Disqus - Latest Comments for Giles Bowkett

Re: When Rails Fails

Giles Bowkett — Thu, 12 Nov 2009 02:42:58 -0000

Wrong. It's not the framework's job to do anything. Rails isn't for lazy people. If you want an error-checking system, you build one.

That would be a ridiculously easy Rails plugin. Just use the ruby2ruby library and grep it.

Re: Places where Rails is not DRY

Giles Bowkett — Tue, 21 Oct 2008 19:27:38 -0000

fwiw, you can solve the first one - migrations up and down - by using the auto_migrations plugin from err the blog.

Re: ReinH — Adding Epicycles: Copernicus on Software Development

Giles Bowkett — Wed, 03 Sep 2008 00:35:03 -0000

NASA uses Ptolemy's system to chart rocket launches. Although fucked-up and legacy, it's easier to use than Copernicus' system for this particular use case.

The best book ever written on Ptolemy and Copernicus (to my knowledge) is Eric J. Lerner's The Big Bang Never Happened.

I sure hope HTML's allowed in comments here.

Re: Why Your Social Website Should Support OpenID

Giles Bowkett — Sat, 12 Apr 2008 20:10:18 -0000

OK. So now I have to respond. I had to make a whole gigantic presentation to answer your last spate of flame-war provocation. So let's see.

OK, so the blog post, "Do Users Really Even Exist?"

http://gilesbowkett.blogspot.com/2008/03/do-use...

The answer is no. Users are a convenient fiction. What actually exists are logins. OpenID assumes users map directly to logins. Because of this it is only useful to Web developers. In real life people share logins with each other or have more than one login. That's my point in the blog post. You dissed it on Twitter saying it was an interesting theory, but that was silly. It's not an interesting theory. It's an observation of the disconnect between how Web developers like to imagine people act and how you can actually see people acting in the real world.

OpenID is Web-developer-centric and based on an assumption that is wrong. It would be cool, IN THEORY, like Communism, but in reality, it's just ridiculous BS. Theories based on ideas which are repeatedly shown to be factually incorrect are theories which will not get you anywhere no matter how pretty they turn out to be.

Also, Microsoft tried to do an OpenID style thing years ago, and failed. Here's the pattern with OpenID and Passport: developers decide it should exist, they build it, and nothing happens. Whenever you have a pattern like that, it means that the real world displays characteristics that people are repeatedly failing to recognize. This is similar to micropayments. Several attempts at micropayments failed in a row before developers decided to give in gracefully and stop providing a technology that the world clearly didn't want. After a while Clay Shirky figured out why:

http://shirky.com/writings/fame_vs_fortune.html

One day Clay Shirky will figure out why OpenID never happened either. And it'll be really freaking interesting to read. But until then OpenID really isn't worth taking very seriously, and when it becomes taking very seriously, it'll only be as a way of leveraging Clay Shirky's remarkable brain. OpenID in and of itself is really not worth taking very seriously.

QED.

Re: A Django Developer’s Views on Rails

Giles Bowkett — Sat, 30 May 2009 18:15:07 -0000

@Aron, I don't mean to appear argumentative but regarding your last point you should check Justin George's comment, a few comments above the comment you made. Loose coupling is on the way for Rails in a very big way. It's going to be an awesome improvement.

Re: Zed Shaw Rant Rant

Giles Bowkett — Wed, 25 Jun 2008 16:27:29 -0000

My gems are totally metal, except for a few which are indie.

Re: @lazytwitter 1: is it possible to hack an iphone with a broken power button? 2: I may have a free T-Mobile SIM card for anyone :p

Giles Bowkett — Tue, 09 Dec 2008 01:34:08 -0000

doh! i have already moved to 2.2.

I can't get a warranty repair, I don't think; it's A) old and B) second-hand.

Re: @lazytweet best bok on RSS?

Giles Bowkett — Tue, 30 Dec 2008 05:06:09 -0000

Need a book on RSS because I'm building something I thought would be insanely easy and in fact it's quite buggy.

