Having skimmed through this Mozilla newsgroup thread, I'm perturbed anew about the new Firefox 2.0 misfeature with regard to RSS feeds and client-side XSL. My original grief was that it cost me a significant bit of pointless debugging time even discovering that it was a brand new gotcha. But, the "we know what's good for you" sentiment is clear - I feel like posting any newsgroup responses would be just more time wasted. At least shoving an arbitrary 512 bytes into the feed fixes the bug. (This seems to compress well: "AAAAAAAAAAAA ... AAAAAAAAAARGH!")
Brilliant: I'll just add that to the body of obscure CSS hacks and bug workarounds that'll someday make me give up web development for raising sheep. Thanks a million, guys.
Some days man, I'm not sure raising sheep would be such a bad idea.
30 years behind the keyboard is a long time.
Especially at 37.
I'm about 6 years behind you, so let me know if you hit the sheep herding phase before me, and let me know how it treats ya. :)
Not to be pedantic, but don't you think there's a lot of random shit a shepherd needs to memorize too?
And you get to work indoors, avoid anthrax and keep your hands free of ovine feces.
I agree that the overriding of RSS styling is wack, but maybe you should go on a sheep-herding vacation if for no other reason than to realize this is a speck on a mote on a flea's arse on the giant ewe of web development. :)
les, your feedback seems to fall into the "IT IS OBJECTIVELY WRONG" bucket. It does fail to account for the behavior of Safari 2 and IE 7, and doesn't account for authors who would like their stylesheet to show only in older browsers without RSS support. Why do you think it's good for the Web for Firefox to diverge from the new de-facto standard? Seems like pissing in the wind to me.
BTW, I can't say framing the decision as some collective character flaw of the Mozilla project is particularly productive.
Joe: Sheep farming has always been my frustrated-programmer joke. It would be something different to memorize - not necessarily easier. But, hmm, I never thought about anthrax. I suppose that would be a job hazard to consider, beyond simple wasted time.
Robert: Admittedly, I'm frustrated. This change breaks things I've invested hours into across several jobs, and into which I've been investing hours again. At least I've been paid for those hours, and I'm glad I don't have to go back and explain / fix the stuff now broken at previous jobs.
From my perspective, I'm seeing all the latest generation browsers now breaking something I thought was standard - but about that I guess was mistaken. At least, it "smelled" like a fairly clean quasi-standard approach to me. I really don't have standards all memorized or even well studied.
But, in the new defacto standard, I get to stuff 512 random bytes into a feed to get the old behavior back. That certainly doesn't "smell" like a clean standard to me, even if IE 7 does it and Firefox 2 now follows. It still doesn't work in Safari. It seems a defacto standard to me in the same way that exploiting CSS parser errors to filter by browser is a defacto standard. Yet another weird edge case to memorize.
From my perspective, the defacto standard has been what FeedBurner and others have been doing with client-side XSL for feeds. And there've been some exciting things starting there - so there's some personal disappointment invoked. And now, all those feeds are broken in that respect. Retroactive wind-blown urine, at this point, I guess.
Maybe all this was really just a neat geeky trick best now forgotten, and the decision really will be the right one in the end. Nonetheless, I don't like it. So, perhaps not "IT IS OBJECTIVELY WRONG". But, from my perspective, it was an unexpected and unwanted surprise.
As for being productive, the decision seems made and others have already registered the same or better objections than I would just rehash. And I can't see with what collective character flaw I've painted the team - I'm perturbed, but I certainly don't think everyone involved is a villain. My apologies if my frustration vents that way. It's just that taking over feed styling from content providers strikes me as "we can do user experience better", and the explanations I've read don't seem to disagree with that notion.
But, I disagree and I don't like it. So, fine, from here on out I use the 512-byte-hack if I want custom XSL styling for feeds. Another trick in the bag. That's pretty much my last word on the subject, unless the capability disappears entirely.
All I have left to say is that I hope the client-side XSL-styled World of Warcraft home page never switches to an Atom or RSS format - whether or not using a non-HTML format for a front page was a good idea in the first place. I thought it was, but that could be my whole problem. :)
My "frustrated with technology" catch-phrase has always been to become a Buddhist Monk. That complete detatchment of technology seems to be quite the fantasy of all coders everywhere. Maybe that's saying something.
By "collective character flaw", I meant that I've read you and other web authors writing "they think they know better than us". That's not how the decision was made. And to be honest, it bothers me somewhat to read a rant on the subject that only focuses on Firefox. To me, it seems like taking advantage of an open product development process because you know we will listen, while other vendors won't. A rant that claims "Firefox and IE and Safari suck" would be different, because then it is an opinion on a behavior. Criticism of browser behavior is welcome, criticism of people based on tech religion is not so welcome.
