Skip to: Navigation | Content | Sidebar | Footer


Full Archives

Mapping on the Web

March 31
I’m currently involved in a project that deals with large–format maps targetted towards high–resolution HP plotters. The finished maps are around 4 feet on both sides. Somehow, these incredibly huge pieces of paper have to be neatly represented on a low–resolution 15 inch monitor. Unless you’re running a custom CGI setup and have access to greatly–detailed vector maps, ala MapQuest, the job of converting a static map to something that’s actually usable on the web is enough to make you lose sleep. First problem: text. At 11 or 12 points, text on the printed map is perfectly legible. When you start shrinking it down even a bit you begin losing that, and when you shrink it down a lot, it’s hopeless. Second problem: detail. Land is complicated and intricate, and maps tend to reflect this. Third problem: file size. The detail you need to capture will choke most image compression methods, despite their best efforts to reduce it. In a perfect world every client that requires maps would be able to offer vector source to work from, and flexibility that allows you to use a number of options to present the material, ranging from SVG to PDF. If you live in this perfect world, please, send me a post card. The rest of us are stuck with reality, however, and we have to make the best of a bad situation. This involves multiple versions of the same map for previews, large format views, and detailed views. Possibly the inclusion of a PDF or two for printing. In this particular case, the client recognized the problem at the onset and with a detailed guide from me, were able to ask their cartographer to generate custom maps that we'd be able to work with. The 6000–odd pixel wide JPGs I ended up receiving were still way too large, so they had to be cut back further. The solution was to present a tiny, reduced version as a preview, which makes no attempt to represent any detail. The preview is clickable, and loads a 2000 pixel wide JPG in a new browser window that an end user may scroll across. As well, 8"x11" PDFs are offered for print usage, and since they do need to target their high–end users, HP–plotter specific high–resolution maps are linked. The client understands that at 500 pixels, the preview map in no way represents the end map other than to serve as a quick overview, almost an icon. The end user is specifically told that a few larger versions exist and are easily accessible, and the solution appears to be working. These type of hacks are regrettable, but in many cases necessary. Custom scripts are not available in every budget, and in many cases the source material is less than optimal for the end use. But a bit of creative thinking and adaptability can come up with a solution that works.

Permalink › | no comments

Too Much Load

March 25
It strikes me as fundamentally wrong that success on the web can bite you in the ass. The direct server–to–client model is great for keeping information centralized and under your control. But when your traffic starts edging upwards it becomes a make–or–break proposition to keep your site running. Bandwidth ain’t free, never has been, and probably never will be. You have to store your site on someone’s server, and keeping that computer up and running can sometimes be a full time job. Not to mention the electric bill. More and more sites are resorting to funding blitzes to stay alive, soliciting donations under the threat of closure. The micro–payment idea never took off, so what else is there? Not–for–profit sites can either saturate themselves with advertising, in a time when advertisers are paying next to nothing, or they can ask for subscription fees, one–time donations, and general support from their audience. I want to see a non–profit organization address this. I want to see a philanthropist who made their millions off the info–tech rush 5 years ago give back to the rest of us. Surely there’s someone out there who sold their dot–com (at the right time) that values the idea of a free web? Somebody who recognizes the good that people donating their time can bring to the rest of us? It will involve time and money. The time, this community can donate. I’m sure there’s no shortage of people willing to invest the time needed to set up a server location, configure the boxes and maintain them, and monitor the service. It’s the money we need, to pay for bandwidth, hardware, and office space. Alternatively, I’d be happy to see someone develop an alternate system for serving up the web — a peer to peer model that allows a central server to call the shots, content-wise, but refers most of the bandwidth on to associates. I'd be more than happy to lend out some of my unused processor/bandwidth time to support these sites. A Seti@Home for the independent web. Somebody make it happen! I refuse to believe that valuable independent publications must die due to something as ridiculous as a common infrastructure problem. This post dedicated to today’s traffic spike, courtesy of Mr. Z.

Permalink › | 2 comments

Discrepant English

March 24
forge — to create, to mold, to develop an original. forge — to counterfeit, to fake, to duplicate an original. How did one word end up meaning two opposite things? We speak such a quirky language, don’t we?