Re: http://adam.blog.heroku.com/past/2008/7/6/the_end_of_bugs/

Giles Bowkett — Tue, 09 Sep 2008 15:25:01 -0000

Absolutely! This is why debugger support is a bad thing. Development practices which are very close to bug-proof exist.

Re: Database Versioning

Giles Bowkett — Mon, 02 Mar 2009 17:55:35 -0000

I think the answer is to use Err's auto_migrations plugin, and never use migrations.

Re: Close to the Problem

Giles Bowkett — Sun, 03 May 2009 20:16:37 -0000

I blogged something very similar not long ago.

Re: Tech Tuesday: The Fiddly Bits

Giles Bowkett — Wed, 25 Feb 2009 19:53:26 -0000

Can you release parts of this system as open source? I'd like to take a deeper look at it, and experiment with it.

Re: Receiving Email with Rails

Giles Bowkett — Fri, 24 Apr 2009 03:14:59 -0000

hey jason - also check out Astrotrain on github. it receives e-mail and issues xmpp or http in response. blatant plug for my employer but it's what enables Tender to turn e-mails into support discussions. like how Tripit turns e-mails into itineraries automagically.

Re: Hint: Expert Rails Books Aren't About Rails

Giles Bowkett — Tue, 02 Jan 2007 14:21:00 -0000

_I think a lot of programmers are going to find themselves looking for jobs in the coming years, and not because their jobs are getting sent to India. In fact, fewer and fewer jobs will go to India for the same reason these guys will be out of jobs. In the past, and increasingly less so now, the knowledge of how to configure different frameworks and technologies to play well together was mistaken for the ability to develop. Programming jobs are filled with people who are more admins than developers. With things like Rails obviating the need for these pradministrators, the jobs will go to the guys who know how to build valuable software._

I wish I believed this, but let me tell you why I don't. Think about the silly sci-fi idea of robots taking over the world. If you look around and compare the present to the past, the big difference isn't that robots are taking over the world -- it's that *everybody* is becoming a programmer. Even watching TV requires programming skill today. You've got to program your remote, or tell your TiVo when to record something. I even have a programmable toaster. I think this trend is like a fifty-year trend, and it's only going to keep going.

I think what you're talking about, really, is very true on some levels, but it's almost that technologies like Rails make building applications so easy that the programmer who approaches his work like a _*designer*_ is going to be the one who succeeds. I think that is very true, but the job market for "programmers" who are really just admins with the ability to write code, I don't see that drying up any time soon.

Re: Declarative Workflow in Rails

Giles Bowkett — Sun, 21 Jan 2007 14:21:00 -0000

Hey Pat -- how do I play around with this?

Re: Java People Must Be Stupid

Giles Bowkett — Mon, 22 Jan 2007 14:00:00 -0000

No way, Pat, I read your blog under normal conditions, and I don't agree at all. It's not really stupid people, it's people who've accepted a stupid idea. Java's design is based on the idea that a language can prevent misuse by making bad things hard to do. It's defensive thinking. Java people are caught up in this paradigm; that's where these questions come from.

It might be more accurate to say that Java people are smarter than they give themselves credit for.

The thing is, telling people they're stupid, when their problem is an equal mix of bad ideas and defensive thinking, that's not the most persuasive argument available.

I think the only compelling argument, from a Java perspective, at least, in terms of Java people I've talked to, is that nobody ever actually accesses the stuff inside Rails which, if it were inside Java, would be hidden. The door isn't locked, but still nobody goes in.

In any OOP language you do want to keep client programmers only using the API, not the internals of the objects the API is made of. That's a good design principle. Java makes it happen by building a fort, locking everything down, the theory being they can't steal it if it's nailed to the floor. Their goal is legitimate, but their way of achieving it is kind of paranoid. Ruby's got the same goal, but a radically different approach to achieving it. It doesn't enforce any kind of security, but makes it easy to build such elegant APIs that entering the "forbidden zone" never even occurs to people, because it just isn't necessary.

It's like the opposite of thinking defensively isn't thinking offensively; the opposite of thinking defensively is thinking *creatively*.

