I just placed a bit of wisdom from Tycho Brahe at Penny Arcade in my quotes file. I credited them and put in the domain name as well. Then I got to wondering... I'm going to have this quotes file forever, effectively. Somehow I don't think they are going to be doing these comics forever. What happens five years or ten years down the road for all these sites? Obviously the corporate ones are going to stick around, but what of the others? What happens to The Morning News? What happens to Pray Station? It doesn't even get into the financial demands of running a popular site (not a problem here). People get bored. They move on. Good sites will disappear. Will they get replaced? Or will the steady demise of the open Internet reduce the flood of self-publishers to a trickle? I see no sign of this happening now, but what about when everyone has "been there, done that?" Right now there's new blood. Eventually the Internet will reach saturation. The novelty will be gone. So will my attribution of the quote be relevant ten years from now? We're all assuming a certain permanence, or even if we're not, we're thinking only a couple years into the future. All long-term thinking is on a large scale, not the small. So what happens in ten years? Who's to know?
For what it's worth, the quote was: If it doesn't have broadband, then it had God Damned Well better have an open bar.
I want an email client that changes my message settings based on rules. For example, when I post to a newsgroup, I want to make my email address ketan@REMOVEketan.org or something like that. I want to be able to easily switch sig files so I can have one for work and others for personal mail. That way possibly offensive quotes ("Pickup Line #11: Chick do now.") don't get sent to, say, a prospective employer. I think every significant piece of software should have a scripting engine so I can make it do things the way it should have done them in the first place. Python would be a good start. Emacs had the right idea (although they might have considered practicality when picking Lisp as their language of choice).
I've gotten asked where I get the weird links. It's not like I tried very hard; I stole them. The community of people who use the web the way people used it in the early middle days ('95 or so) has shrunk greatly (proportionally, of course). I mean, most of the people I know are early middle school web users (as in "old school," not as in 6th grade), but there are still some who follow the new pattern. Well, you follow one link to another to another, and pretty soon you're somewhere totally weird. But people generally don't do that anymore. There have been many statistics coming out over the last few years about the fact that an decreasingly (increasingly? I'm confused) smaller number of websites (MSN, AOL, Yahoo) is taking up an increasingly larger proportion of user time on the web. Part of that has to do with the sorts of users that are coming online. Part of that is that these big players are growing blob-like, adding more and more content and features in an effort to make it unnecessary to go anywhere else. I think, also, that part of the issue is (paradoxically) that there is too much selection out there. There are so many sites that one can like that eventually one's daily visit list becomes far too long to reasonably cover without one's eyes glazing over. I have some 25 sites in my daily list (of which I only have time to visit about 17, and I'm unemployed), another 15 regular (more frequently than weekly) comics, as well as 10 weekly sites. Then there's other, less regular ones, like reviews sites. Finally there are all the links to follow that I get from the above. It's a lot. For example, here's an easy way to find a bunch of decent weblogs. You can often find interesting weblogs by looking for weblog software and then seeing "these sites use us." They generally don't show off crappy sites, although if that's all they have, they will. But if you find, say, five sites for weblogging software (1, 2, 3, 4, 5), you can easily find links to sites using said software. And there goes your whole day.
I'm trying to move away from Blogger, see, into something a bit more full-featured. Unfortunately, some things are just a little too full-featured (slash). And some things aren't written in PHP. And some aren't Open Source (two things necessary for me to tinker with it). And so forth.
Gah. It's 4:04am. I've spent the last hour or more struggling with a malformed POST request. Tabs matter in headers, in case you ever need to implement any part of HTTP.
Mozilla is surprisingly good. I mean, wow. I expected it to still totally suck. Mind you, don't download the (AOL Time Warner) Netscape version. Not because I have a bias against that media conglomerate (which I do), but because there are irritating things about it that justify why I have a bias against that media conglomerate. For example, in the Netscape-branded version, you cannot disable javascript popups. When you install it, you also get these irritating "Free AOL and Unlimited Internet" shortcuts all over the place. And they demand you get a Netscape user id. So get Mozilla instead: same great taste, a whole hell of a lot less of a pain in the ass.
