It's true that Jobs isn't necessarily the best presenter in the world. But I haven't seen anyone who can continually present their products with the purpose, passion, and pride that Jobs does, all of which are on display here. (No doubt because, as you say, he so often reveals monumental products.)
Someone could give a one-off presentation that might wow you more than any single Jobs presentation. But I can't think of anybody in recent times who has delivered more total wow over the course of his lifespan.
Absolutely! I love Jobs. I've tried to watch every keynote he's given. He's magnificent. But I feel that if he was told to give a keynote about a worse product, he'd refuse outright. A part of his passion comes from his belief that his product is the absolute best.
It's also why he's such a good salesperson. He's got the best product. It's also how the Apple web site works: they have an almost pornographic display of their material. I love it. It sells better than any hype could.
I tend to find myself influenced by the karma count more than anything else. I have to log myself out after commenting to increase the cost of checking the change to my net karma.
I'm a strong advocate of hiding the karma counts entirely (for users, submissions, and comments). That greatly reduces the conditioning effect, not to mention karma-related drama and groupthink (as Surowiecki argues, crowd wisdom is best procured when evaluations are performed independently). Submissions and comments can still be ordered by net karma over time, as they are now.
Whatever you choose to do, I think there's a strong argument in favor of experimentalism. Especially since the site isn't a commercial undertaking and can deal with the occasional disruption.
Which, of course, helps Amazon figure out what it doesn't sell, but should. I'd say it's a brilliant idea, but I can't help thinking, what took them so long?
Have you never seen a feature implemented that was (a) kind of obvious, and (b) a long time coming?
I can appreciate that not everyone thinks this was an obvious idea. I'd appreciate it even more if you can respect that I think it was fairly obvious. Thanks.
My tool for eBay had an E-mail to a Friend feature before eBay did. eBay later added that feature, and these days it's prominent on top of every auction.
My tool also had tabbed web browsing before Internet Explorer did, a Microsoft product.
My tool also had a popup-menu where a user could scroll up and down via a mouse (to enter bids from a menu rather than having to type them in) before the iPhone even existed (with their calendar date picker), an Apple product.
It itself was a desktop app before eBay every announced theirs.
It's really not whether a feature is obvious in hindsight. I've done stuff that seems "obvious" right now (never mind just "thought it was obvious in hindsight without implementing anything"), and it was new. You can't think of it that way. Amazon's new feature is quite new and groundbreaking.
But I can see your point. It's because it's a big company, that when it finally got around to implementing something like that, that it's a miracle. In fact, it wasn't until very recently that users have been allowed to select quantities when adding an item, versus in the shopping cart.
And of course, they are a big company, and so part of the news is that they had the panache to allow tracking of other companies products.
Funnily enough, my tool did that exact same thing, earlier--the entire point was to track items a user was planning on buying from eBay, Yahoo!, and Overstock auctions; Amazon was the fourth but I never got around to implementing screen scraping for their auctions, for various reasons, though it had it's own ID reserved in my code/database, as it was originally a planned feature. I think my YC application mentioned creating something similar, but a web app.
Note that when these four or so companies added the features I mentioned to their own products, boy did they get tons of press and praise! :)
And regarding features that are "obvious" that I haven't created, it's not a useful metric. There have been many YC companies that have done something I've thought of, but didn't: an eBay seller tool, a program that runs on your computer and tracks your time and the apps you are using in a web app interface, a tool that backs up your files on your computer via a local network to other computers whenever two or more are running, encrypted and seamlessly, a hands-on survey app for the iPhone, an expense tracker, a work log, all before those companies started YCombinator or switched their idea to their current one. It's all obvious. It's all execution.
Off topic, but interestingly enough, some of the earliest users of my bidding tool for eBay were from San Jose, CA.
Nod. If I had a nickel for every web app I've thought of and could have executed on but didn't, only to see others take the same idea and launch...I might have a buck fifty.
I should qualify that when I say the Amazon feature is "obvious", it's only from my perspective as a user. I would want a site-independent wishlist for the simple reason that I wouldn't want to manage wishlists on multiple sites. Only now, however, is it clear to me what the business incentive is for that.
Perhaps it only appears when a given number of new stories have been submitted since you last looked at the page. Like, say, 20. Possibly user-configurable.
I really like this idea to put a number next to new, making it feel like an inbox. People will end up clicking it based on their need to "empty the inbox" built up from email management.
DEAR GOD NO! Please don't do this, most of us already have enough inboxes.
