*sigh*. I am so angry at Adobe. I've used Photoshop for some years now and have just purchased Adobe Photoshop Lightroom - wicked awesome, by the way. When I bought Lightroom it was part of a package deal - pre-purchase Adobe CS3 and purchase a version 1.0 Adobe Lightroom for a special low price. What does any of that have to do with garamond?

 

Adobe decided to bundle a special professional version of the Garamond fonts called Garamond Premier Pro. The font is named after Claude Garamond - a 16th century French publisher and typeface designer. Remember, back in those days you actually had to create a stamp, or punch for each character in each size to actually be able to print something.

 

So anyway, Adobe has a thing for Garamond, I guess, because in 1998 they sent a guy out to Antwerpen, Belgium to study the metal punches that still remain from Mr. Garamond at the Plantin-Moretus Museum. Now, as an old TEX user, I can attest to the fact that there is such a thing as a font-geek (how can you not love this site, the tagline is: it's all fun and games until someone puts their eye out with a stray descender). What this guy (his name was Bob) did was produce a matched set of fonts for use in different sizes and presentation contexts. Let me just expand on that a bit, because it's extremely relevent to my anger at Adobe.

 

One concept that Bob captured in the Garamond Premier Pro font set is that any character in the set has to be drawn differently depending on its size, whether it is bold or regular stroke, or italic. There are additional nuances, but let's not go into them here. Here's an example using Times New Roman, which is another serif-styled font like garamond, but a bit more common:

This is simply the same character repeated twice, once as a SIZE=1 (smaller, left-hand character) and once as a SIZE=7 (larger, right-hand character). The serifs are the decorative little strokes at the end of the lines (like the tops and bottoms of the H). The question is, should the actual shape of the letter be the same regardless of size?

 

The answer is no. You don't necessarily perceive both H's above to be identical, size aside, particularly if you look at the middle horizontal stroke in the H. If you really want to have fun with this, look at the page using the smallest, medium, or largest sizes as set by the browser. To do this, you use the menu to select  View -> Text Size -> Smallest, for example. I believe that the font is just scaled up and down from the same basic shape in this example and the height of that horizontal stroke in the H moves around based on how it scales and your perception of whether or not it is in the middle. This is just a basic perceptual thing, I think, but guys like Bob overcome it by establishing a set of size-bands for fonts. Think small (six picas), medium (12 picas) and large (20+ picas). The image at the top right, with Hkg repeated is an example of how the Garamond Premier Pro font solves this problem. The characters on the left are presentation size, I believe, which means large fonts. The characters on the right are for small type (five to eight picas) but blown up to be the same size. You can see that the serifs are more delicate. Maybe so, but when the font is at proper size (small), you percieve the match between those characters to be better than using the exact same character shape and weight but just scaling it up or down.

 

Now that you have trained up to be a typesetter, let's throw in the next loop. I had to write a letter on Thursday. I mean a real, honest to goodness business letter on letterhead with a signature and everything. Remember those? And I thought, great, I can throw garamond on and it will really sing. There was a risk to this and I was aware of it because I had actually read the release notes that came with the font. They talked about how the number of glyphs in each set were large enough that you shouldn't use more than two or three different styles of garamond on the same page. If you did that and your printer had only 32M of memory or less you were in for trouble. I accepted the risk, knowing that my kit is pretty good.

 

Oh, what a mistake that was. In retrospect there were two things happening at the same time that were really insidiously bad. The first was that they were right about the fonts. You really didn't want to put too many in a document. My printer didn't have that much memory, as it turned out, and so it really choked on the document. The second problem was the inability of the computer to deal with this condition. It (the printer) seemed unable to tell the computer that the job was cancelled at the front-panel (which is one of the many ways I tried to kill it). So then the computer would resubmit the print job. This persists even after a reboot because the printer can't complete the job. Then, if you try to delete it from the print spooler on the computer, it hangs. Eventually I couldn't even open up the printer control panel without a hang. I couldn't open up Word without a hang. It was terrible. Furthermore, I had installed the fonts on both my server and laptop computers. So I had both computers unusable for word processing all of the sudden and the printer in a questionable state. I lost an entire day of productivity. The solution was to continue to cancel the job on the printer as it appeared and to reboot and purge the file and print job continually. Eventually it settled down.

 

Adobe, don't do me any more favors. There should have been a much stronger warning in those release notes - this is not a font for the casual user. Don't get me wrong, it looks absolutely great on paper, but you really need to know what you are doing and have a serious printer to run this font.

Posted for a friend who asked about this last night over dinner at Olé. The French architect Jean-Pierre Houdin has used some nifty modelling software from Dassault Systemes to propose a very interesting theory about the construction of the Great Pyramid of Giza. You get to explore this theory via a super-flashy online multi-media show in the comfort of your own home. (Warning, this is a heavy-duty show involving a 3D visualization plug-in.) Of course, there was a book deal too.

 

I know, I know, it all sounds like a late April-fools joke. And yes, if you are curious, Houdin does seem to be the surname of Harry Houdini. I think it just adds character to the story. Jean-Pierre makes an appearance too, as you can see, narrating the story via CGI. I actually thought that was kindof cool.

