I agree. Free drinks are basically worthless--to the point of it seeming like a desperate come-on. Free food just means, "You are expected to stay here for lunch and talk business." The big questions are:

1. Money, benefits, vacation.

2. Do I get an office?

3. How much time do I spend fixing other people's bugs?

For an Italian it is very hard to understand how it is possible that the kind of advertising Google, Xobni, and other companies do in order to attract employers can actually work.

I mean, to have something to drink it's ok... and even the cool Ikea desk, but this here matters 0.001%... instead the stress is in the following stuff:

1) What I'm going to do there? I'll have chances to learn more? 2) Salary, is high enough? 3) Environment: it's a too big and chaotic city? Are my coworkers handy or a bunch of antisocial nerds?

After one week two displays, the cool furnitures, and so on will matter zero, it's all the rest that will do the difference.

Of course I don't think you need to attract Italian or European hackers at all, but just to share this difference.

antirez - your points are valid. This page is intended to show that Xobni is a fun place to work. Let me address 1-4.

1. I can't say exactly what we're doing at Xobni yet except for "We're making email suck less" :-) The work here is at the intersection of Machine Learning, NLP, systems programming, and UI stuff.

2. The salaries we offer are very competitive - more than at most competitors. We want people to be happy and not have to worry about money in SF (an expensive city).

3. San Francisco is beautiful. From the Xobni office, it's 10 meters to shopping, 2 blocks to Union Square, 4 blocks to world-class museums, 30 minutes to the ocean, and 4 hours to skiing. Hard to beat that.

We actually give people three displays (not two).

Sounds good? http://www.xobni.com/jobs.php

Cool! Only start a transaction when you reach a threshold - makes perfect sense! Essentially, they solve the micropayments problem by using lazy evaluation.

No, but an IRC channel listed on news.yc would help funnel people to the channel rather than just when someone randomly comes across it in a comment.

sunglasses?

Apparently a lot of the people who are making this kind of comment about the Facebook API haven't actually given it a good look. Their original API does exactly what she's asking for: it let's you use Facebook's friend data, among other things, on your own external sites. This new version still does that, but it also allows you to embed your application inside of Facebook, as well as add data from your application, like Twitter updates, to people's profiles. If you don't want to embed your app inside of Facebook, you don't have to.

Facebook has added features that make it far easier to find and use external applications based on the API, but for some reason people think that's a bad thing? I don't get it. Yeah, if your app relies heavily on Facebook you'll be locked in, but every other API has that same effect. It's the nature of the beast.

I think the reason why this is throwing people off is because Facebook really isn't a discrete application itself. When you use the Google Maps API, you're building on top of that map and embedding it directly into the page. You basically get everything Google was already offering in that widget. With the Flickr API, you get to do neat things with pictures, and put those in your pages. With Facebook, what would you be embedding? At its most basic level, Facebook is the friend graph, with all sorts of other things piled on top of it. Most of what you care about is who is friends with who, and it's harder to visualize that than it is with most other sites out there that offer APIs.

But anyway, they didn't get it backwards. They made the API more valuable to third party developers by integrating it into the site more, while still giving you the freedom to have your entire site live outside of Facebook. In turn, that will make Facebook more valuable to its users, which is kind of the whole point. I guess people expect offering APIs to be some sort of act of altruism.

First off, I like the elegancy of the CPS approach, however...

The continuation-passing approach doesn't scale. Not only to multiple machines, as papersmith points out, but on a single machine. news.yc has (had?) problems with a theoretical DoS which pg has acknowledged in email: http://news.ycombinator.com/comments?id=18083 (post formatting corrupted by pg hiding then unhiding the post).

Dave Roberts has an interesting write-up of the approach: http://www.findinglisp.com/blog/2004/11/web-application-design-rest-of-story.html In email he said that talking about it with pg one time, pg made clear that for Viaweb it was only used for the web site the merchants accessed, not the many hoards of customers who accessed "normal" links, and only the links on the current page for that merchant worked, i.e. if you pressed the browser's Back button all the links on that old page were broken and the server would just present you with the current page again. In summation Dave said if your site is small enough continuations may work fine, but if it's meant to be the next big thing use REST and I agree.

Sounds like a bad plan to me.

