Bloggin' the web, one post at a time

sIFR sucks

Now before you start descending on my like a pack of wolves I’d like to make my case. Which at the moment sounds worse than it is. The sIFR technique is a very good solution for a particular problem. The problem being the severe lack of typefaces. However, that isn’t the problem, the actual problem is much worse.

Some, myself included, keep forgetting that the web is still very young in cultural, sociological and artistic terms. The W3C it’s just had it’s tenth birthday and yet it has delivered so much already. But some things do take time to evolve, at a more generational pace. Typography is one of those things because it is cultural, sociological and artistic. Anybody remember the rise of Desktop Publishing (DTP) at the end of the eighties? Many said at the time that DTP set the evolution of Typography back a hundred years, give or take. And they have been proved to be right. Even though a new impulse for graphic design jumped into the driving seat. Those that did master the use of typefaces pushed the boundaries of typography in a way that had never been envisioned before. And at a pace that had never been achieved before. But bringing a discipline to the masses also brought the level of acceptable typography down, way down. Everybody thought they could do it. Many think they still can. In the past an apprentice asked their master if it was good, now they ask the client.

2005-01-12 / 10:22

42 Comments

By Jeff Croft
01/13/2005 12:57 AM
Reppin' KC MO, Fellas and the ladies know.
Okay. All that makes sense and is well-said, but I just have one question:

What technology would you perfer to use that DOES allow you to kern your letters and use great anti-aliasing?

There isn't an alternative. sIFR may not be perfect, but it's the best thing we've got.
By Andrew Hume
01/13/2005 02:39 AM
London
Good points Egor - all totally valid.

As Jeff has mentioned, sIFR is useful for solving a problem that has no other solution right now.

I accept that it is ultimately a hack, and I hope that we will soon have something better. Meanwhile it's helping to make the web the best it can be. If nothing else comes from it, let's hope it makes someone (W3C?) realise that we need a proper way of embedding fonts on our pages.

" In the past an apprentice asked their master if it was good, now they ask the client."

Loved that by the way. Mind if I quote you one day? :)
By Egor Kloos
01/13/2005 02:40 AM
Rotterdam, The Netherlands, Europe, Earth etc...
Well that's kind of what I'm saying. The option of not using this sIFR technique or any other technique at all is nearly always preferable than using it.

Andrew Hume does caution the unwary but in my post I hint at the fact that those with talent can do great things with it but the rest of us mere mortals will not. The more powerful and easy to use a tool becomes the poorer it's result often becomes. sIFR is such a tool.

My concerns of kerning and anti-aliasing need not be a problem. Kerning is less of a problem when mono spaced fonts are use and Flash anti-aliasing more acceptable when used above certain font size.

All the disadvantages with images for example still trump the sIFR technique even with it's accessibility advantages. That's how strongly I feel. This is purely a subjective view and each has to make their mind up on how to use it. But I fear most will just use it and won't give it a second thought. Hence my rant.
By Egor Kloos
01/13/2005 02:57 AM
Rotterdam, The Netherlands, Europe, Earth etc...
Personally I don't think that 'hack' is a dirty word, I know some zealots that do. sIFR is a welcome addition to the advancement of the web. Also the sIFR shows us all as bright as day the gapping hole in todays GUI web browsers. So by all means I hope this thing keeps on getting better. Till the day software developers find out a way to distribute copyrighted fonts (even if it's just per session) we're stuck in neutral, raring to go.
By Jeff Croft
01/13/2005 03:00 AM
Kansas City
> All the disadvantages with images for example still trump the sIFR technique even with it?s accessibility advantages.

If you are comparing sIFR to images, you are missing the point of sIFR. sIFR is not designed to be used in situations where you would have used images in the past. sIFR is designed to be used in cases where you have hundreds, maybe thousands od headlines that are added and changed very rapidly by people who aren't Photoshop gurus, and therefore image creation for each one is unrealistic.

Remember, Mike Davidson originally started working with flash-based replacement for ESPN.com. On that site, ESPN needed a solution that allowed non-techy and non-designer types (read: editors) to be able to simply type the headline into their CMS and have it outputted in the corporate typeface. sIFR gives them that, and currently, nothing else can.
By Egor Kloos
01/13/2005 03:34 AM
Rotterdam, The Netherlands, Europe, Earth etc...
> ...to be able to simply type the headline into their CMS and have it outputted in the corporate typeface.