This weekend I made this recipe:
8 tablespoons (1 stick) unsalted butter, plus more for dish
6 slices good white bread, crusts removed, torn into 1/4- to1/2-inch pieces
I used some bad HEB baguette that I bought to see if La Madeleine baguettes were worth the extra $0.80. They were. So I heated up the baguette in my toaster oven and threw it into my food processor to make many small crispy crumbs instead of the softer, larger pieces specified above. I think you should go my way. Is better.
5 1/2cups milk
I used soy milk. Worked fine.
1/2 cup all-purpose flour
2 teaspoons salt
You'll probably want to put more on after it's done.
1/4 teaspoon freshly grated nutmeg
1/4 teaspoon freshly ground black pepper
1/4 teaspoon cayenne pepper, or to taste
Make that a slightly heaping 1/4 teaspoon. Don't be too approximate with this, though, because that's close to perfect.
4 1/2cups grated sharp white cheddar cheese (about 18 ounces)
I have strong feelings on cheddar. It must be white. Maybe a little yellow. Definitely not orange. That's not natural. The only sharp cheddar I could find was Vermont Cabot block cheese, which I happily grated myself (realizing with a "doh" when I finished that I could have used the food processor).
2cups grated Gruyére cheese (about 8 ounces) or 1 1/4 cups grated Pecorino Romano cheese (about 5 ounces)
I used the Romano cheese. Buy the block in your supermarket's fancy, expensive-as-hell cheese section.
1pound elbow macaroni
Yeah, ok, so let's do this thing.
Heat the oven to 375°. Butter a 3-quart casserole dish; set aside. Place bread in a medium bowl. In a small saucepan over medium heat, melt 2 tablespoons butter. Pour butter into the bowl with bread, and toss. Set bread crumbs aside.
I didn't have enough melted butter, so I put in a couple teaspoons of olive oil. Didn't seem to hurt.
In a medium saucepan set over medium heat, heat milk. Melt remaining 6 tablespoons butter in a high-sided skillet over medium heat. When butter bubbles, add flour. Cook, whisking, 1 minute.
While whisking, slowly pour in hot milk. Continue cooking, whisking constantly, until the mixture bubbles and becomes thick.
Remove pan from heat. Stir in salt, nutmeg, black pepper, cayenne pepper, 3 cups cheddar cheese, and 1 1/2 cups Gruyére or 1 cup Pecorino Romano; set cheese sauce aside.
Fill a large saucepan with water; bring to a boil. Add macaroni; cook 2 to 3 minutes less than manufacturer's directions, until the outside of pasta is cooked and the inside is underdone. (Different brands of macaroni cook at different rates; be sure to read the instructions.) Transfer macaroni to a colander, rinse under cold running water, and drain well. Stir macaroni into the reserved cheese sauce
Pour mixture into prepared dish. Sprinkle remaining 1 1/2 cups cheddar cheese, 1/2 cup Gruyére or 1/4 cup Pecorino Romano, and bread crumbs over top. Bake until browned on top, about 30 minutes. Transfer dish to a wire rack to cool 5 minutes; serve hot.
Ok, so don't do it this way unless you're going to feed like 8 people immediately. This might not sound like a lot, but this is heavy. I've got about 40% of it left after about 5 servings. And you don't want to leave this sitting around because it gets soggy. It's ok if you bake it warm again, but most definitely don't microwave it. Instead, do this: get a 1 quart container and put in a third of the macaroni. Ideally your 1 quart container will be fairly shallow (~2"). That way when you spread the remaining cheeses and breadcrumbs over the top, they get distributed well. I made this in a deep, round dish. It came out well, but it would have been better in a shallower and wider dish of the sort that I do not have. Anyway, refrigerate the remaining 2/3 of the mix. Try to just bake it as necessary. You do all this, and you'll have a macaroni and cheese that tastes so good, you'll call up your state attorney to fine Kraft millions for false advertising. What they make is not macaroni and cheese. I mean, going from the box, you'd think they'd make it with cheese, and not crap. And yet. Theirs sucks. Make this. It's easy. It's better.
The following is the text of a letter I intend to send to my House Representative and Senators concerning the Consumer Broadband and Digital Television Promotion Act, or the CBDTPA, introduced yesterday in the Senate by Senator Ernest "Fritz" Hollings, D-SC.