I understand that FeedBurner and others have a business interest here, but you have to understand that at least some of the things they do directly conflict with capabilities that are beneficial to browser users (that we can't deliver if it only works with half of the feeds out there). We want our users to have a coherent feed experience, with browser-specific preferences and minimal click-through screens. To that end, the screen that supplants the XSL was termed the "feed preview". It should show what's in the feed, and default to user-specified feed preferences. If the feed doesn't contain ads for aggregators and feed services (and wrong instructions), a preview that shows that stuff deceives our users. Not in an evil way, but in a "bad usability" way. When combined with authors who want a browser feed screen with a stylesheet for older browsers, and Safari/IE7 behavior, the scales tilted away from using the XSL.
I understand that people disagree here. But the fact that all browser vendors have done something that doesn't align with the way they used to do it should tell you that the issue is more complicated than "author's intent", for better or worse.
Robert: You have a great point there - I do in fact think that Firefox, IE, and Safari all suck with this current change in feed handling. And yeah, the only reason I ranted about Firefox is because I think it's the only one of the three there's much chance in affecting. It's also the only browser I use, so it's the first one where the feed handling change really sunk in for me. I really haven't cared about it until just now.
Of course, I've been nowhere near participating in the newsgroups or code checkins or anything whatsoever related to actually contributing to Firefox - so this is all blog-chair quarterbacking from a syndication feeds enthusiast.
As for the "they think they know better than us" issue, I think it can be found in a newsgroup response like Mike Shaver's:
To this, the most vocal respondents say, "But I really want my users to experience my righteous XSLT!" Even if it has flaming GIFs, dancing hamsters, and cat pictures. It's righteous, damn it. (Yeah, the quoted "righteous" is a little off-putting when you actually think the XSLT thing is kind of keen.)
Why not apply the new-and-improved styling by default and bow out when there's a processing instruction? Then, spread the word - "Get rid of your cruddy not-for-you XSLT!" I'm sure you'd get lots of converts grateful to stop screwing with their feeds, which would leave the outlier nutters like me and FeedBurner to keep happily shooting themselves in the feet. Instead, I see, "style 'em all and let hacks sort 'em out," which totally rubs me the wrong way.
At this point, I'm really just mildly disappointed, and feel like I've gotten a downgrade in an otherwise all-around great new browser. I'm not actually looking to farm sheep over this. There is a hack-around available for IE7 and FF2, so at least there's that.
I guess what would make me feel like I'd gotten a decent consolation prize would be if that Feed Preview screen were significantly richer and easier on the eyes than the other browsers' offerings. I'm assuming that's projected for the future. Maybe what I need is to do a checkout and dig around.
CraigM: You may want to reconsider. I've known WAY too many buddhist monk programmers in my day.
conscious Robert: Why do you keep seeing people writing "they think they know better than us"? Because Mozilla has chosen to ignore an Author's preferences when displaying a feed. Period. You can explain it all you want, your reasoning may even be right- but the bottom line is, Mozilla made a conscious decision to ignore styling instructions included by the feed's author. The only reasons I can think of to do this is to be perverse (which I doubt), or because you think you know better (and from here on out, I use the collective you for Mozilla, because I'm tired of typing Mozilla).
You claim it's better for users. I say you think you know better what is good for your users than the feed author does. This is the same feed author that created the feed's content, and the stylesheet, and decided to present the feed to users with the stylesheet. I'm sorry, I just don't buy it.
Another non-starter is the IE7/Safari argument. I don't care what those browsers do; I chose Firefox because it decided to lead, not follow, when it comes to standards. Whether you believe that displaying RSS feeds with stylesheets is a standards issue or not, you can't deny it was a de-facto standard until recently.
And what about those authors who only want their stylesheets applied in older browsers that aren't as enlightened as IE7? Not that I've seen a public outcry from these authors in any of the forums discussing this "feature", but let's assume they must exist. They were smart enought to add the stylesheet to their feed, why not allow them to also add to their feed an indication that a browser's built in styling may trump the linked stylesheet (perhaps a PI)? Wouldn't that allow both types of authors to get what they want without wasting bandwidth in 1/2 kb increments of sheer hack?
I've read the arguments, here, on the buglist, and in the newsgroups. To me, and I suspect, to many others, Mozilla (or those who made this decision within Mozilla) has acted as though they think they know better than web site authors. Explanations such as yours appear to me as an explanation of why you think that, not a refutation that you don't.
Regarding the above: Sorry about the extra "conscious" at the beginning.... I was typing my comment in a text editor, and used the comment form to check the spelling of that word. Forgot to delete it when I pasted in my comment. It is not a poorly-formed attack on Robert :)
I, as a user, don't think feeds should be styled. Feeds are used in ways where styles are not easy to apply, would cause theming chaos, and just look good when all in the same style.
That said, I heard Firefox 2.0 has an ugly default style, which, I guess, should be improved. I like IE7s feed view. Oh, and the arbitrary 512 byte switch is odd - reminds me of Internet Explorer's Friendly HTTP errors... Which weren't that friendly.
Hm, I might not be thinking in the same way a regular user is thinking...