Re: Java People Must Be Stupid

Giles Bowkett — Mon, 22 Jan 2007 16:15:00 -0000

@Pat - I predict many many more angry comments. People will read every word you wrote but the title will weigh so heavily in their minds they'll interpret those words in ways you never intended. Could be a real nightmare.

@Dave - yeah, but again, they reach in to create APIs, not to circumvent them.

@Everybody! There's a lot more to understanding Ruby than intelligence or stupidity. I've met very intelligent people who didn't understand programming at all. I think any programmer who ever leaves their house will do the same thing sooner or later. It generalizes up from there. Criticizing smart Java people who don't understand Ruby is like criticizing smart Americans who don't speak Japanese. It's got nothing to do with intelligence, it's just where you're coming from.

Re: Why You Have To Work For A Startup

Giles Bowkett — Sun, 01 Apr 2007 13:13:00 -0000

Personally, I would amend this to say "Why You Have To Work For A Startup At Least Once." Or expand it to the general case of doing something new and tremendous. I got into the Web in 1994. The dot-com boom might have been underway by then, but that's not what people were calling it. They were calling it "What are you talking about?"

But it's true, the experience you get creating stuff that hasn't existed before is so much more useful than the experience you get implementing somebody else's ideas or going over ground that's already been thoroughly explored.

(Of course the 37 Signals guys would probably amend that even further, given their famous anti-VC stance.)

Re: Why You Have To Work For A Startup

Giles Bowkett — Thu, 05 Apr 2007 13:33:00 -0000

I do interesting, creative stuff all the time. Sometimes I do it at my job, sometimes I do it outside my job, sometimes I do it in spite of my job. Startups are great, but they require too much time, and the thing is, compared to the days when nobody had even heard of the Internet, startups today just aren't doing anything truly revolutionary. The only genuinely revolutionary startup I'm aware of is iRobot. Every other company out there is still busy fulfilling the visions of William Gibson, Neal Stephenson, and Bruce Sterling.

Re: Story Runner top to bottom screencast

Giles Bowkett — Mon, 08 Oct 2007 11:20:00 -0000

In addition to the upcoming TDD with Rails studio, I've also just watched all three of Geoffrey G's Peepcode screencasts on RSpec. (As you can imagine I'm adding your screencast to the mix and looking forward to a nicely tired brain.) Anyway, the reason I bring this up is I'm pretty sure that sometimes GG will record the video of his screencasts in one pass and the audio in another. I had that kind of earmarked as a strategy for my own screencasts, but I'm sharing it as a solution to the X many things at a time problem.

Re: Story Runner top to bottom screencast

Giles Bowkett — Tue, 09 Oct 2007 14:28:00 -0000

Dude. Weak. People are name-dropping me just to get a laugh. I'm the Paris Hilton of bloggers! Oh noez!!1ONE

Re: My Beef With ActiveRecord

Giles Bowkett — Mon, 19 Nov 2007 12:10:00 -0000

OK, at best I'm being pedantic and at worst I'm being wrong, but I think you actually have an issue with Active Record, not ActiveRecord. The whole idea of Active Record (the pattern) is to mix business logic and persistence. Any time you use a pattern you're basically opting for a particular set of tradeoffs. Active Record is a good match for a lot of situations, but not all. Sometimes the tradeoffs are the wrong tradeoffs and the solution is to use a different pattern.

On the other hand, it could be that you could get around this just by writing an AR subclass called ActiveRecordBaseWhichEnforcesBusinessRulesForChristsSake and then building your model off of that rather than off of AR::Base.

Re: My Beef With ActiveRecord

Giles Bowkett — Tue, 20 Nov 2007 17:46:00 -0000

Well, DHH was initially a PHP guy. Rails wasn't actually designed for serious OOP use. I don't even mean that as a dis, it's just true. The concern you're raising comes from a different part of the programming world.

I think you should build it. Producing a variant on ARec which made proper OOP possible would probably make life easier for a lot of people. The question is, can you find a solution simple enough to fit in a Rails plugin? It's obviously worth doing and obviously doable, the only question is how many lines of code it would take.

I think patching AR::Base#initialize is certainly the right idea. I don't have ARec's internal structure memorized, but calling validations from within #init ought to be easy enough. Piers' approach is a great place to start, and the question of protecting attribute assignment is actually still worth debating. After all, you're not trying to turn it into Java. You're just trying to make validates_xyz actually validate xyz.

