The StackOverflow Podcast

Intro, advertising

[0:51]

Atwood:    Uh, the WWDC is going on, right? The Apple conference.

Spolsky:    Yeah. The Apple pilgrimage to the Mecca.

Atwood:   That's right, I think there was some new product they introduced. I don't know what it was; I haven't heard anything about it yet, but I think there might have been a new product introduction.

Spolsky:    Our listeners will probably know more about it than us, 'cause by the time this hits the airwaves...

Atwood:   Yeah. That's right, that's right. So it was fun; I got to meet a few Apple people and...

Spolsky:   Oh, did you actually go?

Atwood:   Oh, well no I didn't go to the actual conference. There was a... just sort of a...

Spolsky:   Oh, there was just, like, a really cool line outside that you just hung out at?

Atwood:   There was a little event that I went to. I was actually invited by Wil Shipley, who wrote the Delicious Monster software.

Spolsky:   Oh yeah, that's awesome!

Atwood:   Yeah, and also he is a user of FogBugz as I found out -

Spolsky:   Yep.

Atwood:   - when I met with him. He had very favorable things to say about FogBugz, so...

Spolsky:   Oh good.

Atwood:   Yeah, and when I talked to him I told him, "yeah, you know.. " talking about Joel, and he's like, "oh is Joel here?!" and I said, "no no no, Joel's in New York." [chuckles]

Spolsky:   I have people for that. I sent my people.

Atwood:   To actually attend these events?

Spolsky:   Yeah, there were a couple of FogCreek people there.

Atwood:   [chuckles] Yeah, for some reason Will kinda reminds me of you a little, I don't know why. Maybe it's just the whole general attitude, I guess. He's very outspoken.

Spolsky:   Is he like stunningly handsome?

Atwood:   [laughs] He's just very outspoken, very opinionated. But yeah: great guy. So, that's where I was at.  I got to do that.  That was kinda fun.  So, one thing that I posted this week and got a lot of feedback on, and maybe... I'm sure you'll have an opinion on as well, was as we work on StackOverflow we're actually building a lot of functionality in JavaScript.

Spolsky:   Oh yeah, just go for it.

Atwood:   Yeah, 'cause it's just really mature now. There's so many common frameworks - we're using JQuery. But, a lot of things we want to do... it just seems very natural to do it in a kind of semi-dynamic sort of way. The editor we're going to use - has a very unfortunate name, WMD, but it's based on MarkDown - and it has a real-time preview and that's a component that's already built.  But that's Javascript.  And then some of the other things we're going to do, such as when you ask a question, it's gonna, out of band, go out and query to see if other questions similar to the one you're asking have already been asked. That's another thing where it just seemed sort of natural to just do it as you type. So yeah, so one of the questions I posed in the blog was, you know, "is it OK to, at least when you're asking a question and when you're actively participating [Spolsky: and editing] and editing, exactly, is it ok to basically require JavaScript?"  I mean, I'm sure.. the site won't be "broken" if you have JavaScript off but it will be probably a pretty poor experience.

Spolsky:   Does anybody really turn off JavaScript nowadays, and like successfully surf the Internets?

Atwood:   Yeah, I was going through my blog...

Spolsky:   It seems like half of all sites would be broken.

Atwood:   Yeah, and I kinda asked that question in 2006.  I was like, "well who really browses the internet with JavaScript off?" and now, it's like 2008 and it seems even more, sort of ridiculous that someone would do that. I think there are obviously accessibility concerns that...

Spolsky:   I'm tempted to scrap IE6.

Atwood:   I think we'll work on IE6.  I'm not too worried about the visual look, my feeling with IE6 is as long as the site kind of works, it doesn't necessarily have to look "right"

Spolsky:   Well, you'll find all kinds of little weird JavaScript weirdnesses with the editor and stuff and it'll, at least our experience is that it takes a lot of time to backwards test all those things and get them working on IE6.

Atwood:   Yeah. Yeah, that was one of the weirder news items that came out of the WWDC was Mobile Me, according to Apple, doesn't support IE6 and that was, I guess, kind of news. But yeah, IE6 is generally reviled sort of like the old Netscape 4.7.

Spolsky:   And apparently... did you hear the reason that Mobile Me doesn't support IE6?

Atwood:   They're Apple. Do they really need a reason for doing the things that they do?

Spolsky:   The reason is that there's some kind of bug with cookies from two-letter domain names, 'cause it's me.com

Atwood:   Really?

Spolsky:   Cause IE6 had a bug. That's what I heard.

Atwood:   Really? That's interesting. I didn't know that; I had never heard that explanation yet. 

Spolsky:   One of these rumors that floats around.

Atwood:   Oh, and also a pronunciation thing, because of course I'm the bad pronunciation king. I have said, "OS EX" before, which is wrong. It's actually..

Spolsky:   No, they will.. yeah..

Atwood:   It's "OS TEN," and I apologize to everyone who was... whose ears were offended by that 'cause I just didn't know.

Spolsky:   I say that too. I can't remember. It's one of those things that I'm never gonna be able to remember.

Atwood:   I found a short post...

Spolsky:  Life is too short to worry about what they call their operating system.

Atwood:   You know, it was funny, their "proof" for it, which was very amusing, which I agree with actually was that in OS X's speech synthesis module, if you type "OS" and the character 'X' it'll actually read it as "OS TEN."

Spolsky:  I'm sure, I'm sure, I've heard Apple employees say "OS X."

Atwood:  Really?

Spolsky:  Yeah.

Atwood:   Yeah, well that's a, you know, according to several sources - authoritative sources - is that it is, in fact, "OS Ten" so I'm trying to say it the right way. So I apologize.

Spolsky:  Well, you're gonna have to remind me 'cause I don't have the brain cells for that particular application right now.

Atwood:   I don't know why I was even searching for that. I think it was just one of those things. It's weird, I'll just be working throughout the day and I'll just have some random thought like, "Oh, I gotta look this up" and of course, it's totally interrupting whatever it is I'm doing. It's very much an Attention Deficit Disorder. But hey: I learn useless things like how to pronounce, you know, OS X.

Spolsky:  There's an article which is probably not worth reading, but the title says it all, by Nicolas Carr called, "Is Google Making Us Stupid?"

Atwood:   You know, you and I are on the same wavelength! I have that up on my screen, I totally was going to talk about that.

Spolsky:  Alright, well I guess I'd better read it.

Atwood:   It's really good. It's really good.

Spolsky:  Now remember: Nicolas Carr, the only reason you've ever even heard about him is because he wrote this book claiming that all these companies spending money on IT projects are a waste of time because all they're doing is fussing around with the fonts and the typesetting in their PowerPoint slides and they're not actually adding any productive value in any way.

Atwood:   Yes.

Spolsky:  So that was sort of his first major... getting himself on the map by claiming that all this, all this IT work we've been doing - adding GUIs and wizzy things and fonts - just creates make-work which doesn't actually create any kind of long-term competitive advantage for any particular firm.

Atwood:   And that, just to be clear with our audience, that book was called "IT Doesn't Matter" as in 'it doesn't matter.'

Spolsky:  Yep.  So "Is Google Making Us Stupid?" is an excellent follow-up probably to that... kinda the large picture.