Well you can do that with images by generating and managing the images on the server-side end of the CMS automatically and loading them up to the website via client-side scripting and no I wasn?t comparing the two. In anycase that?s not my point, the use and misuse of typography is. Furthermore using this technique for corporate branding purposes is very persuasive. Another reason for me to be hesitant to deploy it, never mind mentioning it to the client. There are certain quality levels that would need to be assessed before that would happen.

Just because I have grave reservations on the use of this technique for the WWW in general doesn?t mean I?d never use it, that would be folly.
By Jeff Croft
01/13/2005 04:02 AM
Kansas City
But Egor, if you generate the images server-side, you also don't get to kern the letters. Just like sIFR, right? :)

By Jeff Croft
01/13/2005 04:03 AM
Kansas City
And one other point...

I share your concern for the "use and misuse of typography," but don't you think it's possible to abuse type no matter which technology (CSS, images, sIFR, etc., etc.) you choose?
By Andrew Hume
01/13/2005 04:32 AM
London
It's certainly possible to abuse typography with all those technologies.

However, sIFR has an extra potential for disaster in that the lines within which its use is acceptable are so narrow.

I have pushed it to what I feel should be its limit on my weblog, The Dredge - purely because I want to show off its potential. I believe the same is true at your site Jeff (although as Egor points out, the anti-aliasing in the smaller fonts is very poor - I think you may be better off going back to FIR in your side bar :)).

However, I have a worrying feeling that others will not hold back quite so much, and sIFR will be up for more bad press purely because it is not being used well. And that is the point Egor is making I think: It will be misused.

And I'll tell you something else: If you don't implement it *exactly* correctly, it doesn't work! That's gonna be a heck of a shock to a few cowboys out there! ;) ...I should know ;)

By Phil Sherry
01/13/2005 10:19 AM
Liverpool, UK
I think what you really mean to say is: Flash Sucks

Attacking Shaun Inman's hard work for a problem that lies with Macromedia - as you pointed out - isn't particularly professional of you, or anyone else.
By Egor Kloos
01/13/2005 11:11 AM
Rotterdam
> ... if you generate the images server-side, you also don?t get to kern the letters.
Both have their good and bad point also both are typographically poor and so my objection still stands. Less so if either technique involved the generation of words instead of characters.

> ... don?t you think it?s possible to abuse type no matter which technology you choose?
Yes, however I did feel that an additional cautionary note was in order. And some technologies are more redundant than others. Choose carefully and only when you need to.

> Attacking Shaun Inman?s hard work...
Shaun Inman is the man. I wouldn?t dare knock his work. Maybe you read over the part where I mention the sIFR is a good technique for solving a certain problem and that it highlights another problem with typography and the Internet. I hardy think that?s Inman?s fault.
By Phil Sherry
01/13/2005 11:23 AM
Liverpool, UK
> I hardy think that?s Inman?s fault.

...which is why it's not sIFR that sucks - it's Flash, typography and the Internet that suck ;)
By Rogier
01/13/2005 11:42 AM
Netherlands
> "...it?s Flash, typography and the Internet that suck"

No, it's the people who use them incorrectly that suck :P
By Chris Hester
01/13/2005 12:13 PM
Yorkshire
"Shaun Inman is the man. I wouldn't dare knock his work."

Except you did. By the very title of this post.

What I like about sIFR:
You can copy part or all of the lettering with the mouse. This is a major bonus. Images cannot compete.

What I don't like about sIFR:
It can be slow to load.

Despite it's faults, sIFR is surely outstanding work. Macromedia may well improve their anti-aliasing over time, when sIFR will be ready for them. And who is to say that Mike isn't working on a new version that allows kerning? This is just the beginning!
By Egor Kloos
01/13/2005 02:10 PM
Rotterdam
> Except you did. By the very title of this post.

Well maybe I did a bit of baiting. However it's doesn't imply that sIFR itself is bad but that it doesn't provide the web with improved typography just additional typefaces.
By Jeff Croft
01/13/2005 06:20 PM
Reppin' KC MO, Fellas and the ladies know.
> it doesn?t provide the web with improved typography just additional typefaces.