The only thing I like about non-time-series sites like HN is that it just isn't in their nature to ever be something inbox-like. It's the sole saving grace of having a voting system, especially with vote-ranked threaded discussion -- it keeps it from ever being something I have to "catch up with" regularly.
That's probably the most minimalist incentive you can give to increase New pageviews, and fairly easy to implement. (Edit: updated the original post with this suggestion.)
I have downvote arrows on comments, but at some point they ceased to affect the score. Which is to say, I click on the downvote, and the points go down by one, but upon reloading the page, the score is the same as before.
(And yes, it's happened enough times that I know it isn't just someone voting the comment back up in the intervening period.)
I have downvote arrows on comments, but at some point they ceased to affect the score.
It's part of karma-grump protection. One's downvotes do not count, unless he upvotes first and at least as many times. If you run across something you want to downvote, first find something else to upvote. If you want to rack up downvote ammo to spare, upvote often. For example, if you upvote 10 times in a row, you will have 10 downvotes available. After that, your downvotes will no longer count, unless and until you do more upvoting.
I'm considering adding an additional layer of protection against grumps: to only allow users as many downvotes as they've made upvotes. That way each user's net contribution to the global karma would be at least neutral.
Thanks for pointing this out. I suspect there may be something genuinely broken with my downvoting, though, because despite having only upvoted for a good while, my downvotes continue to not count.
Alternatively, HN is keeping a backlog of my uncounted downvotes, and counting them in order as I make more upvotes. Thus, as long as there's a backlog, any new downvote I make will initially not count.
Personally, I think the >=0 net karma rule goes too far. I've tended to downvote more than I upvote because I drift towards policing behavior (not "grumpness": not all people who downvote more than they upvote can be classified as grumps, which is a highly subjective notion anyway), and generally have high standards for what I consider worthy of an upvote. Social moderation systems have to account for the variety of sociological roles. Not to mention, the net karma of the site may be negative, but a >=0 rule will never allow this state to be represented.
It is also impossible to determine what my net karma contribution is, and thus that the >=0 rule is even in effect. Net karma contribution should appear on a user's profile page, or at the top menu bar. I shouldn't be expected to track my karma contribution manually.
because despite having only upvoted for a good while, my downvotes continue to not count. [...] I drift towards policing behavior
If you have downvoted a single user many times in a row, you might have triggered an anti-karma-bombing feature. http://www.google.com/search?q=site%3Aycombinator.com+karma+... I would email pg at this site to inquire, and be ready to apologize if an apology seemed called for. Good luck.
Certainly good advice but it may leave HN readers with the idea that hiring a VP of Sales person early is enough. Blank explicitly rejects this and suggests that early sales are the responsibility of the founders. In particular they have to treat their product as a hypothesis, which few successful sales folks are able to do. They treat objections as something to overcome. Only the founders can balance the question of whether they are talking to the right prospects with the right message about the right features.
My apologies. What follows is only a portion of what's in the talk.
1. The founders must sell. They must listen to prospects.
2. They cannot hire a VP of sales until they have learned how to sell the product.
3. The founders should view their business as a guess.
4. They can see if they are correct by selling, which will confirm:
a. Who the customer is (who will pay for the product)
b. How to talk about the product
c. If product has the right features
The word he's looking for is "gestalt". Objects that are grouped together are perceived to be related.
Personally, I would not have put those global navigation links anywhere near the user-specific links. Put global navs next to the GitHub logo, which is the global signifier supreme. All the variations Ryan proposes are merely slight improvements on a fundamentally flawed design. It's like trying to sharpen a spoon when what you need a knife.
The user-specific navigation choices are: all | public | private | sources | forks . You can put these in the upper right block instead.
By the way, you can learn the underlying principles behind this stuff very quickly by flipping through an HCI textbook. Just find a cheap, used copy of a well reviewed textbook online. Ryan's post has nice mockups of a real world example, but barely scratches the surface of the theory. More to the point, he entertains a totally flawed approach all throughout.
Which is not particularly well reviewed, but it is adequate, and you can apparently get the 2nd edition for $1.27 + shipping. It's major problem, as with most textbooks, is that's it's written in dry, overly qualified prose. But it should still be skimmable.
Honestly, these things are jam-packed with good, reliable information. I thought I knew a bunch after years of doing Donald Norman-esque analyses, but I still got a lot out of it. Just remember that it's less important to remember the theory than to have been exposed to it. It's enough just to have the concepts rattling around in your brain somewhere.
Someone could give a one-off presentation that might wow you more than any single Jobs presentation. But I can't think of anybody in recent times who has delivered more total wow over the course of his lifespan.