Atwood:   Yeah, no he's a brilliant writer.  I mean, this particular article is in "The Atlantic" so you know he's a real writer because if you get published in "The Atlantic" then you're actually a writer.  So what I do is just sort of hacker-y.  He's really good.  And actually he has linked to my blog before which is always a little thrill when people really, really good link to your blog.

Spolsky:  But not from "The Atlantic?"

Atwood:  Pardon me?

Spolsky:  But he linked to your blog, but not from "The Atlantic?"

Atwood:  No no no, from his own blog which is called "Rough Type." But it's really funny that you bring that up because I literally have that up on my screen and that article - which I did read, actually last night on BART on the way back from this WWDC party - was interesting because it was saying, you know, the way you consume information online sort of changes the way that you actually approach reading fundamentally, which I think you can make kind of a case for this because he had a lot of - it's somewhat anecdotal, right? - but he also cites this study, which was really interesting (in the UK) and I'll read you just this one little paragraph. So, in this study - this is actual science, right? - they said, "it is clear that you are not reading online in the traditional sense. Instead, there are signs of new forms of reading emerging as users "power browse" horizontally through titles, contents, pages, and abstracts, going for "quick wins," it almost seems as though they go online to avoid reading in the traditional sense.

[9:10]

Spolsky: That used to be called speed-reading and they used to sell courses on television. Late night you would tune the television and there would be a person advertising a course in speed reading. So i'm not sure this is entirely completely new-

Atwood: No no no

Spolsky: -this activity.

Atwood: I agree this is not at all new.

Spolsky: Except now we call it "power browsing", instead of "speed reading."

Atwood: It's funny I remember talking to my dad at one point. My dad and I used to be, .... I was always an avid reader when I was a kid, and my dad and I would go to the library and get books for each other and stuff like that.  And my dad asked me. "Oh what books you have read recently?" and I said "you know haven't read a lot of books. I spend most of my time reading stuff on the Internet." And that still is kind true.  I mean I'm reading Clay Shirky book for sure. But-

Spolsky: Yeah.  Have you finished that yet?

Atwood: -the number of books that I read is very very small now.

Spolsky: Have you finished Clay's book?

Atwood: Not not yet, I'm about halfway through now but god I love it. The more I read it the more I like it. and I did follow up with Jarrod from the StackOverflow team. I gave him my original copy of that book while he was here. So I am following up on his homework

[10:10]

Spolsky:  Did you get to the part--the part that I thought was kind of interesting was that he talks about the most popular Meetup groups.  What's kind of interesting is that the most popular Meetup groups are probably not what you would expect--where the heck is this... Meetup, how would I find this on the Internet?--anyway, they're not the things you would expect, there's Chihuahuas is real popular, there's stay-at-home-moms is absolutely huge, and then there are all these sort of not-mainstream religions, like Wiccans, and spirituality, and atheism and stuff.  So it's sort of the kind of people that, like, the mainstream religions have a place to organize, and so they go there.  But the non-mainstream religions use Meetup, because they don't have churches yet.  And I thought that was kind of interesting, and it sort of applies to StackOverflow, in the sense that, if Meetup had defined what Meetups they wanted to create, if they had sort of come up with their own list of things, they would not necessarily have thought of the things that became popular.  They certainly wouldn't have.  When you look at that list, the things are just too bizarre, Chihuahuas and stay-at-home-moms and Wiccans, they never would've--it never would've occurred to them to create Meetups for those particular groups.  Because, they might have thought of Democrats, Republicans, but these are groups that actually have places to meet anyway.  And the way that applies to StackOverflow, I think, is that we might have a temptation to go in there and say, "Let's make an area for Java questions, and an area for Python questions, and an area for .NET questions," and discover that there already are really good places to ask, let's say, Java questions, for example.  But if our system was just "tag this any way you want," then we might start to see tags emerge that reflect maybe a funny list of weird things, like people programming the iPod SDK using Ruby, because there's no good place to talk about that.  Or it may just be sort of more bizarre things that don't actually have great existing communities, and they might emerge, and our top ten most popular tags may not be "Windows Programming," or "Web Programming," they may be something very unusual that we never would've thought of.  

Atwood:  Yeah, I totally agree with that.  And actually, a real-world example that came up - I was talking to Jarrod yesterday - on the ask page we were originally thinking that maybe we would provide some sort of visual list of the existing tags to sort of help users out and understand what tags they can apply to a post.  But then we realized, like actually showing a list and any kind of hierarchy to the list, means that we would be in charge of that.  And when we started thinking though like what the tags should be, and what they would look like, we realized that there's no way to do that and really understand the relationship between these tags.  It's sort of an insoluble problem.  In other words: you don't really know what people want to talk about before they're gonna talk about it.  So you have sort of a chicken-and-egg problem.  So what we eventually realized was, instead of having a visual list, we're just going to have, literally, a text box where as you type we're gonna auto-complete with some of the tags people have already used, so you'll get a hint of "OK, there's .NET, there's Python, there's iPhone, Objective C," things like that, so people don't misspell the tags."

Spolsky:  That's kinda useful.

Atwood:  Yeah, but in general it's basically, there's no hierarchy, it's just whatever people want to talk about, whenever they want to talk about it, you just type in the words.  For example, we were thinking, "OK, you could have top-level categories, like .NET, Python, C, things like that.  But then the relationships become really weird," because you can talk about .NET, and then Python at the same time, because you have, say Iron Python.  And there might be Python frameworks that normally you would only talk about in the context of Python, but then all of a sudden your talking about it in the context of .NET, so if you make that a parent-child relationship, it has multiple parents.  So it becomes a very flat list, which is, I think, simpler in a way.  But it's an example of what you're talking about, where it's just, you don't know what the community wants to talk about, so you don't wanna sort of institutionalize those relationships in code, it doesn't really make sense to do that.

Spolsky:  And I think we may discover that the tags that emerge as the most popular ones might not be what we expected.  

Atwood:  That will probably - almost certainly - be true.  I mean, we're gonna seed the community with your .NET discussion group.  And, for anyone who hasn't heard us talk about that so far, Joel does have on his site, I believe it's discuss.joelonsoftware.com, is that right Joel?

Spolsky:  Right.

Atwood:  So there's an existing community there, and we're gonna turn off--

Spolsky:  Of .NET questions, yeah.

Atwood:  Of .NET questions, thank you, we're gonna turn off that community and say, "Hey guys, StackOverflow is the new, 'approved', blessed version of this."

Spolsky:  [laughs]

Atwood:  And then hopefully we'll get some initial early adopters on that.  We were thinking also about the beginning of the site, it's gonna be a little weird.  So there's really gonna be a couple phases, and it's still scheduled to happen near the end of this month, we're gonna let you look at it, because we figure you're gonna have a lotta changes for us once you see it up and running.  

Spolsky:  You mean me?

Atwood:  So we're gonna go through a cycle with just you, then we're gonna go through a cycle with people on the private beta list.  Which now is (I think) 300 names.  So maybe a two-week cycle that way.  So before anyone actually sees it, we're hoping to get all the "suck" out of it, right?  All the dumb mistakes we made that we didn't think about, having more eyes look at it, by the time it's actually consumed by the public it'll be sort of--

Spolsky:  I don't mind if it sucks when people first see it, but there should be something there.  That's why I wanna have a beta.  Because if people come to it and it's literally empty, they won't really know what to do.  And they'll say "OK, that was fun, why did I just waste my time," and they won't come back.  