This is about the MBTA, which has made a few questionable decisions over the years. Back in the day, the fare collection was backwards, you had to pay when you got off the train. This led to the ridiculous condition that you could get on the T and ride it somewhere and then have trouble getting off. Back in the day, though, it was just the MTA, they hadn't added the B yet (Bay), and the song Charlie on the MTA immortalized the fare problem.

 

Later, during the the seventies, we were saddled with really finicky light rail vehicles (LRVs) on the Green Line.

 

Now, though, we have Charlie Tickets. I don't know where to begin. Well, hopefully all of the stations actually have readers for the Charlie tickets at this point. I've watched the progress of this system as I've moved back and forth between London and Boston. When I visited Boston the first time the Charlie ticket was available, I happily purchased a $10 ticket only to discover that they hadn't actually put Charlie-enabled turnstyles in all the stations.

 

As if to overcompensate for the old pay-on-exit deal, Boston stopped requiring anything but a controlled exit from train stations. Thus the exit protocol, if you will, is different depending on the station you are in. In some stations, you may exit through a ticket-taking turnstyle (although you do not have to scan the ticket to exit the station, just walk through it), or a one-way revolving metal gate, etc. This isn't particularly important, necessarily, except to illustrate that there's a bit of chaos in the system. For example, at Haymarket yesterday, there was no path to channel people entering and exiting the station. It was a mess trying to walk through and around the streams of people trying to go through the turnstyles in both directions. And then there is buying a Charlie ticket in the first place.

 

I've been in high tech for 15 years and the first time I tried I had no idea what was going on. It was that bad. I ended up buying a single-day trip car for $5 when I was trying to buy a stored-value card for $5. The interface was terrible. It's still pretty bad, but at least I could figure it out this time around.

 

So what's with a 'stored-value' card? It's a industry term that never should have gotten on the the printed card, the ticket kiosk interface or anything else, in my opinion. If you go to the Charlie page, they talked about stored-value and clearly this is the term they've decided to use.

 

I think at this point we can trust the general population to understand the concept of a debit card. Now, if it were me, and clearly it's not, I wouldn't even have a separate designation. You have Charlie tickets and Charlie cards. A Charlie card is a debit card for frequent riders. A ticket is a day ticket for infrequent riders. Done. But I wouldn't call the Charlie card a stored-value card or a debit card. I'd call it a Charlie card.

Google is a powerhouse, no doubt. But there are a lot of other companies doing Internet search applications that are arguably better and more comprehensive than Google. What's up with that? Let me take you through three interesting companies in the search space that might help you out.

 

First, a few concepts.

 

A search engine like Google works by literally visiting every web site it can identify and building an index of the words, images and other content it finds there.

 

A meta-search engine is a search engine that uses other search engines as the source of information rather than actually 'crawling' the web. When you enter a search term, it fires off searches on Google, Yahoo and MSN (for example, and hypothetically) and then presents all those results to you. That would be a meta-search.

 

Podzinger is a search engine for podcasts, YouTube and other media files on the Internet. BBN has a long history of innovation in the field of speech recognition and the marriage of their speech technology with media content on the web is exciting. Plus, it's a great place to search for cool podcasts. You can also check out the Zing Index - it shows who and what is being zinged.  Now, when you put up a podcast or YouTube video or other media file there is typical metadata associated with the event. For example, there might be a program description with a list of guests. This stuff is easily searchable, but Podzinger goes much deeper. It actually transcribes the entire event and creates a searchable index from that. One word on usage - if you are looking for a proper name (i.e., Brad Delp), put quotations around it so that you are searching for exactly "Brad Delp", and not everything with a Brad and everything with a Delp in it. Big difference.

 

ChaCha is a meta-search engine that has guides to back up the search. This sounds simple and obvious, but it's extremely powerful. If you do a basic ChaCha search, you will get results that have been focused based on the guides. If you can't find what you are looking for, you ask to connect to a guide. At that point, you begin a two-way chat with a guide who is an experienced Internet searcher in your domain of interest (i.e., home crafts or photography and so forth). As you tell this guide, via the chat session, what you are trying to find, the guide is off looking and popping search results into your browser. It's really a very compelling thing for those of us who just aren't good at finding things on the Internet. There are a lot of people in that category and many of them are too embarrased to admit it. Searching, googling, whatever, can be hard. Anyway, there is one more dimension to this. The guides get paid between $5 and $10 per hour to help guide. It's a great work from home occupation for someone with some Internet savvy.

 

ZoomInfo is a semantic search engine for business information. It is a true search engine - it crawls the web - but it is searching for specific types of information: business information. Zoominfo has information on over 35 million business people including just under 1.5 million board members, 120,000 laywers and 1,200 mimes. By semantic search, they mean that they are able to understand some of the context and relationships in the content they are indexing. This is then used to build a rich profile of an individual or company from multiple sources. The basic search function is free - you get access to this huge database of business information.

 

Those of us that have been around since the start of the widespread adoption of the Internet in the mid-nineties know that Google was not the first or even the best search engine. For a number of years the leader in both of those categories was actually Alta-Vista.