On the other hand... maybe testing the waters with a few different things, and being flexible and willing to experiment might not be stupid. I'd guess that it's better to be sequential rather than parallel though, in order to focus on one thing at a time.

I'm working as an employee for one startup and a founder of another, in two totally different business domains (financial software and internet games, respectively).

It's hard. Startups take more out of you than a regular job will, because you can't just switch into maintenance mode. There've been times where my boss has had a demo for an important client one week (making me work nights & weekends for a couple days to get it ready), and then I've gone straight to preparing a launch for the startup.

I don't think I really could've done it differently, though. I wanted to work at a startup so I could get experience & a sense of what I was getting myself into, and this is my first job since college. The day job is essentially funding my startup - it pays better than YCombinator, with similar advice & experience benefits. I thought about switching to a less intensive job at a big company, but I dunno if they'd hire me when I already have a startup on the side. Plus, I have a pretty good relationship with my boss and the IP policy is relatively lenient at my current employer, reducing the chance of IP problems (he doesn't care as long as I'm not starting a competitor, basically).

Overall, I wouldn't recommend it unless you're in a similar situation - first job out of college, self-funding, and want the training of a startup. If you're older and already have plenty of experience with business, I'd get a job at a big blue chip where nobody knows what anybody else is doing - a coworker said that Thomson Financial is great for this, almost everybody has their own private business that they run out of their cube because the company is incapable of getting anything done itself. If you can get YCombinator or similar funding, I'd take that: you're trading equity for time, which is usually a good tradeoff. If you can afford to, quit your day job and live off savings (I'll probably do this fairly soon, but I want a little more runway so my startup has a greater margin of safety).

Your points definitely make sense. I spoke to these guys recently. They seem pretty bright and what they're working on makes a lot of sense. I'm guessing that they are pretty competitive on the points you bring up regardless of what random perks they are leading with here.

Thanks guys. We'll be pitching at The Next Web (Amsterdam) 1st of June and from there on out market a bit more aggressively. Software's almost done.

--Reinier, http://tipit.to/

You do realize that this would be the death of so many startups, right? They're already demonstrating that they would rather lift ideas from other companies than buy anyone.

I don't care for the founders much at all. They obviously got very lucky in choosing their product. Their technology is as boring as they come (PHP et al.)

More that I don't like about facebook, the product:

They "sell" privacy. So much of their marketing revolves around the idea that what you do on facebook will only be known among those you want it to. And yet here, facebook fails miserably. It lets in exactly the worst people: law enforcement, employers, school officials.

Oh and the content of their photo galleries is trash because the design encourages people to upload every single picture on their harddrive. Early, early myspace even had them beat in this area.

I don't know, I just never liked it or them much at all.

In Italy Ikea is very cheap _but_ considered cool. The smart single living in Milano will have ikea furniture, and so on.

About Aerons and several screens, again this does not match the EU hacker, that mostly uses a single screen (Quake players have mulitple ones), a common chair, but instead vim/emacs, fvwm2 or other high-productivity window manager and so on. Again a cultural mismatch, I don't want to say it's bad, just this message is not going to work outside the US.

Just pique your interest, ok, but if what they are looking for are real smart people they should know that this kind of advertising will hardly work. It's much better to talk about technology, freedom in the hour you should reach the office in the morning, $/cost of the life there, ...

How many of you have a comprehensive time-phased marketing plan for your startup that goes beyond getting on digg, techcrunch and delicious?

Mostly out of laziness in not wanting to have lots of networks open, I opened up a #news.yc on freenode, which has a few people on it from time to time.

Greetings All,

We are currently looking for excited motivated people to help build the premiere live video website online. If you have extreme hacking skills (python, ruby on rails, javascript, flash action script) and you live in the San Francisco area (or are willing to relocate) please drop us a line!

Feel free to email: info@justin.tv

Peace,

Justin.tv Crew (Justin, Emmett, Michael, Kyle, and Jacob)

PS: Look out for our new release early next week - lots of fixes, a new justin.tv channel, and more...!

Unique IPs/day are still climbing. Pageviews per IP are down, however. I think your theory that it may have something to do with application cycles is correct. There are now about 50 people who are extremely busy with new startups.

Whoa, Youtube did what? Link me to that story!