Atwood:  Yeah, no, I totally agree, I think on a professional level too, it's a little scary when Jarrod, Geoff, and I - the StackOverflow team - work on this, we realize it has to be pretty good because there's a lot of eyeballs on it.  But that's the way it should be, I mean, there should be some professional pressure to produce things that are good, that should be the natural order of things.  I was always scared when I was in environments where it didn't matter what you pooped out, people had to accept it, because you were the only people doing it. [laughs] Y'know, they just have to consume whatever it is you happen to produce, and they have to like it, because that's where they work.  I mean, there should be competitive pressure here, right?  If people look at StackOverflow and go "Y'know what, this kinda sucks, and this other site is better," that's the way the world should work, there should be choices so that you have an incentive not to suck.

[17:20]

Spolsky:  But it can be missing limbs. I mean, it can be seriously broken as long as there's questions and answers there, and quality people are giving quality answers to quality questions.  Because everything else we can fix after we launch.  And as long as people feel like they can ask a question and get an answer, they'll be willing to try it out.  They're not gonna--I don't think this audience is gonna really care that much about--especially in the early days, about the visuals being perfect, and the UI being totally slick.

Atwood:  Right, it's a process of evolution, absolutely, and certainly the attitude I've learned from my blog, and I think with web apps in general, you get this sort of healthy attitude toward apps that sort of grow over time or you're sort of gardening, I call it "web gardening."  It's like a marathon, not a sprint, right?  

Spolsky:  Have you ever gone to one of those new websites that everybody's talking about on TechCrunch or whatever, and you get there and the graphics are just stunning, and it's just visually perfect, and the UI, a lot of work has gone into it, and the actual functionality of the site is just blah, it's just uninteresting and boring and you never go back?  And you think "OK, that's the last time I'm ever going there!"

Atwood:  I think there's a lotta stuff like that, and there's--I don't really like to go to TechCrunch, because I feel like it's what I call "venture capital pornography," where it becomes about making money more than actually making things useful and interesting.  And that culture really bothers me, for some reason.  When people complain about Michael Arrington and stuff like that I tell them, "look: just don't go there."

Spolsky:  Well, my complaint about TechCrunch is sort of unrelated, it's that it's presenting itself as a news site, but there's actually no news-gathering activities happening there.

Atwood:  Well, Arrington does post some commentary now and then, and some of his commentary--

Spolsky:  So then it could be a commentary site, yeah.

Atwood:  But it's pretty rare, I mean, for the most part, it's just, "Oh, look, another Web 2.0 site that does X," in that sense, it's not really doing a whole lot.  But it's become a huge phenomenon, obviously.  I mean Arrington is like an industry unto himself now, so take it for what it is, I guess.

Spolsky:  Must be one of these Silicon Valley things.

Atwood:  Yeah, I think so.  I have a friend who sort of has been immersed in that culture for a while.  And actually, this friend, you might have heard, he had a video called "Here Comes Another Bubble," I don't know if you saw this.  It became viral--

Spolsky:  Oh, that was funny!

Atwood:  Have you seen it?

Spolsky:  Yeah. It's hilarious.

Atwood:  It's great, his name is Matt Hempey, he's a really, really smart guy, and part of the reason I think he wrote that is this whole frustration with sort of the Silicon Valley culture where it's sort of navel-gazing at some level.  It's like, you're doing stuff because other people are doing stuff, and the whole TechCrunch cycle of doing that is sort of very funny.  But yeah, you're right: that's one aspect of Silicon Valley culture I don't necessarily care for, is the echo chamber of everybody doing stuff on TechCrunch.  

Spolsky:  I think they really believe that TechCrunch brings you a large number of hits to your website.

Atwood:  But completely empty hits, it's kinda like those Digg links to your site.

Spolsky:  Right.  They're people that don't hang around, and don't really make accounts.  You might actually, in the early days, you might have gotten a few people to sign up that way.

Atwood:  It's kind of like soccer hooligans, right?  They care about whatever immediate pleasure they can get from clicking on your site, like "Oh, look, a funny picture of a dog!" or whatever.

[20:55]

Spolsky:  There's sort of too much of it.  I think TechCrunch sort of posts too much, and there's too many new sites there for any reasonable person to actually go check them out.  At some point, it's just like, "Oh, God, another, yes, I know, it's a site, that's fine, tell me about the good ones, don't tell me about everything."  

Atwood:  Well, I think, too, it lacks context.  I mean, if you're reading a blog post about something, "Y'know, I had this problem," and blah, blah, blah, and in the context of that problem they say "Oh, there was this site that had this cool thing that helped me with my problem."  Then it's contextual.  But TechCrunch is just new site, new site, new site.  New sites because there are new sites.

Spolsky:  They also report on Twitter being down.

Atwood:  Oh, God, their whole axe to grind with Twitter is very, very tedious.  Speaking of which, you said via email you're actually on instant messaging now? Is that true now?

Spolsky:  Yes, but pretty much--oh, look, I invited you!  But you didn't accept my invitation.

Atwood:  Oh, there we go.  My bad.

Spolsky:  Invite again.

Atwood:  I don't even like you, that's the problem, I don't really want--your advances are kind of unwanted from my perspective.  No, I'm just kidding.  

Spolsky:  Very funny.  I haven't done IM because I used to do IRC in the old days, and I know what a time sink it is.  So I just pretty much stopped with the time sink thing.

Atwood:  Yeah, I'm with you.  I'm a bit of a Luddite when it comes to instant messaging.  Like, I've really, really resisted it for a long, long time.  Now Joel's having an out-of-band conversation with me on--

Spolsky:  Now you just inbanded it, you, you cheater!

Atwood:  I just thought it was funny.  But hey, it's your little Reddit icon, for the Joel Reddit, it's nice.

Spolsky:  Yeah, that's what I use as my avatar these days.  Well, I had to do IM because we opened a second office for Fog Creek, because we just overflowed this one, and we haven't yet consolidated everybody in the new, big office because the new big office right now is a construction site.  Last time we were there, it was almost completely demolished except for some tiles which were stuck to the floor, and there were a couple of old workers sitting on the floor with toothpicks or something, trying to scrape them off, these last remaining tiles.  And I was trying to get them upgraded to wood toothpicks, because I thought that might go a little faster, but no go.  I think it's just gonna take us six months to get this construction done.  I hope not, I hope in September we'll be moving.

Atwood:  I just read your Inc. article, it sounds like this space is gonna be really awesome.  So I think it'll be one of those things that's worth waiting for, it's gonna be really cool.  And definitely, y'know, Office Snapshots--was that the correct name of the site that we referenced a few [episodes ago]?

[23:28]

Spolsky:  I think so, I think that is it, yeah.

