still a loose federation of rabble-rousers at whatwg.org, then got “invited” to co-chair when the whole thing got subsumed into the W3C. And the FIRST AND ONLY THING they did was propose this versioning scheme. When they didn’t get their way, they took their marbles and went home, did whatever the hell they felt like (in secret), and sprang it on the world 8 months later.
So yeah, when I read Microsoft apologists crying about how *super confusing* standards are, and how *difficult* they are to implement, as if Microsoft were just a HELPLESS FUCKING VICTIM of a cruel cruel world — as if they were even making a GOOD FAITH EFFORT TO PLAY ON THE OPEN WEB — I just want to vomit. On them.
P.S. - Chris posted Versioning and html[5] on April 12. Guess what Microsoft introduced 3 days later? Good faith effort, my ass.
Comment by Mark — Wednesday, March 19, 2008 @ 11:55 am
Sometimes I think MSFT does what it does, or does not do, to browsers just to keep the Software as a Service people at bay while they squeeze every nickel out of their old thick client and/or ASP implementations before joining the rest of the 21st century.
Othertimes, I think they’ve just got their heads in the sand hoping that standards and browsers other than theirs would just go away.
Most of the time I just think they’re stupidly stuck in their own little world.
Comment by Mean Dean — Wednesday, March 19, 2008 @ 12:47 pm
I really can’t see Joel as a MS apologist here. What he says is true, “IT’S THE USERS STUPID”. Sure we would all like to get rid of past baggage and go to real “standards” (whatever TF they are), but that is not practical. Backward compatility like it or not is a MUST. If you haven’t been writing software long enough to know that then go change your diapers. You may not like MS, you may despise them. That doesn’t change the fact they OWN the browser market by a large margin. I don’t use thier browser, but that doesn’t matter and I’m smart enough to know it. Is linux a free and BETTER OS? I think there is no doubt, is your mother or aunt using? NO way. MS “OWNS” the desktop as well and shows no signs of quickly losing it ,despite the Vista setbacks. So stop blasting Joel for what is obviously true. N00bs.
Comment by ZetaPrime — Wednesday, March 19, 2008 @ 1:18 pm
> Is linux a free and BETTER OS? I think there is no doubt, is your mother or aunt using? NO way.
Actually, my mother does use Linux. Migration was not without its problems, but they’re all smoothed out now and she’s perfectly happy running Firefox and Thunderbird on Ubuntu. Though I’d bet real money she couldn’t come up with even 1 of those 3 names if you asked her. They’re just “web browser”, “email”, and “computer” to her.
I mostly agree with you. Microsoft has a monopoly on the desktop, no question. They have a majority browser share, no question. Neither of these things will last forever, though. The best way to combat X-UA-Compatible (or the next scheme that Microsoft comes up with, or the next, or the next) is to reduce their market share. Help a friend install Firefox. Help your parents migrate to Linux. Whatever. Every little bit helps.
Comment by Mark — Wednesday, March 19, 2008 @ 1:40 pm
So the web is now a clogged septic system. Lets create something that keeps all the crud plugging up the system. WTF?
I’m for breaking everything in order to make it work right. I love backwards compatibality, but sometimes it’s just beter to say fuck it and start all over. So what if you site looks like poo for a day or two put up a “we’re sorry page” every one else will have one. Feed readers should still work. This might bother people for a day or so maybe even a week (until the site gets fixed), but people can adapt we’re amazing in that regard.
I’ve been a software developer for over a decade and chances are you’re using something that has my code in it… if it means anything.
Time for a flush.
Comment by Ken — Wednesday, March 19, 2008 @ 1:59 pm
1. Spolsky develops PHP software using his own language, Wasabi, not just with Microsoft tools.
2. Spolsky has scorchingly criticized Microsoft repeatedly.
3. WTF are you talking about with the AdSense line? Spolsky owns his own company and posts rarely, and, minor correction here, *there is no AdSense on that page or on the site*.
I own your book but this is just stupid.
Comment by Ryan — Wednesday, March 19, 2008 @ 4:00 pm
HTML is a failure. We to all stop trying to fix it by creating standards only a rocket scientist could understand.
What’s wrong with starting over? Just stop trying to improve HTML and the browsers implementation of it. Do something new, simple, straightforward, easy to implement, and truly standard. It has to be different enough that it won’t get confused with HTML. And stop letting crappy pages get through the development process.
Joel is right. Where the hell is the W3C browser that everyone can test against?