Permalink › | 1 comments

Hack Hotbot, Part III

March 20

As the world’s attention has been shifted elsewhere, not much of note has happened recently in the wide world of design. For the sake of having interesting new things on this site, here are a few notes from my own battle in the trenches: the Hack Hotbot contest. Hopefully they’ll be of some use to anyone joining me in this raving masochism.

  • Ugly code — When you invite people over to your home for dinner, it’s considered rude for them to speak up about your cleaning habits. However, when you invite someone into your home, you usually spend a few minutes frantically cleaning to make it presentable.

    I’m going to be a bit rude here, pardon my manners HotWired — yep, Hotbot is a mess. It was lauded for converting to a CSS layout a few months back, but I wonder how many people have had their hands in there since. You can almost see the different coding styles at work. Best advice: draw a sketch of the main structural elements, to give you an idea of which containers you can hack for various effect.

    Keep in mind that anything you do can only be done to the style sheet. There is no way to modify the HTML source, so consider it set in stone.

  • Hard–coded pixel values in tables — I said it was a mess, right? There are a few different spots where the tables are given set widths. I’m not against tables for tabular data, but these particular tables seem unnecessary. Not only that, but they are 600 pixel plus, so it limits the design side of things to pretty much one configuration: search box on top, search results/filters right below.

    I suspect the layouts of the new skins will be quite similar to each other. There isn’t much room to maneuver here, so variation will be more likely in the colour scheme, fonts, and background–images.

  • Extra style sheet — After all your hard work, there’s a further style sheet being called in called ‘or.css’ which over–writes a few elements in the original style sheet. This might not be a problem, except that some rules are marked !important — exactly the hack I had in mind to over–ride them.

    The main element I’ve clashed with so far has been the logo — but in the end I won. A bit of lateral thinking says, okay, it’s relatively positioned top and left, but that doesn’t mean it has to stay there. Top and Left positions aren’t the only thing determining where it sits on the page. Thanks to some crafty hacking of the margins and a few strategic !important placements, I’ve managed to move it to where I need it.

  • Shifting elements — So when you first load the page, you get an announcement table telling you about the Hack contest, which you can close. Which means that the search bar below will shift in place once you’ve closed it. Which means any background image you’ve placed to contain the two which relies on pixel values gets bunged up. Make sure to test with both the box open and closed.

    Once you close it, it doesn’t look like you can get it back without flushing your cookies. So I’d recommend Viewing source and saving a copy of the HTML to your hard drive to work from, to make sure you’re seeing the announcement window. Heck, I’d recommend that regardless, it saves time with all the FTPing back and forth.

  • Lack of containers — One thing I’ve had to face is that there aren’t many elements TO hack. You've got a few stand–alones sitting in the body, but nothing containing them. Which may or may not be a problem, depending on your design ideas. It certainly was with my own.

    Here’s a $50 tip that may or may not have occured to you: with CSS, it’s possible to over–ride most every element. Including the <form> element. You can even apply background images to it. wink wink nudge nudge.

For all my complaining, I have to give credit to the guys at HotWired — this was a great idea for a contest, and one I’ve jumped at. It takes guts to open up your source to the world, bugs and all.

I have to wonder about the entries though. I’m sure I will be blown away by what is accomplished with creative CSS and some non–linear thinking, but the bar seems rather high for entry to this. Not only do you have to be a code jockey, capable of making sense of the standards–soup, but you have to be a great designer to put it all together into something attractive. What we saw recently with the WThRemix contest was that the ‘CSS as structure’ crowd were gung–ho to throw their hat in the ring, but the ‘CSS as beauty’ crowd shied away.

This contest is for people who are both. How many of those are there, I wonder. I hope they put up a big page of entries once the contest is over, I’d love to browse and see if I’m totally out to lunch here or if my theories hold up.

Permalink › | 1 comments

Hack Hotbot, part II

March 13

Hey, I like the initiative that Wired and Hotbot have shown recently just as much as the rest of the web development community, so Terra–Lycos is okay in my books.

But I was a little jaded after learning that the Hack Hotbot contest isn’t open to the ‘world–wide’ half of the web, only U.S. citizens, so I posted about it. Well, it would seem our good friends at Terra–Lycos did a bit of surfing this afternoon to find out who was mentioning their contest. I had a small swarm of lycos.com activity around noon to one, the referral coming from what looks like an intranet search (powered by Inktomi) for “hack hotbot”. One of the pages hit was my contact form, and I was sent an e–mail:

Thank you for the HotBot mention within your site. We are running a skinning contest which may interest you. We are giving away a 57" TV as the grand prize. Your site tells me that you would make a mean HotBot skin for the contest.