To some degree, I concur. sIFR (or any Flash-based text) is certainly not God's gift to typographers, but I still maintain that it's a marked improvment over anything we've had before.

Like it or not, typeface choice is part of typography, and this gives us a great deal of selection. Flash does anti-alias text, which Windows doesn't do (by default). True, it's not the most perfect anti-aliasing ever, but it's still better than nothing. The next version of Flash is slated to have much, much better anti-aliasing.

In closing, I don't think that Shaun, Mike, Mark, or anyone else involved in sIFR ever <em>claimed</em> it was a robust tool for rich typography. They only claimed it will let you render XHTML text in Flash using the typeface of your choice -- and it does that brilliantly.
By cheekygeek
01/13/2005 08:15 PM
You've chosen a sensational sounding headline that has very little to do with the point(s) you've made in your article. sIFR can be misused. So can a hammer. So hammers suck?

Don't condemn the tool when the problem is inept tool USERS.

I don't have a problem in utilizing a tool within narrow parameters (replacing only larger headlines, with font.swf files that don't have kerning issues) if it is better than what I had before for that purpose and if it degrades gracefully.

I'm impressed with what I see on sites that use it WELL and I'm determined to learn to use it appropriately.

sIFR is waaaaay cool and I'm disappointed that you've chosen such a sensational sounding headline to make so many points that have nothing to do with sIFR itself.
By Mike D.
01/13/2005 09:07 PM
Seatown
Ok, before we get into the subjectives, let's start with a few facts:

1. You mentioned you didn't like Flash's anti-aliasing. Flash 8 will have an entirely new anti-aliasing engine codenamed "Saffron" which will be almost as good as OS X's quartz anti-aliasing (and waaaaay better than ClearType). All existing sIFR implementations will automatically enjoy this better anti-aliasing as soon as people download Flash 8.
By Mike D.
01/13/2005 09:07 PM
Seatown
2. You seem very adamant about wanting to adjust the tracking on your typefaces. If it's that important to you, you can always just open up your typeface in Fontographer or any other such font editing tool and adjust the default kerning yourself. It's not that difficult.

3. You seem to be comparing a sIFR workflow to a workflow where a designer manually sets display type for every page that gets published. If you run a site which publishes so infrequently and you don't mind opening up Photoshop or Illustrator and manually setting your type for every article, then keep doing that. I think that 99% of sites out there are not in that boat though. Automation is an inescapable aspect of web publishing and sIFR is an <em>automated</em> way to improve your typography.
By Mike D.
01/13/2005 09:08 PM
Seatown
And now on to some subjective stuff:

I feel like the crux of your complaint with sIFR is the democratization it provides in the design process. Your argument is not just that trained typographers are <em>better suited</em> to make type decisions than plebians (which is undoubtedly true), but that <em>only</em> trained typographers should even be allowed to make these decisions. It's the same argument that traditional typesetters made when desktop publishing became popular. It's the same argument that tax accountants made when TurboTax got popular. It's the same argument that always comes from the experts when power is put into the hands of the masses... no matter what the subject. I'm not disagreeing with you that in a perfect world, trained designers would set all type, but there simply aren't enough trained designers to keep up with the raw amount of material being published on the web these days. That is half the phenomenon of the web itself: the ability for anyone to publish. If you want to say that sIFR can be abused, I wouldn't argue that. But so can HTML and every other web technology. There are standards purists out there who believe that if you can't write perfect HTML 100% of the time, you shouldn't even be allowed to publish on the web. Now that's just silly.
By Mike D.
01/13/2005 09:08 PM
Seatown
On to anti-aliasing. I personally don't have a problem with Flash's anti-aliasing because I don't use it at tiny sizes. At the sizes I use it at (20-point or higher), it looks marginally worse than Quartz and much better than ClearType. If you have a problem with how 12-point anti-aliased text looks in Flash, then don't use it! It's like the old joke about the patient and the doctor --

Patient: "Doctor, it hurts when I bang my head against the wall."

Doctor: "Then don't bang your head against the wall."

Pretty simply really.

And finally, on to image-replacement. I don't see how image-replacement has any advantage over sIFR unless you are talking about an entirely manual workflow where you are physically crafting each image in Photoshop every time. Under that circumstance, you have more typographical control, but you don't have any of the advantages sIFR has like zoomable, selectable, and copyable text.