Comment by Mike — Wednesday, March 19, 2008 @ 5:01 pm
> HTML is a failure.
I wish everything I did was a failure like HTML.
Comment by Aristotle Pagaltzis — Wednesday, March 19, 2008 @ 5:46 pm
There is a straightforward strategy that Joel doesn’t mention: modify IE8 so that it doesn’t claim to be Internet Explorer in the User-Agent header. That will suppress server-side IE-specific hacks (at least in standards mode) and IE8 will get the same thing Mozilla gets.
What should it report instead? Some unique token not yet used in the User-Agent space.
Comment by John Cowan — Wednesday, March 19, 2008 @ 5:47 pm
Surely all the IE6 CSS fixes that depend on the parsing bug, won’t affect IE8 - assuming they fix the parsing bug.
The document Event / passed Event in JavaScript switch/’bug’, will carry on working.
How long is it going to take Google to fix Google Maps for IE8? Not long I would guess.
Why don’t web page writers/programmers allow their pages to degrade gracefully - do I need another button that does nothing when you click it? Or requires me to download a plug-in to use the web-page/site?
And surely you should be able to have both IE6/7 and IE8 on your system, so if it doesn’t work in IE8 you can open it in IE6/7 - without cluttering up IE8 with compatibility code that would be hard to debug and might be a security risk, etc.
I thought the Vista / IE8 thing was supposed to be an clean sweep, do it right from root level up, type rewrite (the sort of thing that killed Netscape according to Joel.)
Comment by Bobby — Wednesday, March 19, 2008 @ 6:30 pm
The change the user-agent string, occurred to me almost immediately, so I rushed over to comment on it, only to discovery 100s of others had beaten me to it.
As to what to change it to, Explore Internet (an action instead of a thing), or Internet Browser (perhaps too limited sounding - “can you fill in forms?”), View Internet - but VI is taken :)
Comment by Bobby — Wednesday, March 19, 2008 @ 6:39 pm
I must admit, I almost bought Joel’s article when I read it.
But I’ve just spent 6 hours rebuilding a CSS layout from scratch because IE7’s Page Zoom feature JUST DOESN’T FUCKING WORK. It gets very tedious to design around IE’s limitations all the time.
Let me put it another way — this time, ![if IE gte 6] style conditional comments were the only thing that got me through this sane.
But why have I never thought to myself, “I want this style to apply just to Opera/Safari/Firefox”? That’s right — because they almost always render pages in the same way! It’s ONLY IE that continually has this problem.
Comment by GuruJ — Thursday, March 20, 2008 @ 1:03 am
I just love all the people who say standards/html is a failure. I also love the people who say IE is the standard we should follow. Yet, if you write html, or css, what published standard are you following? What published standard does IE attempt to follow (badly)? Where is this so called “IE standard” published?
Then there are those who say we should feel sorry for Microsoft and forgive them because they have to have backwards compatibility or they will break the web and all those Fortune 500 intranets. Well, many businesses live and die by decisions and mistakes they’ve made in the past. Microsoft must live by their past mistakes and live/die by them.
Comment by Rob — Thursday, March 20, 2008 @ 8:10 am
@Ryan
I agree that Spolsky occasionally smokes Microsoft, but just as he’s smoking their pole:
“Why are the Microsoft Office file formats so complicated?”
http://www.joelonsoftware.com/items/2008/02/19.html
Read that and then reread this post…
Hi, I’m Desktop Developer Barbie. Pull my string and I say, “Binary formats are tough! Let’s go shopping!”
Comment by Dean Landolt — Thursday, March 20, 2008 @ 10:22 am
> P.S. - Chris posted Versioning and html[5] on April 12. Guess what Microsoft introduced 3 days later? Good faith effort, my ass.
I’m virtually certain that you disagree, but arguably, Silverlight and Flash are a solution to a different problem than HTML 5. Web content, and web sites based on content, should be written to conform to HTML standards as much as possible — exceptions are limited to things like YouTube, because there hasn’t been a good enough standards-based solution to web-based video, though hopefully the HTML5 stuff will fix this.
But I’m not sure I believe that applications are a special case of documents. And that’s kind of the premise behind writing real applications using AJAX. I’ll certainly grant that Google Spreadsheet is an impressive demonstration of what you can do with AJAX, but just because applications *can* be built that way doesn’t mean that they should.