[edited several times because it was just too long]
I am a software developer in Austin, in the 10th Congressional district. I am writing to express my strong objections to SB 2048, the Consumer Broadband and Digital Television Promotion Act, or the CBDTPA. I am gravely concerned that this bill, if made law, would do great harm to the average American citizen. I believe the bill sacrifices the good of the American people for the good of a relatively small number of media corporations.
My most serious concern with this bill is the lack of protection it affords for the American citizen. While it contains language meant to protect individual consumers, I fear the protections as codified in this bill are neither strong enough nor explicit enough to guarantee my rights under the fair use doctrine, among other rights. Beyond simply stating that fair use should be protected, in order to make any sort of copyright control mechanism acceptable, Congress should pass a law affirming what the rights of consumers are. For example, as a result of Supreme Court ruling in 1986 in the Betamax case, the government cannot take away the right to "time shift" a recording of a television program. However, there is also nothing preventing the technology industry from taking that right away through technical means outside the scope of the law. Before any control mechanisms are implemented in technological devices, it is imperative that Congress guarantee the basic fair use rights of Americans.
By the provisions intended by supporters of the bill, all electronic devices would have mechanisms to verify that all content has been properly acquired. In effect, televisions, computers, compact disc players, and many other electronic devices will actively watching Americans as though they were criminals. These devices would implicitly assume that Americans are thieves and that theft can only be prevented by installing monitoring devices in every American home. This is contrary to American values. The device is not the crime, piracy is. In order to protect the interests of all parties, we must maintain the legality of electronic devices while prosecuting their misuse. Criminalize the deed, not the device. Appropriate laws already exist without compromising the rights of Americans. Furthermore, any digital rights management technology would require registration of individual users, an unprecedented invasion of privacy for entertainment.
We must also consider the impact to the nation's technology industry, widely recognized as one of the most important sectors of the United States economy. The entertainment industry grossed roughly on the order of $40 billion in 2001. By comparison, the combined software, electronics, computer hardware, and telecommunications industries are estimated to have grossed on the order of $500 billion last year, or over ten times the size of the entertainment industry. In spite of this tremendous disparity, we find the peculiar situation of the tail wagging the dog. This bill means considerable harm to the technology industry, as attested to the Senate by Leslie Vadasz, co-founder of Intel, one of the largest semiconductor manufacturers in the United States and the world. He testified as to the damage to the industry should the federal government mandate technological specifications to solve the problem of piracy. The American businessmen and women who have made this country's economy the strongest in the world will have to bear the cost of this technology, reducing the ability of American companies to compete both here and around the world.
Despite the claims of media conglomerates, it is not at all apparent that these measures are necessary. I have been using the Internet as a means for acquiring new music for several years, including using primitive precursors of the now infamous Napster. In that time, I have purchased nearly one hundred music albums as well as many DVDs. I would not have purchased many of the albums had they not been readily available through these illegal means. The reasons for my illegal acquisition of music were solely for evaluation of the music. In nearly every case, if, after sampling an album, I found it to my liking, I acquired it legitimately. While illegal, file sharing services satisfy a need in the market where the music industry has failed. Consumers do not want to steal music, but given the unrelenting blandness of standard music distribution methods, such as radio and television, we are forced to seek other avenues in order to satisfy our tastes. One attempt by the music industry to enter this market, known as Duet or PressPlay, has so far been a dismal failure due to a limited selection of music and draconian control mechanisms, not the cost. In addition, the laws currently in effect have been successfully used to prosecute pirates, including just recently when a counterfeit DVD operation in New York City was shut down.
One of the other arguments made in favor of this bill is that making rich media readily available for download is what is needed to create widespread customer demand for broadband technologies. However, with the existence of illegal content online, broadband nevertheless remains in the early stages of adoption. There is no reason to believe that legitimizing this process will accelerate its adoption. In addition, the federal government should not be influencing the market in this case; broadband Internet access is hardly of the same importance as electricity or a telephone. Any hurdles to broadband adoption are a result of telecommunications companies dragging their feet far more than a lack of demand.
Examining the history of the relationship between the entertainment and technology industries, we find in several instances the former expressed their concerns and fears over a technological advancement, be it the VCR or the audio cassette. In these cases, the fears of the entertainment industry were overblown. In fact, after embracing technologies they originally opposed, the movie studios and record labels managed to increase their revenues and profits, in spite of piracy. While no system is perfect, to some extent fraud is the cost of doing business. No mechanism that protects the rights of law-abiding citizens will be perfect in quashing fraud. The banking and credit card industries have managed to be profitable while still maintaining a fair degree of protection for the rights of consumers. The entertainment industry should be no different.