Atwood:  People really like that.  Like, a bunch of people, like I had never heard of it, and then you found it and people are like "Wow."  People spend all this time actually browsing it, and I got all this really good feedback.  That was a great find, and I think my point is that it's worth waiting for this stuff, because people really appreciate it.  There's someone out there who cares about this stuff.  Particularly the people who work for you, we don't talk too much about the people who work for you, but I'm assuming they care about this stuff like you do, and appreciate this stuff.  So I think it'll be worth it even if it takes, if some people have to work in the hallway for a while, right?

Spolsky:  There's a couple of people doubled up in offices, including me, but--

Atwood:  What, you have multiple people in your office?

Spolsky:  Yeah, Michael Pryor is right here.  

Atwood:  Oh, nice.

Spolsky:  If I tell any really good jokes, you'll hear some sort of laughing in the background there.

Atwood:  [laughs] No, I don't think we'll have that problem.  Just kidding.

Spolsky:  What were we talking about?  Oh, the new office.  Yeah, the reason I read Office Snapshots, actually, is--everybody else is reading it to sort of daydream about nicer places to work--and I read it to make sure that nobody has a better working environment than we do.

Atwood:  [laughs]

Spolsky:  I'm always looking at it, and I'm like, "Oh, come on, you just put a bunch of chairs in a big room, with some tables!"  And then it's cool, because you've got like little rubber duckies lined up on the top of somebody's monitor.

Atwood:  I think what I'm gonna do is, I'm gonna wait until you guys move in, and I'm gonna gift you guys a Rock Band setup, so you can actually have that for your new space.

Spolsky:  Awesome!

Atwood:  As an office warming gift, I think that will make sense.

Spolsky:  That reminds me, I got a gigantic projector, a gigantic projector screen for the new lunchroom.

Atwood:  You will definitely need that.  So let's see... Another thing with StackOverflow, we actually did our first server deployment.  So we actually took the code we're working on and deployed it to the CrystalTech server that we have dedicated for that purpose.  We had, actually, very, very annoying problems with the deployment, which was sort of amusing, I guess.

Spolsky:  Can I see it? Can I go there yet?

Atwood:  You could, it's--

Spolsky:  Just IM me the URL, and all the people won't hear it, and I can go there and I can--

Atwood:  I will.  So the problem we actually had was, so we're using ASP.NET MVC, and ASP.NET MVC is still under development, sort of.  So we have a little bit of a dependency there that's kind of weird.  So my hope is that by the time, like basically a month, less than a month, even, something deployable will go out, where there's like, what Microsoft calls a "Go Live" license, where you're using stuff that's not fully baked yet, but they guarantee that there's gonna be no more breaking changes in the API's and things like that.  So the next version won't cause you to have to rewrite all your code and stuff like that.  So my hope is, we'll have a Go Live license sometime before that happens.  But we actually entered this weird .NET DLL hell.  You remember, of course everybody knows DLL hell, right? I've never had it in .NET, before!  But this was a first for me.

Spolsky:  Except it called them like "assemblies" just to be confusing.

[26:30]

Atwood:  Well, it was really obscure, just settings in the config file that were very, very, very subtle, that turned out to be really relevant under IS7, but not relevant--so there's two ways you can run the site, you can run it in the Visual Studio embedded web browser, which used to be codenamed Cassini, and of course, you can also run it in IS7 on the server.  It turns out that certain sections of the config file become super-important under IS7, and there are really tweaky things that can be wrong with it, that are very--like I'm talking single character things. [laughs] The sort of things that blow up the space shuttle, right, like "Oh, there was this semicolon in the wrong place, so everything exploded."  That was not fun to deal with.  But we did deploy to the server, which was kind of nice.

Spolsky:  Cool.  We use - while I'm plugging products here - we use this thing called FinalBuilder for everything building and deploying.

Atwood:  I've actually used a really old version of that, way back when.

Spolsky:  The new version's really cool: it's got a web-based interface now, so you can--everybody can kind of FinalBuild from their desks, if you have multiple people.  And it basically knows how to do stuff, so instead of, like a makefile you're basically typing command lines at it, but FinalBuilder knows how to do things like check-out from source code, for all the major source code repositories, it knows how to launch Visual Studio and do a full build, it just knows all kinds of stuff, I mean, it knows about thousands and thousands of different--not thousands, hundreds--of different apps that a developer might use, and it will present you with options like in dialog box form, to launch those apps and do various things.  And it allows you to make kind of a FinalBuild script where it's just one click to get the whole thing built and out on a server.  And then you could set up one that gets it to a test server, and one that promotes it to a live server, and that way you don't have to remember all 37 steps that it takes to get something built and deployed.

Atwood:  Yeah, that's cool.  I think one thing about having a tool that's sort of out-of-band from your development environment is that it's aware of all these different methods of building, and all these other different applications, so...

Spolsky:  It knows how to copy files, it knows how to like, copy this file, that file a little later than that file, and do all kinds of cool stuff.

Atwood:  I've done stuff like that with MSBuild, which is Microsoft's--I guess you would say, "Ant clone".  And it does a lot of the same stuff, and there's actually a set of community tasks you can plug in to it that are essentially .NET code, that assemble [garbled] a bunch of functionality.  Such as check-in, check-out, so as a part of your build you can get everything in the source control, things like that.  But MSBuild is kind of painful to work with, and I'm guessing Ant is even more painful, because I remember in my XML thing that I posted, people pointed to Ant as sort of the penultimate example of bad XML.

Spolsky:  OK, I was hoping that they would say it was bad, I was wondering if they were like, "How could you hate on XML, we have Ant, look how great it is!"

[29:35]

Atwood:  There are still people who are very, very upset about that post.  I was really surprised how deep that went.  Because I just was like "Well, maybe there's alternatives to XML," and people were, like, mortally wounded by that.  Like, [sounding hurt] "I can't believe you said that!"  One of my favorite things anyone has ever posted on my blog is "I have lost all remaining respect for you, based on this post."

Spolsky:  [laughs]

Atwood:  Like, they had a little reservoir of respect for me, and it was really low--

Spolsky:  They'd worn down most of it.

Atwood:  --they had like a half gallon of respect left, and now that's all gone, because of what I wrote.  I found that--that almost killed me when I read that, that was very, very funny.

Spolsky:  Y'know what I find really strange?

Atwood:  What's that?

Spolsky:  There's something that I find really strange, which is that people will send me an email saying "I agree with everything you say, except for this one thing."

Atwood:  Oh yeah, that happens to me all the time too.

Spolsky:  And that's like--that's good, that's cool, I'm glad that you're still thinking critically, but it's sort of--did it ever occur to you that maybe the fact that you agree with everything I say means that you should in some way give me some extra authority, and on this one thing that I'm saying that you disagree with, just open your mind a little bit maybe?  Maybe actually, I'm right about everything?  Is it possible [laughs] or is that--

Atwood:  Do you really believe that you're right about everything? [laughs]

Spolsky:  No, no, no.  But if you agree with everything, then fine--you should agree with about half of what I say, that would be about right--but if you're really agreeing with everything I'm saying, and there's that one thing that you don't agree with, maybe the problem is that one thing, not--I'll give you another example, not to talk about politics in any way, shape or form, but if you go to the Hillary Clinton website right now and look on the forums, you'll find an endless stream of people saying that they love Hillary, they totally support Hillary, they will never in a million years ever vote for Barack Obama (who actually won) and they're gonna vote for McCain instead, and it's kind of weird that they're such devoted fans of Hillary, yet when she says "OK, I lost, let's all vote for Barack Obama, because he's the one who's got the closest platform to the one that I support," they're not willing to take that extra step.  Like they're so devoted to Hillary that they're not willing to be devoted to her in this way.  It's kinda weird.  And maybe it just shows like a real independent-mindedness that maybe it's unique to Americans or programmers, or maybe everybody's that way, that they're just not willing to--even the people that they have the ultimate respect for, they're not really willing to give them respect.  They're not gonna go the last mile and actually be forced to disagree, [delete/believe?] something that they disagree with.