Given that, I think an application-programming environment that plugs into a browser is an entirely reasonable endeavor. Then the only problem is that neither Sivlerlight nor Flash are open standards and we ought to have an open standards-based implementation of such a thing. Which I agree with, by the way.
Comment by Andy Norris — Thursday, March 20, 2008 @ 1:33 pm
Man, I enjoyed this. Thanks Mark.
To anyone claiming Joel as some sort of usability engineer (nee one of the best even) then why are we uninstalling FogBugz at my company this weekend? The FogBugz homegrown(!) VBScript derivative just doesn’t cut it for performance, and the UI caused howls of protest from multi-bloody-click-resistant dev/qa staff (you know you’re in trouble when assembly programmers complain about a Web UI).
He’s a fair writer, but his company stuff would be lost without the resultant marketing to the tech crowd (’Hire Smart Devs!’ Hey that’s me. Joel is /so/ cool and correct), but post the ‘When I was at Microsoft I put Microsoft VBA in Microsoft Excel when I was at Microsoft’ stuff he’s been all down-hill…
@Ryan - grow out of the man-crush thing, time to flip the Bozo bit on ol Joel I’m afraid…
Comment by David — Thursday, March 20, 2008 @ 3:09 pm
Mike from comment 60 has inspired me. “What’s wrong with starting over” indeed? Let’s ditch this HTML thing and use Gopher! C’mon, whaddya say? :)
Comment by roberthahn — Thursday, March 20, 2008 @ 3:53 pm
> HTML is a failure…
> Whats wrong with starting over?
Its being done. See http://www.curl.com
Comment by rustom — Friday, March 21, 2008 @ 10:08 am
I was at the treadmill on the gym reading Frank P. Ginac’s “Customer Oriented Software Quality Assurance” when it hit me.
I think a point Mr. Spolksy is missing is how often both developers and customers have to go through dramatic changes with each and every iteration of a Microsoft Product.
Whether it’s the leap from IE6 to 8 w/all the kludge in between - or the hapless Word 2007 user who now effectively has to relearn the menu system because for some reason, well actually I don’t know the reason why they radically changed the menu system.
So too with developers, whether you’re one of the VB users whose investment of code gets wiped every 3 to 4 years, or a C#/.NET developer who now finds time invested in their ill conceived template/skinning compelled to start over again with yet another “Microsoft” standard … as opposed to a compiler that simply works and plays well with the W3C.
Fact is, it continues to make less and less sense to stick with their products because they morph into entirely something else every couple of years. Case in point, IE 8.
Comment by Mean Dean — Friday, March 21, 2008 @ 1:22 pm
@Mean Dean… you’ve totally hit the nail on the head. I’m dealing with the aftermath of the Microsoft “standard” roller coaster right now. Some of my customers who built and bought solutions built exclusivly for the Microsoft platform of the time. This means IE 5.0, VB, some weird end-of-life Microsoft middleware, Microsoft J++ and Office 2000. There’s some hokey IBM shit in there too.
Fiscal difficulties and horrific consultingware architecture made it difficult to update their applications, so we ended up with all sorts of nightmarish scenarios… multiple versions of Office that are impossible to update, and webpages that require things like Citrix to access with IE5.
But somehow, lots of people have this crazy idea that Microsoft is the “safe play” and you have a “roadmap” to work with. But my customers using open source backends are having none of the problems that these folks have.
Comment by Duff — Friday, March 21, 2008 @ 7:14 pm
@ (#77) rustom 10:08 am
>> Whats wrong with starting over?
> Its being done. See
> http://www.curl.com
[This partly off-topic, but...] somehow I cannot convince myself to trust a company that’s either ignorant of; doesn’t care; or doesn’t know how to set proper HTTP Content-Type headers on its php-served pages (of which site’s but the top one, above, is accessible sans error warning to standards-compliant browsers).
Call me picky, but, to repeat John Gruber’s one-liner comment on Joel Spolsky’s rant, before he bailed out in favor of Mark’s superior, errr…, exegesis [lookitup!]: “you reap what you sow.”
Comment by Ianf — Saturday, March 22, 2008 @ 7:47 am
This page reads like a Bush/Cheney cartoon fever-dream. Not everything is good or evil. Joel’s merely pointing out that Microsoft’s chickens have come home to roost. Those of you who take umbrage at words like “pragmatist” and “idealist” are just strutting around with chips on your shoulders. A truly enlightened geek is both.
Comment by kokorozashi — Tuesday, March 25, 2008 @ 3:18 am