While I don't doubt that this type of measure would go a long way to stamp out piracy in the United States, it does so by intruding too greatly both on the consumer as well as business concerns. There is a solution to the piracy problem, but it cannot come out of treating Americans like common criminals or burdening the technology industry with further regulation. As such, I strongly urge you to oppose the CBDTPA and all similar measures that only benefit a privileged few to the disadvantage of both individuals and businesses. I would appreciate hearing your position on this issue. Thank you.
Two things that make me think that most people and media don't "get" the Web. One, most documents on the web are still not dated. With magazines and newspapers, there's standard boilerplate, and each edition is pushed out at once. With the web, documents go out when they're done, or at least they should, and there's no way to infer that time by correlating with other dates and times listed elsewhere on that site. Two, most news articles still don't link. Even the most obvious links are generally omitted by most news organizations. That's just pathetic. No doubt it's an editorial decision to keep people from visiting other sites, but that just makes it even dumber. Linking made the web. Putting a news article on the web using HTML formatting is not publishing for the web.*
* On the other hand, I get seriously annoyed by the (Slash)dot-head types that insist on linking everything that could possibly be linked. TODO: formulate a general policy on when documents tangential to the main thrust of a sentence/paragraph/story should be linked and when they shouldn't.
I'm seeing more and more sites requiring cookies simply to browse the site. For instance, I was looking up locations for a clothing retailer. They used MSN Expedia as a source for their maps. Fair enough. But apparently just to look at the map, I have to allow all sorts of .expedia.msn.com and .msn.com cookies to be set. Just to look at one image. Law.com won't let you read any articles without using cookies. They're somewhat less intelligent about it; they try to set the cookie and then check it in javascript embedded in the page. If you disable javascript, you can view the articles just fine. Or you could download them using a non-graphical client like curl or wget and view it locally. But that ignores the point. I can think of no instance where cookies can be justified on technical grounds. Session management is a solved problem. Now, transparent logins are convenient, but that's it for cookies. Mainly, I think the website programmers are just lazy. Either that or some exec made a bone-headed decision to try to coerce the web-browsing population at large to enable cookies. Of course, I'll just go elsewhere. And when that becomes too difficult, no doubt there will be a plugin for Mozilla that allows you to specify that certain sites get random values for the cookies they set. Trying to coerce users with this sort of strong-arming is stupid and counter-productive. At best, they set off an arms race with those trying to circumvent these measures. At worst, they drive customers away. Why go through all that trouble when they can accomplish 99% of what they want by doing a tiny amount more work (in many cases, most of which they're already doing)? Audiogalaxy still works in Lynx! The whole point of this web business is that anybody can do it. And yet we still find publishers of all shapes and sizes trying to lock users into their own fenced of pieces of the web. Companies who long for the early 90s when the only access was Compuserve or AOL or Prodigy, completely controlled environments. Not only do people not want that, it's obvious that people don't want that. Look at the number of people subscribed to an online service in 1995 and compare it to today. These people don't understand how to serve us, plain and simple.
I just noticed that I am now the #1 (my resume on ketan.org) and #2 (this site) results on google searches for "Ketan Gangatirkar." It's about time, considering I haven't played Diplomacy in years, and those are the majority of hits out there. These sites are also the #1 and #2 hits for "Gangatirkar." Google also seems to do the right thing on simple misspellings of my last name. Astonishingly, there are now Gangatirkars online who are not people I know (thankfully); it's a rare name in India too. Ironically one of the hits is on a fansite for the band The Police; I just added their greatest hits compilation to my wish list yesterday. I'd better pick up gangatirkar.com before it's too late.
Thanks to my new job, I have gotten for free a SXSW Interactive pass. Speakers and panelists include Richard Stallman, Lawrence Lessig (who you may remember from Eldred vs. Ashcroft, Joshua Davis, Matthew Haughey, Dan Gillmor, Po Bronson, and a bunch more. God. The worst part of this is having to choose which one to attend. For example, "Surviving Your Own Collaborative Project" and "Freelance Forum: Going Solo" are both at the same time on Monday. If I get too far ahead of myself, I dream of participating more actively next year with amphetameme and crankpot.
