Empire of (white on orange) XML

I actually felt a brief and surprising flare of hope the other week, when Dave was saying that rather than the incomprehensible and unlocalizable white-on-orange “XML” button to indicate the presence of an RSS feed, people ought to use a white-on-orange “Subscribe” button. Sure, it’s problematic on newspaper sites, where they tend to use that word to mean “give us your credit card number, and we’ll ship sliced dead trees to your house,” but a step away from “XML” as button text is a step in the right direction.

Unfortunately, that was apparently a brief detour, and in reaction to the Microsoft RSS Team’s heretical use of a white on orange image with no text, now we’re back to:

Note to those who think the white-on-orange XML icon isn’t suitable for international use, consider that the People’s Daily, of the People’s Republic of China, uses the icon, as-is, without reinventing it. I figure if it’s good enough for the Chinese, it should be good enough for Microsoft.

Um. Dave? That’s the English edition of the People’s Daily. That’s why everything looks familiar to you: it’s supposed to. If instead you look at the GB-2312 Chinese edition, you’ll find that virtually nothing looks familiar, and the only things that do, URLs and email addresses and one lonely link reading RSS订阅, are the result of us insisting that the whole world use our character set. Try this on for size: how about if everyone around the world, no matter what language they speak, uses as the label on the submit button for every web form (picking a couple of characters at random, which I hope don’t say something like “dog puke”) “搜索” and for the reset button, “希索”?


Comment by Micah #
2005-10-16 02:34:50

”RSS订阅” is, interrestingly, a compromise: ”Subscribe to RSS”.

Comment by Richard #
2005-10-16 02:45:23

Can someone remind me, if something that’s good enough for a Communist dictatorship is good enough for a heartless multinational corporation, then why is that thing good for me?

Comment by Aristotle Pagaltzis #
2005-10-16 03:04:48

Why do we even need those damned buttons at all? Is there any use case that autodiscovery can’t handle? (And why does Winer think he is the arbiter of everything to do with syndication?)

Comment by Phil Ringnalda #
2005-10-16 16:27:38

That depends on whether PaceAnchorSupport is accepted (both in the ”formally accepted into the RFC for Atom autodiscovery” sense and in the informal ”enough aggregators support it, successfully, for Atom and for RSS” sense). My use case for autodiscovery-isn’t-everything has always been someone stuck with a CMS and a corporate process that allows them some freedom to affect things in <body> and absolutely no freedom to affect things in <head>.

Comment by Phil Ringnalda #
2005-10-16 10:46:49

For anyone who wants to comment before I figure out how to unbreak what I’ve broken: don’t use any newlines. If you just run everything together in one paragraph, or do your own <p>…</p><p> without line breaks between them, it’ll let you through.

Dave’s objection was that ”Subscribe” is also English – thanks to late night writing, I didn’t make it clear that the advantage of ”Subscribe” is that it is a word, and thus localizable. If I’m picking the right glyph from what Micah tells us is ”subscribe to” then a Chinese site or Chinese copy of IE would have a white-on-orange ”阅” button, a French site or browser would have a white-on-orange ”Souscrivez”.

Comment by Phil Ringnalda #
2005-10-16 12:13:59

Okay, I think I fixed the comments, so there shouldn’t be any more need to avoid newlines.

Feel free to paragraph again.

Comment by Dave Winer #
2005-10-16 12:13:50

Phil, for a guy whose comment system is broken, you sure take a lot of cheap shots. It would be nice if you wouldn’t translate my writing for others, instead, quote it. I didn’t object to Subscribe (how unfair of you to say I did) I actually proposed it. Methinks you’re trying to score some personal victory instead of have a meeting of minds. Now cut it out, okay?

Comment by Phil Wilson #
2005-10-16 13:30:28

Actually Dave, you *did* object to ”subscribe” in your response, because you say ”one way to do something, no matter how flawed that one way is, is better than two, no matter how much better the second way is.” and you’ve already said, many times, that you think the ”one way” is the white-on-orange XML button (here, http://archive.scripting.com/2005/10/08#When:4:23:06PM two weeks after you wrote about using ”subscribe” here http://www.reallysimplesyndication.com/2005/09/26#a970). If this is a mis-understanding, I’m sorry, and I look forward to a clarification of your position.

Additionally, you completely ignore Phil’s point that despite what you’ve said, the white-on-orange XML button *isn’t* good enough for the Chinese.


Comment by Jacques Distler #
2005-10-16 12:57:37

How … umh … sad. Does anyone still care what Dave thinks about his misbegotten little orange button?

One would have thought that he would, long ago, have learned that a red octagonal sign means ”stop”, no matter what text (or alphabet) is used on it. From that, an alternate-reality Dave would have concluded that an orange button with white lettering (no matter what text or alphabet is used on it) means, ”this is a feed; subscribe to it.”

Comment by Jacques Distler #
2005-10-16 15:08:10

Just to make things even more confusing, Dave uses the same non-localizable ”White ’XML’ on an Orange Background” GIF image to link to both and OPML file and an RSS feed.

After beating us over the head that this button (and no variation thereon!) indicates an RSS feed, he goes ahead and uses it to link to an OPML file.

Me, I’m using CSS to create ”White on Orange” buttons which actually unambiguously and localizably indicate what kind of feed they link to.

If I want to create a button to link to an OPML file, I’ll be damned sure to use a different colour.

