I find that leaves a slightly bad taste in my mouth. I wish it hadn't been done, but I guess I can understand why.
In the discussion a link was provided to a non-scribd version, so here it is. It's a 1.8MB ZIP of a PDF.
The idea is that it counts backwards in sleep cycles from the time your alarm has to go off so that you wake up during a light phase of sleep. Waking up in light sleep (vs deep sleep) will let you wake up feeling more awake, alert and refreshed.
If you're interested, the URL is http://sleepyti.me
I'd love to hear any comments or suggestions about the page, so if you have any feel free to let me know!
There's a watch on the market that monitors your movements in order to spot when you're lightly sleeping. You can set an alarm, and a window of time before that alarm. If, during that window of time, the watch detects you moving, it'll go off then. Otherwise, it'll go off at the specified time even if you're deep asleep.
The idea with http://sleepyti.me is basically to emulate this without the need for external hardware. I'm in the same boat as you--that is, I have trouble actually falling asleep sometimes.
I originally had 'time to fall asleep' option that would allow the user to specify how long it usually took to fall asleep. Unfortunately, no one really ever used it. It's a lot easier to say "the average human takes fourteen minutes to fall asleep," than to try to calculate that in beforehand.
My advice is to give yourself a half hour or so to fall asleep. If it's been longer than that and you're still awake, get up, walk around, go back to bed in 15 minutes (and recalculate from sleepyti.me!). Research says that laying in bed longer than a half hour while awake will actually make insomnia worse. If you're sitting in bed for over a half an hour every night, you might consider talking to a doctor about sleep aids. It helps me sleep better just to know that I could take an Ambien if I need to.
Although I was originally going to try to compensate for this variation, I instead decided to go with the average sleep cycle times instead. There is always going to be some variance in the length of cycles (usually between about 75 minutes and about 110 minutes), but the average is pretty well documented to be about an hour and a half.
Although it's definitely true that some people will have drastically differently timed sleep cycles, this is usually due to some other neurological condition. My theory is that short of that, the law of averages will help me out :)
So perhaps I should go to sleep at 21:30. Or at 23:30.
There's a big difference.
I do love the idea - the challenge it trying to work out whether I'm average.
It's too late to delete or edit it, so I'll have to let it stand, but I'm sorry for replying in anything other than an explicitly positive and constructive tone.
I expect this by end of day, please.
These two threads basically show what I have always said... PDFs are wrong for the web (and any 'tool' that displays them).
Ideally a site that eventually replaces HN, or an update to HN, will have a single location where similar submissions can congregate, and then we get a single discussion thread. In this case we could then have put the direct links as alternative sources, and retained a single discussion location.
But we don't have that, so I provided an alternative source that was close to the surface, instead of buried in a thread.
Which they seem to be doing :)
And what makes your doctor the authority on sleep? They inevitably get their information from scientific studies not conducted by themselves, and the author of an ebook can do the exact same thing.
There's no reason to disparage the ebook just because it doesn't look like the conventional form of authority.
Then it should cite the sources.
A few don't lead anywhere, guess that's a symptom of linking out of a two year old ebook?
Go to sleep early, eat right, don't eat weird shit at night, and you'll feel better. Witchcraft, I say.