One of the panels I attended at SXSW was called "Trends in how the Internet Connects People." The panelists were James Hong (Hot or Not), Scott Heiferman (Meetup.com, and Brad Fitzpatrick (Live Journal). There were some interesting things that I may write up later, but the most interesting part was a digression into why Google bought Blogger. Brad claimed that Evan (the main Blogger guy) had said Google didn't really have an agenda, that they bought it just because it was cool. I didn't buy that at all. There was some speculation that access to weblogs would help Google make its search results more useful and timely because the normal web is more static. Informational pages, FAQs, etc. on a given site don't get created on a daily basis (I call this "slow content"). Over the whole WWW, there are many such things created on a given day, but Google doesn't know when to look at them. However, Google knows that webloggers are finding these things on a daily basis and posting about them ("fast content"). Google knows where the weblogs are and it can thus use them to find the newer and updated slow content.
From this I gleaned an insight. People find interesting information in ways that cannot be emulated by crawler programs. There is too much out there on the WWW for even Google to index completely in a short period of time. It might take a month for Google to find out that some Britney Spears biography has been updated. But a weblogger might find that in a day and post about it. Google can then pick up on that and index the bio weeks before it would have otherwise. And if no webloggers link it, it's a good indication that people don't much care about it so Google can wait a few weeks to index it. From that Google can also infer some information on how useful the link is in ways that they cannot currently from slow content.
This is the same genius that Google is built on: rather than attempt to emulate human judgment (as Altavista and others did in the pre-Google days), Google skip the unreliable guesswork entirely and extract the human judgment that has already been expressed in the web by people constructing web pages. Webloggers do exactly that, except much faster than the more traditional, slow web that existed a few years ago. Each weblogger that Google tracks and indexes is in some sense a human equivalent of a Google web crawler, finding interesting information faster and more intuitively than a program ever could. The example at the panel was weblogs about chihuahuas. There's no way that Google would have a web crawler looking for information about chihuahuas. And, it turns out, they don't need one. They have the chihuahua webloggers doing that for them already. All they have to do is visit every now and reap the fruits. If you have a weblog, you have been deputized. You are part of the Google search posse. You work for them now. And this is a good thing.
I've coined what I call the "30 second rule." If you want people to visit a web site regularly, you have to give them at least 30 seconds. If you provide enough on your site to keep them engaged for that long, they will come again. If they don't stay that long, they won't come back. Obviously that's over-simplifying quite a bit, but I'm going to make a point of providing at least 30 seconds worth of content every few days. Maybe that just reflects my impatience and attention span, but everyone has less of both these days, especially on the web.
blogger remixes tribal december 2003 - You can remix blogger? With tribal grooves? And they do this every month?
brown phlegm - Recurring theme, #1
columbia kidnap adoption - Makes me wonder if I should report this guy to the FBI
engineer get any benefit in autocratic goverment - I dunno. Try moving to Syria.
download wav. mc hammer can't touch this - Recurring theme, #2
foley's coupon - I can just imagine someone trying to clip a coupon out of their monitor.
how does nuclear fission warm the earth's core - uh... through nuclear fission...
jonestown flavor aid pictures - I can understand why someone would want to see pictures of Jonestown, but why is it so important that they see pictures of the Flavor Aid? To put into their marketing materials? "Flavor Aid: Jim Jones Likes it."
list of common american names - Oh man, if you use my list, you are so screwed.
Innumerable "hurricane isabella" (including some very creative spellings) searches that are even more boring than these other ones.
coughing up yellow and brown phlegm - Recurring theme, #1
november 1, 2003 forecast - ketan.org merges with the weather channel
al pacino blinds for the window - I'm trying to imagine how that would work and it just creeps me out.
my heart will go on dion wav file - Recurring theme, #2
There is a particular kind of idiocy is embodied in flash intro pages that have the "skip intro" link in the movie itself, rather than the page in which it is embedded.