Atwood:  No, it's definitely a similar sentiment, and I've heard it expressed several times exactly that way, where "I agree with everything you said, except for this one thing," and that's another thing I find very amusing.  We also, by the way--I can't remember, we must have made some reference to politics in the last podcast, and we immediately got a negative feedback on that, so we do have to be careful.

Spolsky:  Now you see, this time, notice I was referring to something that happened in politics without taking sides on it.

Atwood:  Yes, good, well-done.

Spolsky:  I think what it was in the last podcast that I said - which I'll just repeat now for good measure - is that the trouble with people using anecdotes instead of actual scientific facts in order to tell little stories about things, it's like: it's all nice and innocent when it comes to blogs, but when journalism fails us, you can go to war by mistake.  When the journalists fail to do their homework.  There's the politics one more time.

Atwood:  Yeah, that's scary.  I tell you, every time I get--I occasionally click through to some political blog, and I'm always really, really sorry that I did, because people get so angry.  The whole political blog I just don't understand it, because it's like--

Spolsky:  But you know, they only read the blogs that they agree with 100%.  

Atwood:  It's like juggling chainsaws.  It's just like a thankless thing, that there's just so much fighting that goes on.  I mean there's enough fighting over just general writing that goes on, but politics just [sigh] I don't get it.  I guess if that's what you're into, and that's what you want to do, I guess it's like lawyers, they get paid to argue, basically, so they become--

Spolsky:  Some people really like it, I mean, you probably knew some people in college that were in the campus political union or whatever, and the idea of having a debate about something that they could not care less about is enormously appealing.  

Atwood:  Of course, here we are talking about politics.  I'm sorry, I've done exactly what we're not supposed to do, and for that I--well, it's a meta-discussion, we're talking about talking about politics.

Spolsky:  Alright, let's talk about--the feedback I got that I kinda liked, actually, various people have been emailing me, some people said they want to talk a little bit more about [the] business side of things. I dunno even what errr...

Atwood:  We still have your book list, we haven't talked about that.  You wanna talk about that?

Spolsky: Yeah, oh jeez. Well I've got 4 question all queued up.

Atwood: Oh no, no, let's do questions. I feel bad too because we've done a lot of blah blah blah and haven't done a good job really getting the questions in the last few podcasts. So maybe that is the right thing to do.

Spolsky: Lets' do a bunch of questions today. I'm gonna start with John Topley asking a business question since some people want to talk more about business.

John Topley: [Intro snipped] What are your thoughts on affiliate programs, such as the new 37signals affiliate program?

Spolsky: Err, no it is a complete waste of time. I love 37Signals - more power to them. What they are going to get: they are going to get people doing this. Half of them are going to be people who would of signed up anyway.  That are making their own account and paying themselves their own affiliate payment just as a way of scrounging a few pennies.  Umm, it's just a huge waste of time.

Atwood: So you don't like affiliate programs at all? In any way?

Spolsky: They just don't drive any business. No.

Atwood: It's hard for me to take that seriously because I got to tell you I derive a significant percentage of my income today from affiliate traffic on Amazon.

Spolsky: It's just Amazon.

Atwood: Well...

Spolsky: Amazon is different. And the reason Amazon is different is that Amazon sells millions and millions of products to millions and millions of people and there are millions and millions people with Amazon links and Amazon accounts and millions and millions willing to buy from Amazon so if you link to a book at Amazon its going to be some substantial percentage of people that click on that link already have an Amazon account are interested in the book, wanna buy from Amazon and do it. But with 37Signals they are not at that scale, they don't have enough products really for it to be worth it for anybody to make a lot of effort building these affiliate sites. What we got, I'll tell you why, because we did it and it was a waste of time and every month I had to write all these $17 dollar checks and mail them to four corners of the earth. It was an enormous waste of time. What we discovered is that the biggest category was people - of actual affiliate payments - was people who set up there own affiliate account to make their own purchase. Which you can make against the rules but, you know, you can't really police that stuff and who cares anyway? So...

Atwood: Amazon polices that very well actually, but go ahead.

Spolsky: Yeah, well they police it because it's kind of a big deal. I mean it's their... a big thing. The thing about Amazon is it's sort of like, it's such a huge exception in e-commerce. It's the only affiliate programs that works, literally. I mean, can you think of any others?

Atwood: Well, I participate in several others. I can tell you what they are.

Spolsky: Do you get...

Atwood: ...eBay just launched surprisingly - they had outsourced theirs to Commission Junction, one of the biggest sort of third party ones that would do it if you want to out source the whole... like the writing of the $17 checks you would go to Commission Junction. They become a middle man, they offer your catalogue whatever it is and people can do it through there. So I've done stuff through Commission Junction, specifically through Newegg 'cause I buy a lot of computer hardware and I tend to recommend hardware - you've probably seen me post about that in the past. So I affiliate it through Commission Junction.

Spolsky: Do you get a lot of, compared to Amazon a lot of money from those other ones?

[37:40]

Atwood:  Well, that's the thing.  Amazon's affiliate program is actually [on] very, very favorable terms.  They have extremely favorable terms, which I didn't really appreciate until I started using other affiliate programs.  I know on a previous podcast we've also talked about sort of the other big dog--well, it's not really affiliate, but AdSense.

Spolsky:  Yeah, that's a different story.

Atwood:  Different thing, but Commission Junction for sure, Amazon, and eBay.  eBay now has their own affiliate program, which is insane that they didn't have it earlier, because eBay's such an 800-pound gorilla. 