Anyway, that's about all I have for now. And since we're talking about typography, I thought I'd point out that in Safari, I'm noticing a lot of overlapping text and lack of paragraph breaks on this page. For instance, the entire entry comes through as one paragraph and in the comments section, where it says "Comment by _____" is overlapping the first line of the comment. Also the grey footer text is laying right on the bottom border of the page. Probably just a quick CSS thing but I thought I'd let you know.

Cheers,

Mike

P.S. I also just tried to post this comment and I got an error message saying "The maximum number of characters in a comment is 15. This comment contains 4190." ???? I'll try to split it up a little but hopefully that 15 character thing is a typo.
By Jeff Croft
01/13/2005 09:19 PM
Reppin' KC MO, Fellas and the ladies know.
Sometimes I wish someone would actually force a 15 character limit on Mike D. :)
By Joe Grossberg
01/13/2005 09:37 PM
Arlington, VA
"IT?S noted drawback is ITS?S inability to allow ... IT?S poor anti-aliasing."

In all those cases, you mean "its" (sans apostrophe).

Sorry for the pedantic nitpick, but one man's kerning is another man's punctuation.
By Mike D.
01/13/2005 09:55 PM
Seatown
Croftie: Here's my homage to you, in haiku form (less than 15 words):

Oh Football Genius
Such a Shame To Suck So Bad
You're Worse than the Chiefs

:)
By Joe Clark
01/14/2005 12:07 AM
Toronto, obviously
There are standards purists out there who believe that if you can?t write perfect HTML 100% of the time, you shouldn?t even be allowed to publish on the web. Now that?s just silly.

And it's something I cannot recall reading even once. You're not producing Web pages in that case (you're producing something else), but the Web is big enough to take what you give it anyway, no matter how mangled.
By Mike D.
01/14/2005 12:30 AM
Seatown
Hi Joe. I didn't actually mean you, by that statement. And in fact, I think your latest statement that "If your site has valid code or something trivially close to same, you are working with, and within, Web standards." is more or less right on.

What I'm talking about are the people who compare HTML to a compiled language. A typical line you hear from these sorts of people is "If you have an error in C++, your stuff won't compile. Period." Mmmmkay. Well, this isn't C++... there's leniency built in and my feeling is that that leniency won't be going away anytime soon. That doesn't mean you should let the leniency turn you into a lazy fool when it comes to coding... it just means you shouldn't expect every single person who publishes content on the web to be error-free. Teach them why it's best to make as few errors as possible, give them the tools to prevent these errors, and then leave them be. That's how I feel about the situation.
By Jason Santa Maria
01/14/2005 01:08 AM
Philadelphia
Well, Mike already put to words what I was going to say... albeit a few hundred more words. There are plenty of problems with ALL web technologies. Its creative people like Inman and Davidson who find ways around them with stuff like this. Why don't you re-title the article "Flash sucks" because it doesn't seem like your problem is with sIFR, its with the eccentricities inherent in Flash with using such a technique.
By luke
01/14/2005 01:35 AM
It's always nice to see another side to the story, even the opposite side. Although I don't agree that sIFR sucks, you make some good points. The web is young, and I'm sure it will evolve to great heights as the years come, however, no one makes you use sIFR and people who do should know to use it with limitations. I think when it's overly used it can become tacky, but site's like shauninman.com use it with a beauty that should be appreciated. He's created something original, and it's always nice to see something unique on the web these days.

By the way, this is also a very nice site.
By Egor Kloos
01/14/2005 02:05 AM
Rotterdam
I'm not a fan of renaming title after it's been published. Even if I was over zealous in choosing such a title. So I'll leave it as is.

Everybody can debate whether it should be something else. That what the comments are for.

What I meant to say with this post may not have been clearly put by being slightly ambiguous and the harsh ending doesn't help either. Thank God I'm not a copywriter ;)

I don't have that much of a problem with the kerning or with Flash. As I've mentioned in my comments above. Replacing typography this way is certainly handy and I'll more than likely make use of it myself. Especially after reviewing the third release candidate, pretty nifty coding.