Comment by Danny #
2005-10-17 08:52:20

When I expressed my own flare of hope at the move from [XML] to [Subscribe], Morten Frederiksen reminded me of his 3 year old Syndication Subscription Service. This intermediary allows any-aggregator subscription from a single icon/link, and one option for that is a white-on-green [Subscribe] button. It would seem that not only is thought needed in this domain but research too…

Comment by Mark #
2005-10-17 09:25:51

> It would seem that not only is thought needed in this domain but research too…

You must be new here.

Comment by Danny #
2005-10-17 08:59:45

PS. I’m confused over Dave’s position – his post re. [Subscribe] to which you link does seem to favour that over [XML]. Maybe that suggests an exception to Dave’s rule:

One way to do something, no matter how flawed that way is, is better than two, no matter how much better the second way is.

– what if the something is ’communicate’?

Comment by Phil Wilson #
2005-10-17 09:14:11

He did suggest there would be exceptions:

”I’m sure some smartass will find an exception to that, and I’m sure there is one, but ingeneral that’s true.”

Danny, you know it’s true. ;)

Comment by Danny #
2005-10-17 14:16:46

I dread to think what this would make someone who demonstrated it wasn’t true in general.

Comment by Phil Ringnalda #
2005-10-17 15:26:27

I think Mark has us covered. Me, I think I’ll just moronically change my Funky RSS buttons (the funk has worn rather thin, now that I know more about Dublin Core than I once did) to ”Subscribe” text with white-on-orange CSS, and find something to worry about other than who uses what image, and who that allies them with or against.

Comment by Anil Dash #
2005-10-17 17:26:10

Y’all let me know how it turns out! I can’t wait to see!

Comment by Phil Ringnalda #
2005-10-17 19:14:30

As is all the other lint in my navel, my ever-changing reactions toward comments is fascinating to me.

There was a time (still is, to a slightly different extent) when my reaction would be a simple and pure ”oh holy crap, Anil Dash himself just left a comment on one of my posts one-one-one.”

Now? Hmm, does Anil actually link to that home page, or to dashes.com/anil/? Doesn’t he use his work email in comments? What time was this comment left, should he still have been at work, or not? Who owns that IP address, and is it in a likely part of the world? Better view source on anildash.com, see if it’s actually a cloaked spam page: I know I’ve seen a non-weblog home page for him, but I don’t know if that’s the URL or not.

Between that reaction, always in the background these days but more pronounced than usual because of that too close to ”Great post! Keep up the good work! Visit my [porn site].” comment, and the fact that people are telling me that they’ve gotten spam comments (in the hundreds, in at least some cases) using a misspelled version of my domain name for the email, I’m swinging back to thinking that PGP-signed comments are the way to go, and people need to just suck it up and jump over the knowledge barrier to getting it done.

Comment by Jacques Distler #
2005-10-17 20:29:18

Yes, staring at the source of all your commenters’ web sites is no way to live.

I don’t know that PGP-signed comments will fix the problem. Instead, you may end up staring at the signatures on Anil’s key.

At least, if you’ve gotten one signed comment (that you trust) from Anil, you don’t have to agonize over others.

Comment by Phil Ringnalda #
2005-10-17 21:10:57

My brain was well into the process of sending my fingers the signals for ”it fixes my problem” when I realized that of course it doesn’t.

So, I have a signed comment purportedly from Anil. I need to either have my own reasonably dependable URL for him, say from a subscription, or I need to go to a site I trust myself to navigate to, which blogrolls him (in a reasonable trustworthy way, certainly not through including anything from a third-party site), to be sure I’ve gone where I expected to go to retrieve his key. Otherwise, I may well have verified that the comment was left by someone who controls ani1dash.com or dashes.org/anil/.

So, if an old friend I know
Drops by to say hello
Would I still see suspicion in your eyes?
We can’t go on together
With suspicious minds
And we can’t build our dreams
On suspicious minds
Comment by Sam Ruby #
2005-10-17 21:13:24

I’m not done evaluating it, but OpenID may be better.

Requires nothing to be installed on the client. Once set up, requires no interaction for the user. And when all is said and done, you know that the owner of the URL left has vouched for the comment. Of course, this means a lot more if the URL is dashes.com instead of mypornsite.com…

Comment by Jacques Distler #
2005-10-17 22:20:21

OpenID sounds interesting, but I haven’t had the energy to think through all the security implications of the Specification.

The things I worry about are malicious hosts spoofing the authentication transaction and stealing a user’s credentials. Or replay attacks or …

The ”Requires nothing to be installed on the client” aspect sounds good, but does lead to some worries, no?

In the case of Anil Dash, or Sam Ruby, there’s a non-negligible chance that the PGP key in question will actually be ”trusted” by Phil Ringnalda, as those individuals likely have overlapping webs of trust.

Phil would, then, have two pieces of data to satisfy him: a URL controlled by ”Anil” (with a <link> to his key) and a chain of signatures which confer trust to that key.

If we didn’t have to worry about IRI spoofing (etc), we might even bootstrap one into the other.

Comment by Sam Ruby #
2005-10-18 05:46:25

Having implemented OpenID, one of the more surprising thing to me is that there are no credentials to steal. Instead, new secrets are created each interaction.

2005-10-16 10:06:16

Response to Phil Ringnalda

I tried to comment in response to this post on Phil Rignalda’s but his comment system would take it.

Name (required)
E-mail (required - never shown publicly)
Your Comment (smaller size | larger size)
You may use <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <del datetime="" cite=""> <dd> <dl> <dt> <em> <i> <ins datetime="" cite=""> <kbd> <li> <ol> <p> <pre> <q cite=""> <samp> <strong> <sub> <sup> <ul> in your comment.