Spolsky:  Right.  So these are ginormous compared to 37Signals.  And what happens is that you wind up with a very small population of potential clickers, and so forth.  So I was listening what happened when we did our affiliate program.  Like I say, a bunch of people were just doing it for themselves, a bunch of people just never got any affiliate commissions at all, a bunch of people were these professional affiliate-link building website type people, whose day job - apparently, ever since they got laid off from the peanut-butter plant - consists of maintaining a gigantic website with about a million affiliate links on it, and maybe trying to grab some traffic somehow to it, through some kind of devious domain-squatting type thing.  But the main point is that we would look at these sites that were linking to us, and it would be a page with about 400 affiliate links on it and nothing else.  That nobody would ever really go to, that's being studiously maintained by some stay-at-home mom who (I'm afraid to say) is probably not making very much money on anything.  Maybe on the Amazon links, but certainly not on the links to little software products.  So it was just not doing it for us.  And the other thing is--hey, you just emailed me the other day that old article that Anil Dash wrote about how Google Answers failed because they were trying to pay people.  You get the exact same problem with affiliate payments: as soon as people start thinking "I'm linking to Fog Creek software, or 37signals because I'm getting fourteen cents on the sale" or fourteen dollars or whatever, then they might also think "I've just sold out for fourteen dollars, and I don't really need the fourteen dollars."  As soon as you make it explicit, that you're rewarding them with money, then--this is another one of those cognitive dissonance things, right?--then they understand why they're doing it, they're doing it for the money.  And they're thinking of this tradeoff, "I'm providing this link because of the money."  And that--what psychologists call an extrinsic motivation, the motivation because of something external, the money--displaces their own personal desire to link to you because they like you.  In other words, the best you can do is take a bunch of people who just love Basecamp, or Campfire, or any of 37signals' products, because they're really good products, and all those people that would have talked about you happily and would have linked to you, are now suddenly thinking "Yeah, I'm linking to them and I'm getting four dollars."  And they're sort of forgetting how much they like the product in the first place.  So it sort of displaces that kind of motivation.  So if anything, they actually become less motivated to link to you.  All those people that would have happily linked to you and  then thought "I must be linking to you because your product is good."  And every time they link to you or mention you, they'd be reminding themselves how much they like the product.  And now, every time they link to you they're reminding themselves how much they like those four bucks. 

Atwood:  It was interesting when I read about that, I actually considered "Oh, I should link to them," and I realized no, I shouldn't link to them, because I don't really use their product.  I have used one of their products, like some small project they did on the side, but you should only link to things you really use.  I mean, that's sort of my philosophy.  So you're right, the very introduction of money made me think of it in a way that I shouldn't have, which is I should link to things that I use, as a normal course of talking about it.

Spolsky:  Yeah, you realize that you're losing a little bit of credibility.  I actually--it's kind of interesting, but I actually at one point--I think I was linking--when I linked to books in the early days, I don't even know if they had an affiliate program at Amazon at that time, or maybe I didn't know about it--and when I would mention the name of a book, I would link to Amazon because it was the only website in the year 2000 that had a complete list of every book, and those became the canonical link target for a book, was "Click here to buy it on Amazon" because you would find out more information about the book, and you could buy it right there.  And so that just became, even before it was one of many commercial sites, it was the canonical target link address for a book.  And at some point they added the affiliates program, and I said "What the heck." and I went in and created an affiliate program there, and over time, I've made like $300 or something, and I've kinda come to regret it, because I'd almost rather have a website that I knew was 100% profit-free just because you sort of have more legitimacy when you're not--I mean honestly, if you start linking to 37signals to get those four dollars, then you're gonna lose a lot more credibility than those four dollars are worth.

Atwood:  I don't have much credibility though, so...

Spolsky:  That's right, you only have a little bit of credibility left! [laughs]

Atwood:  [laughs] I have like an ounce or two. They may lose all remaining respect for me if I do that.  But no, that's a great point, and Amazon is a useful link target, right?  Like even if you don't care about the commerce part of it, you would link to Amazon because "Hey, there's user reviews,"  which I talked about on my blog.  There's information about the book, how many pages it is, who published it, so it's kind of--

Spolsky:  Unless the author has his own private page, it's a good canonical link for any book.

[43:28]

Atwood:  Right, that's why it's also a little different when you think about arbitrarily linking to stuff to make money, does the link actually have value, if somebody clicked on it and they weren't really gonna buy, would it be informational at least, so they could learn? Like the whole Wikipedia aspect, which is a non-profit, which is a big difference.  And also, an objective source.  Like, Amazon isn't necessarily objective about the book, because they want you to buy it, right?  So I think that's a great point, and there's a whole related conversation I won't have now about why Wikipedia is such a powerful link target, where I have this opinion - I actually don't think it's an opinion, I think it's a fact - where Wikipedia's gonna own every link there is, but I'm not sure that's a bad thing.

Spolsky:  Uh oh. No, they're always gonna wanna be encyclopedia-like, whatever that means.  They're not gonna wanna own the link for custom corporate mugs with your corporate logo embossed on them.

Atwood:  True.

Spolsky:  And there are gonna be people searching for that.

Atwood:  No, there's probably a category of stuff that wouldn't end up on Wikipedia.  But it's pretty much like a black hole of gravity, though, it really does draw in so strongly that it's hard to resist.

Spolsky:  You think they'll ever have all the little minutiae about programming that StackOverflow's gonna have?  

Atwood:  No, no. I guess that's true too, there's also stuff that can fall between the cracks.  And it's the whole long tail thing, right?  I mean there's maybe, let's say, a million top-level topics that Wikipedia will totally own, right? As number one link results.

Spolsky:  But they're real aggressive about weeding out stuff that they just don't think is interesting, like every single character in the Smurfs.  Well, actually that's probably still there, but there's large categories of things they kind of aggressively weed out, and the say "This is just too detailed," or "not encyclopedia-worthy."

Atwood:  Well, I don't wanna talk too much about Wikipedia because I want to get to more questions, but actually what you're referring to is "inclusionists" versus "deletionists." So there's two schools of thought there, one is "Everything there ever could be should have an article," right?  That's inclusionists. And then there's deletionists, which is, the classic Wikipedia thing is--I forget the exact term, but it's like "This is not notable," I think that's it, "not notable"--

Spolsky:  Will there ever be an article that's like "Items that were in Joel Spolsky's trash can on June 6 2008"?

Atwood:  Y'know, the inclusionists would say "Hey, if people wanna read about it, if someone's motivated to post it, then by definition there's enough people in the world who care about it, why not?"  Because it's not like you have a physical bookshelf.  It's not like you're taking up space, so it's really interesting.  And, coming back to something we talked about earlier, that's actually the blog entry that Nicholas Carr linked to, was "Inclusionists vs. Deletionists," because he was talking about Wikipedia, and evidently had done a Google search and found my little article on it.  It's an interesting topic.  But hey, let's get to the next question.

Spolsky:  Alright, here's a question by Matthew Glidden, I don't remember what it's about, but I'll play it anyway.

Matthew Glidden:  Hi, Jeff and Joel, this is Matthew Glidden, from Cambridge, Massachusetts, and I've been listening to each of the podcasts and enjoying them.  I had a question that related to a blog that Seth Godin recently posted, where he posited that it was better for people to be the best specialist you could be rather than a jack-of-all-trades, which would I think be part of any sort of startup, you'd have to put on so many hats that even though you had a particularly strong skill, you out of necessity would have to handle other things.  And I was wondering what you thought, if you had read what Seth posted about this a few days ago, and--

Spolsky:  Did you read that, Jeff?

Atwood:  The thing he's talking about from Seth Godin?

Spolsky:  Yeah.

Atwood:  No, I haven't.  But I have read the quote "specialization is for insects," which I thought was funny.  I think there's really two stories here.  There's the whole long-tail principle of--you talked about it earlier with your example of the communities from Clay Shirkey's book--you have these communities of really unusual things, where people have been able to find each other through the Internet.  That it's kind of obscure, or so weird that there's just not enough people in the world - say in Des Moines, Iowa - to find each other and do this thing.  But on the Internet they can come together, right?  There's a large enough number for them to do that.  So that's specialization, so I think it works in both dimensions, and I think it depends what problem you're trying to solve.  I think there's a place for being a generalist, and there's a place for being a specialist.  When it comes to software development, I'm a little more torn there, because I'm definitely a generalist, so I'm gonna be a little defensive about that.  I don't get really narrow.