Regards,

Lincoln Jackson
Sr. Director, Search Products

I was a little surprised, to say the least. I’ve spent a bit of time formulating my response, and here it is:

Well gee, Lincoln, I appreciate the kind words, but you didn’t even read the post you are referring to before auto–replying.

In fact, I specifically expressed my desire to enter the Hack Hotbot contest that ‘may interest me’, so I knew about it ahead of time. Which means this is probably a form letter, so I guess I can forgive the wording on that part.

But let’s go back to my post for a second. I mentioned that I’d love to enter the contest, except for the fact that in your official rules, point 6 (General Conditions) states that the “Contest is open to legal U.S. residents who are 18 years of age or older.” This means that I, as a Canadian, am not even allowed to enter. (we Canadians know better than to enter a contest without checking these things first, since a lot of American companies like to do that to us)

So here’s where we’re at: you decided the contest was important enough to take up my time telling me about it, but you didn’t think I was important enough to spend the time making sure I was eligible to enter. I’m a bit hurt Lincoln, and I’m sure you can see where I’m coming from.

But I’m a pretty reasonable guy, and I think we can work this out. Opening the rules to allow the international design community a fair chance would be highly appreciated not only by me, but by the rest of the designers who felt a bit slighted that they couldn't get a fair shot. Since we’re halfway through the contest, a deadline extension of a week would be great too.

Oh, sure, it might cost you a few extra bucks to ship the TV if somebody in the Netherlands takes the prize, but if you’ve already dealt with the logistics of getting a bigass TV to someone on the opposite coast, it really won’t be that much harder to get it to another country, will it?

I’ll await with baited breath your response,

Dave.

Let’s put the pressure on them to open this up. I want a shot at a damn iPod.

update — 9:21am, 3/14/2003: I’ve received a few responses from Terra/Lycos, including a great explanation from one of them who really does care about this, and regrets the way it was handled. The official word is that non–US citizens are recommended to enter the contest through US mailing addresses or in collaboration with US citizens.

I feel a bit hesitant though, since the reason it’s US–only to begin with comes from their legal department. I wonder if there will be any problems if the winner happens to be a non–US citizen entering through a third party.

Still, I have in writing from Lincoln that this is okay, so if anyone enters and has problems, get in touch with me after the contest.

Feel free to use the comments below to hook up with a US partner.

Permalink › | 3 comments

The Perils of Gamma

March 12
When designing for web, you keep in mind that everyone else will see your work differently than you. They may have more pixels or less, more contrast or less, and a completely different colour profile, guaranteed. This is something that has been bothering me more and more. At home, I design on a 15" LCD monitor at 1024x768. This I find somewhat constraining, but I love my monitor and won’t give it up for anything but a bigger LCD. At work I run a 17" CRT at 1600x1024 — an offbeat resolution, but it gives me the space I need. My work display makes everything smaller. This is generally a good thing, since my tendency is to use fonts at a smaller size than many are comfortable with. I can pretty safely assume that if I can read it, so can you. The difference that irks me though is colour. The LCD has a lighter gamma profile with slightly richer colour, resulting in flat, dull green–toned images when viewed on the CRT. There’s a definite greeny–yellow tinge to everything I create on the LCD, so it displays richer blues than the CRT. When designing on the CRT and viewing at home, colours don’t shift as much — images appear just a tad lighter. I could tweak and adjust and try to bring these two profiles closer together, but I’d just be shooting myself in the foot. I can’t predict to any certainty at all what the general public sees, since gamma can be so different monitor to monitor. The only acceptable way to deal with this, as with many other issues in web design, is to view your work on as many different system configurations as possible and just hope for the best. So my setup, while clunky for development, works. But it’s so painful. Take my recent redesign, for example. After laying out in Photoshop what I thought was a great new colour scheme, I made a point of sending a .jpg to work. Result? Ass. Complete ass. A few level/saturation adjustment layers later and I had what I wanted, give or take. Not being content until it was just so I spent the intervening month tweaking on and off. What was once an elegant, multi–layer .psd file grew to about 15 different source files with rebuilds of different graphics. Every new change seemed to result in a new file or two, since the tweaks were local rather than global changes. All in the name of better colour–matching. Some days I envy the print world. Other days I shrug it off and say hey, it’s all good. At least I’m not stuck futzing with traps and bleeds and linked files and postscript rendering problems, right?