I predict that Google will move into the job search market. It's just another kind of search. It's basically a combination of Froogle and Google Local. In fact, I think they'll eventually roll out a number of specialized searches. They're crawling the web pages as it is. It'll be just like with Froogle, I think. Froogle's biggest advantage (besides being part of Google) is that its database is compiled automatically as part of the regular Google crawl. Other price engines, like Pricewatch or Pricescan or Dealtime or..., require active effort to put items into the database. That's a clear scalability problem. Ditto for the job sites. They all feed back on each other with recruitment companies spamming Monster and HotJobs and Dice and .... with the same listings over and over. I think this is something that Google could put together rather easily. They've already done most of the work. They crawl the web. They can infer location. And they've laid the infrastructure for specialized searches. I promise you it will happen.
I'm trying out Opera as a web browser and mail client on my laptop. Mozilla is a great browser, but there is something up with my laptop that makes it dog slow, especially after returning from hibernation (most likely the slow hard drive). Really, it's sad that it's taken me this long to give up and try something new. The last time I tried Opera was around version 3 or 4; they're on version 7 now. It is clearly very fast and well-designed, with lots of features for the "power user." Whether it's right for me remains to be seen, but I could see myself forking over the $20 educational price. Or I could keep running the ad-ware version, which is very unobtrusive.
One of the things that keeps me from using GMail full-time is that there is no way to export mail from it. I am a packrat, with email archives going back to 1997. I'd use my ketan.org email address for incoming, so I'd be able to save those as I forward them, but I wouldn't be able to save the message I send through GMail. One easy way they could set that up is to have a preference that all sent messages be automatically bcc-ed to a user-selected email address. Then I could capture both incoming and outgoing messages for the benefit of future generations.
I've been getting joe-jobbed a lot recently. As a result, I have implemented more stringent rules on what email I actually receive. You are no longer likely to get through by sending an email to random addresses @ketan.org. Mails sent there will be checked eventually, but I cannot guarantee it will happen in a timely manner because the whole point is to reduce the amount of mail I look at. So if you took joy in inventing email addresses for me, well, those days are sadly over.
I am getting joe-jobbed like crazy lately. I think I had something like 1500 bounces on Sunday alone from spammers using made-up addresses at ketan.org as the From addresses. I am sadly becoming one of those people for whom email is an increasingly impractical medium. Plain old spam sent to me is easy to handle; it's the misdirected bounces that are really getting me. The ones that really make me mad are the snarky "your message was flagged as spam." Somehow it never occurred to these sysadmins that spammers don't use their real email addresses. Almost as bad are the challenge/response systems that automatically reply and ask that you click on a link to prove that you're real. Note to world: automatically replying to emails is bad, no matter how good the reasons may seem.
There's little point in having a site search these days since you can just tell Google to restrict a search's scope to a single site. It's a little jarring for the user experience, though, since Google's results page will look completely different from the rest of your site. It'd be nice if Google allowed you to submit your own CSS link to be embedded in the results page. That way, "my" site search would look like the rest of my site, except of course with the appropriate credit due Google somewhere on the page.
I get mad at sites that gratuitously use Flash and only Flash. Musicians seem to be especially bad about it. At least one other person agrees with me and compiled a list of five mistakes band and label sites make. As a corollary to that, splash pages are stupid, especially ones that you have to click through every time you visit. Someone else followed up with a couple more. I think Franz Ferdinand's web site is basically unusable because it doesn't follow the rules.
The Justice Department is making quiet noises about requiring ISPs to log all customer Internet use. Web hits, chat logs, file transfers, etc. would be included in this. The destruction of privacy is obvious, and you don't have to think very hard to guess my opinion. More than being offensive, however, such a policy would just be stupid. Anybody with anything to hide (and lots of people without) would switch to using encryption, such as Trillian's SecureIM, PGP, tunneling over SSL, etc. It would be useless in catching criminals. Then there would be the terabytes and petabytes of storage required, which would likely bankrupt most ISPs. I doubt this would pass, because it would be too expensive and would accomplish nothing. Even if it was practical to save all that data, even if it wasn't a gross violation of the 4th amendment, and even if mobsters, child pornographers, terrorists, etc. used unencrypted channels (which they don't even today), the government wouldn't be able to analyze all of the data; it's just too much. This is just an attempt to avoid doing the real and hard work of criminal investigation while still appearing to be doing something. Technology cannot solve social problems. It can be used as a tool, but like all tools, its effectiveness is narrow, and it cannot replace a real solution.