benjoffe

One of the most important things to get right with lightbox libraries is making sure they work well when the user has already scrolled down the page, this one fails in that case: http://i.imgur.com/Gh1Vd.png

Most clients don't allow this option as it makes the network less efficient. The swarm is efficient due to the fact that users download the pieces randomly and therefore are able to re-seed pieces that a low percentage of other users have.

That's still an interface (perhaps I'm being a bit too technical).

At least with the JavaScript version you're serving actual text which can be read by screen readers, copy/pasted etc.

I've been waiting for this since 1996! After ten years without any sequel I asked Adam in an email if I could make a free tribute sequel since it had appeared to become abandonware and got this reply:

"Hi, We are planning to make it but haven't had an opportunity to start it. Maybe in a couple years.."

A few years later it seemed really puzzling, and frankly I thought he was in denial that the sequel would ever be made. So excited to see this finally come to fruition!

The problem with your multiple device solution is that it puts burden on the recipient, who's highly likely to have any idea what NOFWD is. The fingerprinting solution you outline next is highly flakey as well, as it would require a critical mass of users to gather the necessary data, and even then the kind of algorithms you'd need are orders of magnitude more complex than the ones Facebook etc. use, as those catch instances of users logging in from foreign countries (or perhaps states), and would not catch cases where the 3rd party is in the same city, a very likely case.

A typical scenario I imagine is that the recipient will lose the message, and end up emailing or calling back for the same information, which will annoy all parties (or possibly worse, they'll lose the information and ignore it forever), as well as require the information be transmitted more times than usual (something someone security conscious probably won't like).

I hate to be so critical of this service, but I just can't see this being useful. I wouldn't call this hitting a nail with a sledgehammer, it's more like hitting a nail with a six foot ceramic feather.

Personally I would hate to receive email like this, if I viewed the email on my phone I won't be able to view it on my work computer. If I'm working from home the next day I won't be able to view the message (ever again!). If I want to view the email in several months time I'm assuming it's highly unlikely I'll be able to.

Maybe there's a market for this, and if so good for you, but my honest feeling is that any desire for such a 'protection' is better solved by either talking to the person on the phone or in person, or getting them to sign an NDA, or if the recipient is so untrustworthy then don't engage with them.

When I view ~10 year old web sites and forums it's rare enough that external links work at all, with all these link shorteners it's going to be even worse.

I imagine in the not too distant future that link shorteners will no longer be used (for one of many reasons), if/when that day comes most link shortening sites will have no incentive to keep their service available, and it would be hugely tempting for them to modify existing links to display relevant targeted ads, instead of the destination pages. Hell I wouldn't be surprised if many of them exist purely with that intention.

Ah I see, thanks, there are certainly better examples though, after a few minutes I found 'unshakeable'* which google tries to correct to the American version, yet is standard British English.

* https://www.google.co.uk/search?q=unshakeable (even makes the mistake on the UK Google site).

In that case the sequence should be 1, 1, 2, 720... as zero factorial equals one.

https://www.google.com/search?q=0!

While the story at the end about the bug report is humourous, it's a pretty mean thing to do. Perhaps 3 days lost work was longer than the expected damage, but one would imagine it would waste at least a few hours of someone's time, which is not something I'd feel good about.

On an unrelated note, sometimes using a computer not configured to British English I get spelling corrections for words like "humourous" above, as a sanity check I usually Google the word, it would be really helpful if Google displayed a message that it was the British version of the word, instead of displaying: "Did you mean humorous?".

It would be better if clicking in the gaps between day boxes would choose the nearest day box, so that the mouse would not flash between arrow and hand while sliding across it (this is a pretty common UX practice, I think it has a specific name).

> Why not pass it and then never use?

Separation of powers makes it difficult in many countries to do that (legislature and judiciary being separate).

Now that I think about it this must be why iOS uses 9 minutes for the snooze interval, to avoid similar overlap (always wondered).

There's also a rise in late adolescence when risk taking and suicide are much higher.

You can put your super into a huge variety of places, or even run your own fund. Iirc there are about half a million or so such private 'funds' in Australia.

> He also completely misses the point about javascript's global-by-default variables and how easy it is to introduce them with typos

Using strict mode will avoid these kinds of mistakes.

Google fails after a few zoom outs for me, screenshot: http://i.imgur.com/Kss1x.png

One way is to set a threshold for whether or not to join the dots, if it's above the threshold perform a search between those points to see if they are between the dots or if they diverge, this search can be complicated or simple, google seems to use a fair bit of cpu power when zooming out on tan(x) so I'm guessing they're using a search here rather than some alternative 'clever' induction from the nature of the curve itself.

Well they have very different business models, neither company's licensing here seems unreasonable.

What other character would you suggest? At least multi line strings arent super common (nor essential) so extra difficulty isn't a major problem.

Precisely, I can't think of any situation where a user would search for a substring of a single word in the country name. If there is some edge-case where the first letter or so can be omitted that should be handled with the alternate-spellings feature.

> Sorting by artist seems to be broken - if I sort by artist, it should logically order each artist's albums alphabetically as well. I'm not sure what secondary ordering its actually using - it's not alphabetical by track or ordered by track number (all 1, all 2, ...). iTunes calls the manner of sorting that I suggest "album by artist".

It appears to be a stable sort, preserving the existing order of tracks that share the same album name. So to sort by 'album by artist' first sort by album then by artist. IIRC that's how it also works on windows, and is a feature I quite like.

Furthermore this occurred in 2004.

There's a plug for the 'jelly belly' brand at the end of the video where they show their logo, I'm pretty sure that means they would have either gotten them for free or they sponsored the video.

This was exactly the kind of disclaimer on the thinkgeek 'submit t-shirt idea' page, at least 6 months or so ago. I would link to the page that had it but it seems it's now behind a user login page.

Sounds like a problem with the fact that users and pages can take facebook.com/xxx urls, it's not future proof. Although the overview page resides at /about/timeline, it's understandable that they want to redirect the /timeline path there too.

Legal issues aside, I feel like payment is more likely to be received if similar yet less hostile message is provided as they still have the option of getting the site build by someone else for a similar price out of spite. (Note the 'Cotton Hill' link on the page is a competitor's site -- that's a pretty nasty thing to link to)

I'm guessing (hoping) it will use their closure compiler (or something else) to fallback to JavaScript for other browsers.

What would happen to music that is currently say 60 years old? That would have been copyright-free for 10 years and all of a sudden it is under copyright again?

I've never reviewed any journal contracts before, but wouldn't there be some clause that forbids publishing content that intentionally damages the publisher's brand? If so couldn't TechCrunch sue the author here?