Permalink › | no comments

Hack Hotbot! (Or don’t even bother)

March 11

The old Wired search engine, HotBot, recently made news by switching to an all–CSS design. Now they’re rippling the waters again with their latest news:

Hack HotBot is a unique design contest. We are asking our users to “HACK” HotBot's look & feel by creating their own customized skin.

Serious Designers Wanted. Faint at heart need not apply. Knowledge of CSS technology is necessary to enter. This must be your own design work. Copyright and trademark rules will be enforced. Winning entries may be offered to all HotBot users to select as their skin. Your design will be shown to the HotBot world.

Now this is a design contest I’d love to enter. The prizes are great (52" TV? iPod? I’m so there), and the recognition would also be sweet.

…except that they’ve completely failed to recognize the “World–Wide” part of the Web, and limited entrants to United States residents only. No such luck for this po’ boy from Vancouver, CANADA.

Permalink › | 1 comments

Worldwide.

March 10
Sometimes it’s interesting to take a step back and consider how ordinary the extraordinary has become. I have a box. In this box, I keep a collection of things that have some meaning to me; photographs, drawings, maybe the odd story or essay I’m proud of. This box is a collection of my treasures, and rather than hide it behind a tile in the wall, I’ve chosen to store it safely in public. And people see it. Some of them wander by and crack the lid. A brief glance is all they take, and they keep wandering. Others are more curious, and browse around the box and see what I’ve chosen to store in it. The occasional one will call a few of their friends to come and see what's in this box they found. What's amazing to me is from how far away these people travel to see my collection. Some come from Europe, others from the Middle East. Before the advance of the telephone, this would seem impossible. A hundred and fifty years ago, no one on the planet would believe my story. Even 50 years after Mr. Graham Bell made his famous call, the idea would still have seemed far–fetched. Only today when we have a world–wide network of information could this possibly seem mundane, ordinary even. I make this post to commemorate that today marks the last time zone. For all its faults, Sitemeter has a great feature that tracks time zones of the last 100 users — the usual suspects (GMT, EST, PST) have been out in full force all along, but the less–populated and less–developed (GMT+4, -12, and -2 for example) have eluded me. Thanks to a spike this morning, I’ve now collected all 24 of the major on–the–hour zones (:30’s haven’t been tracked). Which is commonplace for most of you, no doubt, but blows my mind.

Permalink › | no comments

WThRemix Entries In

March 8
Wow. ‘Underwhelmed’ comes to mind. The entries for the W3C redesign contest are in, and it’s pretty clear that the goals of the contest weren’t entirely met. The theory: redesign W3.org to be far more attractive, useful, and organized than its current incarnation. Not officially sanctioned by the organization itself, but still a worthwhile goal. In web design there are three camps these days: the amateur hack (who rarely cares about standards), the hardcore W3 geek (who rarely has a lick of design talent), and the designer (who creates beauty that seldom validates). The contest represents the second category in spades. It’s obvious the latter category, the people who were supposed to enter, didn’t. Let’s be clear — pure CSS and XHTML design can be quite effective, but due to the boxy nature of DIVs and the lack of diversity in available fonts, they all tend to be unified by a certain style. There’s a distinct look to CSS–based design, and while some are far better than others, it’s generally not hard to tell at a glance whether a site was mocked up in Photoshop first or coded from the ground up. So it seems that the designers this contest targetted collectively shrugged it off. Why? Lack of exposure? It hit K10k and others. Lack of interest? Quite possibly. Many designers would rather work with the uniform consistency of Flash than the tricky tinkering of CSS. Lack of payoff? The prizes weren’t exactly drop–what–you’re–doing–and–bang–off–an–entry calibre. Whatever the reason, it happened. Pretty high and mighty position for someone who didn’t submit an entry? Maybe, but am I wrong? I’d have liked to enter, in hindsight, but in my case I just had better things to do. There are a few great entries though, this one and this one are quite strong, and while busy, even this one is a step up from the rest. Familiar with the work of about half the judges on the list, I’d guess I have picked at least two of the top three finishers. My money is on Radu.