To call it a major advancement is what I find questionalble. An advancement yes, but some seem think it will change the web. Well I doubt that very much. No matter what everybody says this is in the end not a structural addition for Web typography but a solution for a very specific issue relating to the lack of typefaces. I don't think it could hurt to turn the hype down a notch.
By Egor Kloos
01/14/2005 02:11 AM
Rotterdam
Mike, I haven't been able to reproduce the promblems with Safari you mentioned. This site was initially designed on a Mac and built for Safari. If anybody else has these problems with Safari let me know. So I can get my lazy ass on the case.

btw. I'm now a Firefox convert.
By John
01/14/2005 02:21 AM
Australia
sIFR has hit the scene in a big way and I note it is still Beta.

I am very impressed with what some are doing, considering the limitations of MM Flash...
Eg.. http://www.disenorama.com/

More power to sIFR's future development and the constructive input that many have given to the project to date. I wish sIFR a happy 2005.

By Jeff Croft
01/14/2005 03:40 AM
Kansas City
Egor-

I don't really believe that sIFR itself will "change the web," but I do believe that these types of ad hoc solutions will eventually for the W3C and other major players into giving us a standard way to design with a healthy typeface choice on the web, just as we can in print.

sIFR isn't the end -- it's just a means to get there.
By Chris Hester
01/14/2005 12:09 PM
Yorkshire
The point is also that sIFR (can't we have a better name for that, Mike?) is still in its infancy. So imagine how great it will be in later versions. Anything is better than a fixed choice of a few average typefaces.
By nat bolton
01/14/2005 08:19 PM
"Merely adding your own typeface per character as if you were block printing doesn?t solve the current Web typography problem we?ve all been stuck with for this past decade"

I must have missed it when people ran out in the streets proclaming sIFR the savior of web typography. Seriously, who ever said this solved the whole web typography problem?
By Brent O'Connor
01/16/2005 05:21 PM
Wow, nice little geek war! :)
By Bryan
01/24/2005 11:55 PM
It's lipstick on a pig...
By Mike Steinbaugh
02/03/2005 07:16 AM
Los Angeles, CA
Just a thought, but I don't think the title matches up with the post very well. Maybe "Lack of typeface embedding sucks" or "Where are the fonts?" but "sIFR sucks" just sounds inflammatory to me. I'm sure Shaun isn't offended, but still, it seems out of place.
By Kris
02/13/2005 10:33 AM
"The lack of typography will hurt the WWW in ways we don?t really understand yet. It may actually kill the WWW as we know it."

Of course not. What an egocentric thing to say, for a designer.
By Charles
07/11/2005 12:12 PM
UK
Have a look at this:

http://www.fontweaver.com

Just been released.

It adresses other problems such as word wrapping and requires no plugins.

Kerning is being worked on.

What do you think?
By Egor Kloos
07/11/2005 12:47 PM
Rotterdam, The Netherlands, Europe, Earth etc...
Fontweaver:
"Fontweaver currently runs on the Windows operating system only"

You've got to be kidding right!? Right out of the box this solution is useless to me, pitty. Also if the pages on your site are anything to go by it would seem to be an even less desirable solution due to the mass of extra HTML never mind the lack of semantics. Doubling up you code is pretty horrid.

How would such a solution tie into the CMS? sIFR doesn't require extra attention when constructing the templates, at least not so much as this solution does. Sorry, but the sIFR solutions has still got this one beat. Fontweaver is in my view even more of a hack than sIFR is. I'll keep my eye on the project but I'll never use this as it is now.
By Charles
07/11/2005 01:11 PM
UK
I hear where you're coming from Egor.

If I may address your points.

Windows: We're planning a Mac version but have not the resources to do so right now.

Doubling up code and CMS: FW will, pretty soon, have the ability to work with automatic replacement of content in CSS classes. We've got proof of concept working, but are still tidying it up. It'll work differently to the existing methods because all you'll have to do is append a Fontweaver Font name to the "font-family" attribute in whatever class you wish in order to make it work. That's what we're trying to achieve, anyway.

We're trying to give as many options as possible to users. There are a bunch of plugins in the works for other dynamic stuff that uses JavaScript and, eventually, the ability to plug your own JavaScript functions (such as menu/calender generation) into the wizards.

Don't forget, it's very new.

All comments appreciated.
By anonymous
01/01/2006 11:25 PM
Check out the discussion here:

http://aasted.org/adblock/viewtopic.php?t=1687
Commenting is not available in this channel entry.