[48:27]

Spolsky:  I think a lot of people people misunderstood what Seth was saying.  He basically said, and I'm going to read the quote from Seth's website: 

If I need an animator, I can find the world's best animator. If I need a bond to insure my movie, I can find the best broker at selling completion bonds.

[...]

Not the second-best or someone who will try really hard or someone who is pretty good at that and also good at other things. [...]

When choice is limited, I want a generalist. When selection is difficult, a jack of all trades is just fine.

But whenever possible, please bring me a brilliant specialist.

and I think where Seth is coming from here is not talking about whether a person - and this I think Mathew is where you're misinterpreting it a little bit - there's a marketing concept (which Seth is obviously familiar with, being a marketing expert) called positioning and the concept of positioning came from one of the books on my list - I think - by Ries and Trout which is called (surprisingly) Positioning, and the idea is that a given product or a given company or a given product line has to occupy exactly one slot in people's brains.  So, for example, Hertz can be the car rental agency; Avis can't have that slot.  And so Avis had to come up with a different slot, they came up with "We're number two, we try harder" and they occupied a slightly different slot in people's minds.  And similarly, Microsoft with Windows came up with "We're the operating system that runs all the computers on everybody's desks" and so that wasn't available when Apple needed an operating system [sic] and so they have to be the cool home alternative computer that's really powerful and sleekly designed and so forth and they're trying to occupy kind of a different position in your mind.  So similarly you know, if you were going to go to a restaurant and you wanted some Thai food, would you go to a Thai restaurant, or would you go to a restaurant that sort of had Thai, Chinese, Japanese and Mexican food all on the menu.  I think you might guess that the Thai restaurant is going to be a little bit better at Thai food, that's what you really want.

[50:41]

Atwood: What if you want a burrito on the side, you might go to the-

[50:47]

Spolsky:  [laughs] no, you don't want a burrito, no you want Thai... Yeah, so the marketing experts always have this approach that it's better to have a narrower product rather than a wider product for many reasons.  And in fact, one of the reasons that so many of these Web2.0 companies fail is that they try to be too horizontal, they try to have a product for everybody and they end up getting nobody, because they're just not being specific enough and people have very specific needs, and if you have a very specific product that meets someone's specific need, that person will find you - especially in the age of Google - whereas if you have a product that solves everybody's problems then you won't, really.  So for example, if you're a plumber, and you're just getting started, and you're putting a big yellow-pages ad, you can just say "I'm a plumber, I can do anything.  I'm kinda good at all kinds of stuff, and I'm just learning, but I'll be able to do just about any plumbing task you have," and you can put that ad in the yellow-pages, and you'll look just like every other plumber that's there, and you won't get any phone calls.  On the other hand, if you say "I really understand those German Dornbracht faucets.  I'm an expert on Dornbracht faucets," it might seem like you're reducing the potential audience of customers that you have, but what you're actually doing is--you have no customers, what you're doing is, you're now attracting a bunch of new customers who will say "Oh, good, I got a Dornbracht faucet, I'm gonna call that plumber, because that's the plumber who knows about my kind of faucet." And all of a sudden, you start getting calls.  And that's sort of a general marketing principle: if you're having trouble attracting an audience, the first thing to try doing is figure out if you could just narrow to a much more specific audience and cater specifically to them, provide a better product for them.  So I think that's what Seth was talking about.  I don't think he was really talking about--like, the CEO of a company, especially a startup, does need to wear every single hat, and needs to be the accountant, and the programmer, and the system administrator, and the payroll specialist, and the HR specialist, and all those things, the marketer, for their company.  That's definitely true.  But I think a startup or a young company is always gonna do better if they can narrow the number of potential customers that the serve, rather than trying to go off after a horizontal market.  

Atwood:  Right, I think that's a great point, and I love Seth, I think he's got a lot of really interesting things to say, but sometimes I think he does a little--he does engage a little bit in the marketing doublespeak.  I mean, he does it in a very cool way, that makes you feel sort of enlightened, but it's interesting on this post (you sent me the link out of band, which was nice) he puts a whole coda at the end explaining his position, because I guess a lot of people misunderstood what he was saying.  And basically, he's saying two things at the same time, he's not really taking a position, he's saying it's OK to specialize in being a generalist.  Well, what does that even mean?  Specialize in being a generalist?  Now we're in 1984, right?  But still, it is good to think about this stuff, as long as you don't take it too seriously.

Spolsky:  Specializing in being a generalist would be being the kind of person that's like maybe a management consultant, where we hire you because you're willing to see the big picture, and you know a little bit about everything, and you can actually come in and give sort of a general solution.  Or, let's take an example of, a startup company may need somebody to be a CEO who is a generalist, and knows about lots of these things, and that's what they're good at.  Whereas maybe at another phase in the lifecycle of that company, they might need somebody who's really good at sales, or marketing, or technical skills, or something like that.

Atwood:  Right, and then to your previous point of where it really is valuable for a business perspective to specialize, I like Larry O'Brien's blog.  He wrote a neat entry where he talked about how his blog helps him get business, and he pointed out something interesting, that he made this one blog post about Sabre, which is, I think, the airline reservation software, it's been around a long time.  

Spolsky:  Yep.

Atwood:  And that single post had resulted in more business for him than all--he had been very into, I think, pen, the pen stuff he had been doing, and he was like "Yeah, I spent a lot of time on this pen stuff, but it never really resulted in any business for me.  Whereas, this one Sabre post I made resulted in all this business for me."  So that's an example of specialization, right?  The people who care about the Sabre stuff found him, probably through Google, and saw that he knew about it and were hiring him, and so on and so forth.

Spolsky:  There's a real tendency, you'll see a lot of young people that start out as independent IT consultants, and they'll take any job, they just want a job, and so people say "What do you do?" and they say "Well, everything, really.  I can learn any programming language, any API, whatever you throw at me I'll learn it, I'll be good at it, it's fine, I can do all this stuff.  Do you want me [to be] a system administrator, do you want me to write code, do you want me as a DBA, I can do all that stuff, and I'll learn it, and I'm young, and enthusiastic," and this is just not appealing.  Whereas if they came out on the very first day and said "Y'know what I'm an expert at? Stored procedures in Microsoft SQL Server 2000.  I don't really know much about this 2005 thing.  And I don't know about functions.  Or views, or SELECT statements.  But stored procedures?  Boy, do I know that cold!"  And y'know what?  There's gonna be somebody who needs a consultant and wants some help on stored procedures in SQL Server, and they're gonna find you.  And now you have a customer.

Atwood:  Right.  Well, that does run contrary, because if you just call the sales department, the answer is always "Yes," right?

Spolsky:  Well, OK, if you get somebody calling you, fine.  There's a difference between sales and marketing.  The sales department always says "yes."  See, there's a tendency to think the audience of my potential customers, the number of people with a problem, the percentage of people in the world that have a problem, times the problems in the world that I can solve, and so the larger the number of problems that I can solve, the larger the number of people that are my potential customers, and therefore, the more customers I should have theoretically, and that seems to make a lot of sense.  Unfortunately, you haven't really explained how these people find you, and why they would want to use you.  And it's real easy to specialize.  You can be the restaurant that has outdoor seating, so you can bring your dog--and you can just put in an ad, saying "The food's not so good, but boy are we dog-friendly!  We'll bring your dog a little bowl of water and a treat while you eat your hamburgers."