Permalink › | 2 comments

Photo Friday

March 6
This is fun.

Permalink › | no comments

Weblogs as Advertising

March 4
What Anil has been talking about is bound to fail. For now. Short summary: Project Blogger is looking to pay weblog publishers willing to hawk merchandise. It’s easy to get sucked into the “everybody is doing it” mentality that surrounds weblogs today. But they’re not. Poll 25 random people on the street. 1 or 2 of them might have a LiveJournal or a BlogSpot–hosted diary if you’re lucky. Even they probably haven’t heard of Movable Type or Blogger. Blogger has just reached the million users mark. That’s one million web logs, that’s not one million people. I’d guess a maximum 400k individual people. Are they all keeping active logs? I doubt it. If Blogger sees one hundred thousand posts a week system–wide I’d be impressed. Where am I going with all this? It should be obvious. The numbers aren’t attractive to advertisers. Why put up $10k or so on such a small audience, a technically savvy audience who can generally tell when they’re being marketed to? There may be a few high–profile attempts in the next few months, but the idea isn’t going to catch on. Yet. The weblog is still the playground of the technically savvy. This is changing though, and won’t be true for long (reference: all the 14 year olds posting about their crushes on LiveJournal). The masses will pick up on the idea sooner or later. The internet was a playground for geeks not too long ago, now everybody’s grandmother has an e–mail address. Even when the numbers start looking good enough to make this idea profitable, the people Project Blogger requires to deliver its message are, well, just people. Some will be far more interesting than most, and as with any medium a natural hierarchy will sort out and certain people will be considered authoritive. (reference: Jason Kottke) Those are the ones that will be most attractive to advertisers. And they’re also the ones who may have to pursue ways to pay off the higher bandwidth costs their popularity has caused. What better way than advertising? We’ve seen all this before. This is exactly what happened to early independent content publishers on the web, and today we happily accept that ads are a small price to pay for continued use of their sites. Since we’re so obviously not willing to pay. (reference: recent rumours of Salon’s yet–to–actually–happen closure) It’ll happen again with weblogs. And we’ll grumble, and moan, and accept. Weblogs were a shot in the arm for the increasing consolidation of major web content publishers, but they will inevitably follow the same path. This is just the start. (All numbers are blind guesses due to lack of information or the motivation to seek it. This is an opinion piece based on intuition rather than fact.)

Permalink › | no comments

Can we drop it?

March 3
I’ve been tracking Netscape 4.x usage the past 6 months on TheCounter.com. Since they track over 350 million visitors a month to various sites that use their service, it’s pretty safe to say they’re sampling a broad cross–section of the web–using population. Netscape 4 has lived entirely too long. Developers have been supporting it since 1997. As it approaches its sixth birthday, old coding practices are still rampant on the web as people accept that if even a small percentage of the population is using it, they must cater to it. In internet years where everything once moved 10 times the speed of sound, if not faster, this has been a considerable setback for people who want to move on to the latest and greatest. The good news: according to these stats, NS4 usage is at all time lows hovering around 1 percent. This means that less people use NS4 today than use a 640x480 resolution, a different percentage of the population that developers stopped catering to long ago. The logic governing NS4 support must therefore apply to 640x480 support if consistency is strived for, however most will completely ignore low–resolution monitors but still code for the broken browser. The stats for the past few months are encouraging. Based on over 350 million users, here’s from August, 2002 onwards: 8/02 - 10,549,936 (2%) 9/02 - 9,869,530 (2%) 10/02 - 9,507,201 (2%) 11/02 - 7,935,732 (2%) 12/02 - 6,008,601 (1%) 01/03 - 6,982,143 (1%) February’s results are skewed because they only aggregate ~115 million rather than 350 million. Increasing the numbers threefold, February weighs in at 6 million again. For a browser that has been around for so long to remain active has been a considerable bee in every commercial developer’s bonnet. We’re starting to see more developers drop it, and the more that follow suit, the more likely the hangers–on will finally upgrade as they see an increasingly broken web.

Permalink › | no comments