Re: My Beef With ActiveRecord

Giles Bowkett — Wed, 21 Nov 2007 12:27:00 -0000

Oh, and *I* blog to let the world see how smart I am. You just link-spammed your own comments.

Re: Refactoring with Shared Example Groups

Giles Bowkett — Wed, 14 May 2008 08:58:00 -0000

On the tests, I would say test-driven refactoring ftw. If the tests pass all the way through, the refactoring succeeded; if the tests targeted implementation specifics, first kill those tests (and write new ones if necessary).

Re: Refactoring with Shared Example Groups

Giles Bowkett — Wed, 14 May 2008 09:01:00 -0000

Sorry, that comment was kind of dumb. Twitterbrain, responding too quick.

Re: method finder for Ruby

Giles Bowkett — Sat, 04 Oct 2008 11:12:00 -0000

Sorry Pat but I think Dr. Nic already did this, in 2006:

http://drnicwilliams.com/2006/10/12/my-irbrc-fo...

Re: Testing protected and private methods in Ruby

Giles Bowkett — Tue, 28 Oct 2008 05:31:00 -0000

This is crazy talk. How the hell are you supposed to have untested methods in the first place? Where did they come from? Did you develop them TDD/BDD and then just throw away the specs, like burning a secret message? This is fucking ridiculous. If you have code that came into being without being tested, you fucked up. If you threw specs away you made your code not-refactorable. I think using private or protected at all is a childish self-indulgence, but throwing away specs or developing without specs is just absurd.

Re: Tagaholic - Hirb - Irb On The Good Stuff

Giles Bowkett — Fri, 13 Mar 2009 20:50:47 -0000

You might find some useful stuff in my irb gem utility_belt. It's under-maintained (sorry) but has some cool features, e.g., the ability to edit irb code in vi, verbosity hacks for Rails and for regular irb, etc.

Re: Propane for Campfire group chat | micro review - mac app

Giles Bowkett — Wed, 07 Jan 2009 19:34:54 -0000

Hey, this is a cool review, but that's not all - try dragging a Twitter URL into the window, or a graphic

Re: Propane for Campfire group chat | micro review - mac app

Giles Bowkett — Wed, 07 Jan 2009 19:34:54 -0000

Hey, this is a cool review, but that's not all - try dragging a Twitter URL into the window, or a graphic

Re: From Geek to Freak: How I Gained 34 lbs. of Muscle in 4 Weeks

Giles Bowkett — Fri, 11 May 2007 13:23:48 -0000

Hi Tim - you really ought to do a whole book about this. I actually found this by googling a comment you made in your talk with Marci Alboher at Google, about cutting your weight dramatically. Between cutting your weight dramatically, increasing your muscle dramatically, and lowering your cholesterol dramatically, you have a whole book there. Also - you have a whole book there which could seriously improve my personal life expectancy. I want my life expectancy to improve! I personally will be massively indebted to you if you document the methods and theory behind these astonishing transformations. I need to make similar transformations, both for superficial reasons and for important ones.

Re: From Geek to Freak: How I Gained 34 lbs. of Muscle in 4 Weeks

Giles Bowkett — Tue, 15 May 2007 02:58:30 -0000

Tim - you have GOT to make this a book!

Re: Poaching Customers on Twitter )) Skepsis

Giles Bowkett — Wed, 06 May 2009 19:18:31 -0000

I personally don't believe that kind of poaching is a good idea because you can get into endless bickerfests about the definition of your features. I mean they could spend twenty tweets arguing over that only to find out you're only ever interested in this one Thai place which is equally accessible from both (for example).

Re: The Modern Babel

Giles Bowkett — Fri, 29 May 2009 11:04:30 -0000

First, you really shouldn't say Oriental but Asian. It might be a nonissue in the South but it's offensive on the West Coast. Don't ask me why, it just seems to be the case.