[56:57]

Atwood:  Right.  Yeah, no, great point.  And I do love Seth, he's got a lot of really interesting things to say.  But I think that sometimes there's a little bit of double-speak there.

Spolsky:  The marketing

Atwood:  Yeah.  It is marketing, you know?  Very good marketing, it's the kind you sort of like yourself for reading it, but it's still marketing!  You know, let's be clear about this.  But yeah, so let's do another question, that was a really good one.

Spolsky:  Okay, well here's an answer actually:

[57:22]

Jim McKeeth: Hello Jeff and Joel, this is Jim McKeeth and I have the solution to your password storage challenge: what you do is reverse your premise, instead of trying to synchronise stored passwords across multiple computers, instead you need a reproducible way of generating a secure password.  And the solution to that is passwordmaker.org, it's an open source program that uses a secure hash, your super-secret password and the URL of the site you're wanting a password for to regenerate the same password every time for each site.  Since it uses a hash you can't reverse it, they can't ever get your super-secret password, all your passwords are independently secure and reproducible.  Check it out: passwordmaker.org

[58:11]

Spolsky: Did you get that?

Atwood: Yeah, I'm looking at it now... but to me this is another piece of software that I have to install, which makes it kind of... I don't know, it's not my preferred solution.

Spolsky: Uh, no, but this is the cool part: it never occurred to me, that it's not a way
of storing a password for every site, it's a way of consistently regenerating the same
secure password for every site. So you don't have to store anything anywhere.

Atwood: Right. I understand how it works, I'm just a little torn by the whole concept.

Spolsky: Yeah, sick and tired of it? Oh look, they've got a widget, they've got a Javascript, they've got a Yahoo! widget. You know, if they have a Yahoo! widget you could
just add it to your My Yahoo page, you don't even have to install.

Atwood: I don't know. I can be a little... difficult when it comes to topics like this.
Nobody has to use the solution that I use, which is I feel some of this stuff should be in
your brain. Of course I'm the guy who doesn't even like using a tool.

[59:06]

Spolsky: Wait, in your brain? I have like 600 websites that I had to make accounts for.

Atwood: Right but somebody made point in a comment in a post I made, which was that you only really have 3 passwords; You have the super secure one you're going to use for the things that matter, like money stuff.

Spolsky: Yeah.

Atwood: You have one you use for stuff that is kind of important, you want a password that doesn't suck too much there. Then you have passwords for sites that, really, does it even matter? If someone compromised your password to this dinky little, like hammocks.com? Does that matter?

[59:44]

Spolsky: Yeah, they've got my credit card at hammocks.com! But this is not the... I don't, I mean that would be good, but I actually use Password Safe and I generate a unique strong password for every site.

Atwood: You're like a hardcore geek: that's why.

Spolsky: Yeah, 'cause after a while, it's possible to get, to be, anal retentive in every
possible direction. You can get obsessive compulsive about... [laughs]

Atwood: And again, it's good to bring these solutions to light, so everybody can find one
they like. Just because I don't care for this one, doesn't mean it's wrong, of course. But
you know, I'm definitely on board with this whole OpenID thing, I mean I really want to
see that take root, because I believe in it [laughs]

Spolsky: This looks kind of popular, Password Maker, "5 tools to bullet-proof Firefox"...
OK, anyway, www.passwordmaker.org, thanks Jim for that suggestion. There's some other suggestions in the inbox, maybe we'll get to them next week. Anything else we have for this week?

[1:00:32]

Atwood: Well I should point out, it's actually not "www", it's just passwordmaker.org.

Spolsky: That is true. The www, it...

Atwood: [laughs]

Spolsky: What happened?

Atwood: I had a funny blog entry, well I thought it was funny, this show "Home Movies",
which was an animated thing...

Spolsky: It's a sex site!

Atwood: What?!

Spolsky: No, I'm just kidding. Just kidding.

Atwood: They had a whole skit about whether you should say "www dot",  you know, I just found it very amusing. My wife and I, we would quote it to each other all the time, so
I'll put a link to it in the summary. But yeah, it is annoying that when you talk about
sites, you actually can get it wrong. Like "Oh, I said www but it doesn't actually belong,
doesn't begin with www", or vice-versa, say you said "passwordmaker.org", or "well, it's
actually www.passwordmaker.org" but is that inferred, right? Do you really need to say it?
Is it really... I don't know. The whole thing is just annoying. You know, I wish we had
kind of just decided on one or the other and everybody did it that way.

[1:01:27]

Spolsky: You know what the number one search used to be at google.com?

Atwood: What?

Spolsky: "google.com"

Atwood: Oh, right. Of course.

Spolsky: And I think number two was Yahoo. [laughs]... The reason, they finally figured
out, is because they have a little bit of Javascript in the page, that moved the focus to
the search box. So it's not that people were searching for "google.com" on google.com
because lots of people are retarded in that particular way. It's that they just started
typing "google.com", thinking that the focus would be on the address bar, like it's
supposed to be, only to discover that Google had stolen the focus and put it in that box.

Atwood: Right. Oh, the focus-stealing. Man, that drives me nuts.

Spolsky: Right, they actually take the keyboard focus away from the address bar and put it in the search box, which just saves you reaching for the mouse, so it's a nice thing,
but...

Atwood: I think that's... I know we're at one hour, but I want to spend 2 minutes talking
about that, I did have a blog entry talking about "the evils of focus stealing", which can
be really traumatic... depending on what you're doing, but...

Spolsky: Yeah, you could be typing.

Atwood: The browser one is particularly egregious, because the default Google search page in Firefox does that and I had to finally turn it off... [phone rings]... Sorry, we'll
have to edit that out. I forgot I had a phone here, I get so few calls.

Spolsky: [laughs] That was me, actually. [laughs]

Atwood: I had to actually change the default behaviour of Firefox, in that regard. I use
about:blank, that's always my homepage, because that's what I like. That's my homepage.

Spolsky: Yeah, I used that for a long time.

Atwood: Yeah! But it was really traumatic because I would fire up a new browser, press
Alt-D, to go to the top, and all of a sudden I'm typing in the Google search box.

Spolsky: Or half the letters are in the top, and then half the letters are down in the
Google search box, depending on how the Javascript thing launches up.

Atwood: [phone rings] Right. Oh, this phone: it's not good, it's not good.

Spolsky:  Something even more annoying is the -- just throw it at the wall, real hard.

Atwood:  It's actually a call I need to take.  So let's end it just here.

[1:03:25]

Spolsky:  Let's end it just here.  You have been listening to StackOverflow podcast number nine, if you have feedback, suggestions or anything you want to talk about, record in ninety second sound clip, MP3 or Ogg Vorbis or whatever the case may be and email it to podcast@stackoverflow.com  Thanks for listening and we'll see you next week.

Atwood: Yep, bye.

[1:03:45 Ends]

Outro, advertising

[1:04:41]