Second, learning Latin is so much more than structural similarity or language foundation. If you can't read Latin, you can't really read English, either. Your assumption about looking up words is completely inaccurate. If you can read Latin and Ancient Greek you'll have an infinitely finer-grained understanding of etymology in English and you'll be able to spot a word's entire history just in its spelling. Life without classical languages is like life without sex or LSD. I feel sorry for anyone who doesn't learn these languages. Latin makes you much better at logic and Greek will turn your understanding of words inside-out.

Also, that stuff about sentence structure is mostly bullshit. Latin and Ancient Greek use much more complex verb and noun forms than any of their descendent languages. That's why they teach you about logic and interpretation. Every sentence in either language carries a complex, highly specific structure, and it trains you to think in this very complex, highly specific way. You seriously should not go through life without learning Latin.

And this:

you can’t express anything in one language that can’t be communicated in another.

You have no idea how wrong you are. That's the most egregious falsehood I've seen since Uncle Bob's keynote. It's just complete and utter bullshit. Absolutely false. You can't express anything in a given language that you can't *summarize* in another, but if that crazy statement was true, poets would not exist.

For instance, Pablo Neruda wrote a line in a poem that goes like this: "mi voz buscaba el viento para tocar su oido." A literal translation could be "my voice sought the wind to touch her hearing." Except literal translation fails here, because "tocar" carries an additional meaning. In Spanish you don't play an instrument, you touch it. So you could pair the sentence with another literal translation: "my voice sought the wind to play her hearing." But to say this loses the poetry and nuance is overly generous. It doesn't even parse. It doesn't even make sense. The idea that he wants to touch her hearing and play her hearing like a musical instrument, all from far away, via the wind - you cannot express that in English without a paragraph.

It's just like how in Ruby you can write in three lines something Java requires a whole page of text for. Think how ridiculous your argument would be if you were saying all computer languages were equivalent. What you're saying here is equally ridiculous, and for a very similar reason.

Re: The Modern Babel

Giles Bowkett — Fri, 29 May 2009 19:06:24 -0000

The other big reason to learn Latin is if you want to get into law school. When I was in high school, my mom was a lawyer, and many people noticed that I liked to argue, so people often assumed that I would become a lawyer. Anyway the etymology thing, having studied both Latin and Ancient Greek, etymology surprises me like once every seven years.

I think the more interesting idea is that every kid should learn some programming languages. Alan Kay did a real neat talk where he said you should teach every kid programming because it's only when you know how to automate something that you really know *how* to do it at all. Wish I had a video URL or something but I've only heard of it, not seen it myself.

Re: yeah its ok

Giles Bowkett — Wed, 14 Nov 2007 17:55:00 -0000

Might be better just to have a remembering table, or, to be more Railsy, a memories table, and have a memory be just the user_id and the token.

Or even abuse the class_name option on association classes:

has_many :logged_in_users, :class_name => "LoggedIn"

and then:

LoggedIn.remember(@user)

Re: irb & script/console tips

Giles Bowkett — Sun, 28 Dec 2008 21:34:00 -0000

Thanks Avdi! If you do gem install utility_belt you get pretty much all of this automatically, plus vi integration. You can be in irb, go into vi, code up anything you want, and on exit irb will eval it. You can do the same with emacs or TextMate also. You also get programmatic control of the clipboard, googling stuff, pastie – I wrote it before git existed – and a bunch of other crap I don’t even remember.

Re: irb & script/console tips

Giles Bowkett — Sun, 28 Dec 2008 21:35:00 -0000

Wow, that was a stupid comment. I wrote it before gitHUB existed, I meant to say – and therefore also before gist.

Re: Five reasons to use a static site generator instead of Wordpress - Guestlist

Giles Bowkett — Thu, 22 Oct 2009 17:04:47 -0000

My Hacker News reformatter site Hacker Newspaper uses the same principle. I have no interest in the community features on Hacker News; I just want links. The site only supports that limited use case, but it's much, much faster than Hacker News.

Re: Ducking an iPhone Annoyance

Giles Bowkett — Thu, 04 Dec 2008 15:06:32 -0000

You can also just turn off autocorrect as of the 2.2 firmware. (I seriously would have gotten rid of my iPhone if this wasn't the case.)