Snooze timing question

[androidtopp]

How does snooze timing work? I ask this becuase with a time display that includes seconds, I finally am able to notice this potential issue (vs my previous alarm, which did not have a second display.)

Let's say my alarm goes off at 5:00:00. And I hit snooze at 5:00:10. Given a 9 minute snooze period, does the alarm go off again at 5:09:00, or 5:09:10? I ask becuase I noticed during a snooze marathon Monday morning, that after X snoozes, the time was a minute or so later than "Original Alarm Time + 9*X". So I'm wondering if it's a time issue in my network, a bug in the alarm code, or just a by design behavior that maybe all clocks follow, but that I've never noticed.

Not that it's making me late - I have the snoozing to blame for that...

[soxfan]

I don't have an answer for you, but I can tell you that I've noticed the same thing. I usually hit the snooze five or six times each morning, with snooze set for 7 minutes. By the time I actually get up the time is usually a minute or so more than 7x5 (or 6).

[androidtopp]

OK - good - maybe I'm not crazy. I know the first time the alarm goes off, it's pretty accurate - within a second or so of the clock flipping over, the alarm is sounding. So maybe there's just some natural skew in there? But even if it was a couple seconds off each time, I think I'm seeing more "off-ness" than a couple seconds x 5 or 6 snoozes would account for.

[soxfan]

I don't know, I'm usually too groggy to take notice, but I think your theory about the alarm going off 9 minutes (or 7, in my case) after you've pressed the snooze button, including the extra seconds from the time the alarm sounds to the time the button is pressed is pretty accurate. I have my alarm set to random mix so maybe sometimes the song has a few seconds of silence at the beginning, or isn't quite loud enough to wake me immediately. Twelve extra seconds x 5 presses of the snooze = an extra minute. I think that's about right in my case anyway.

Easy enough to test. I'll just have to do it when I'm a little more alert :)

[max.spicer]

How does snooze timing work? I ask this becuase with a time display that includes seconds, I finally am able to notice this potential issue (vs my previous alarm, which did not have a second display.)

Let's say my alarm goes off at 5:00:00. And I hit snooze at 5:00:10. Given a 9 minute snooze period, does the alarm go off again at 5:09:00, or 5:09:10? I ask becuase I noticed during a snooze marathon Monday morning, that after X snoozes, the time was a minute or so later than "Original Alarm Time + 9*X". So I'm wondering if it's a time issue in my network, a bug in the alarm code, or just a by design behavior that maybe all clocks follow, but that I've never noticed.

Not that it's making me late - I have the snoozing to blame for that...

It's configurable, but by default it's 9 minutes after you hit snooze. If you hit snooze at 08:23:12 and leave the settings at the default, the alarm will go off again at 08:32:12. The web ui lets you choose a snooze length in minutes between 1 and 30.

Max

PS If you really want, you could actually set it to be any time in seconds between 0 seconds and a very, very large number of seconds (years). You'd have to use the cli or some other mechanism to do this though (e.g. editing the prefs by hand).

[max.spicer]

OK - good - maybe I'm not crazy. I know the first time the alarm goes off, it's pretty accurate - within a second or so of the clock flipping over, the alarm is sounding. So maybe there's just some natural skew in there? But even if it was a couple seconds off each time, I think I'm seeing more "off-ness" than a couple seconds x 5 or 6 snoozes would account for.

Interesting. It never occurred to me that snooze should do anything other than add x minutes to the time at which you hit the snooze button. Are you saying that if you hit the snooze button it should just add x minutes to the original alarm time? So if the alarm goes off at 07:30:00 and you hit snooze at 07:33:42, the snooze should stop at 07:39:00 (assuming a 9 minute snooze)? Is this how most alarms work?

Max

[dwilliams01]

I've always like the ones where you could be mid-snooze and hit the snooze again to bump it another 9 minutes...

It's really important to be able to hit the snooze about 1 minute before it goes off again to get another 9 minutes started as well.

As for the seconds drift, I guess that I've never had an alarm clock with a seconds display to really notice. Certainly didn't have anything that I would normally let go off for more than a couple of seconds (beeping kinds of alarms).

[soxfan]

Interesting. It never occurred to me that snooze should do anything other than add x minutes to the time at which you hit the snooze button. Are you saying that if you hit the snooze button it should just add x minutes to the original alarm time? So if the alarm goes off at 07:30:00 and you hit snooze at 07:33:42, the snooze should stop at 07:39:00 (assuming a 9 minute snooze)? Is this how most alarms work?

Max
I think it's more of a seconds rounding thing. I think if you had the alarm set for 7:30, with a 9 minute snooze time, whether you hit the snooze at 7:30:05 or 7:30:45 the next alarm would go off at 7:39. In your example, if you waited until 7:33:42 to hit the snooze the next alarm would go off at 7:42:00. I don't think the Boom way, going to the exact second, is wrong; just a little different than a lot of us are used to.

[max.spicer]

I think it's more of a seconds rounding thing. I think if you had the alarm set for 7:30, with a 9 minute snooze time, whether you hit the snooze at 7:30:05 or 7:30:45 the next alarm would go off at 7:39. In your example, if you waited until 7:33:42 to hit the snooze the next alarm would go off at 7:42:00. I don't think the Boom way, going to the exact second, is wrong; just a little different than a lot of us are used to.

It doesn't do that. It sets a snooze timer for the time now (including seconds) + 9*60. If it's doing anything else, that's a bug.

Max

[androidtopp]

Well, this morning, incredibly enough, I was alert enough to do a little testing.

My alarm went off at 4:42:00 (don't ask), and I snoozed at 4:42:10. The next aalrm went off at 4:51:10 - so exactly nine minutes after I depressed the snooze button. I snoozed again at 4:51:25, and got another wakeup at 5:00:25. So Max's explanation is exactly what I observed this morning.

I'm not *complaining,* just trying to work it out in my head. Since I've never had an alarm clock that so clearly displays the seconds, I've never taken notice until just now.

And to add to dwilliams point - this is the first time I personally have used an alarm where I didn't frantically try to make the thing shut up - I actually like the song I'm waking up to. So I'm inclined to let it go for a few more seconds, and that's probably where I'm noticing things drifting away from the even minute marks.

Being able to snooze in mid-snooze might be nice...but for me it'd be dangerous. Who knows how many button presses get registered when I'm fumbling in the dark...I'd snooze all the way past my express train.

[soxfan]

It doesn't do that. It sets a snooze timer for the time now (including seconds) + 9*60. If it's doing anything else, that's a bug.

Max
I guess I didn't word my response properly. I wasn't saying the Boom did this. I was saying that I think most traditional alarm clocks only work at the minute level. So, whether you press the snooze at 7:30:05 or 7:30:30 or 7:30:50 the next alarm will go off at 7:39:00 (sticking with the 9 minute snooze time). As I suspected, and you've confirmed, the Boom goes right to the second level. And I agree with androidtopp; I'm not complaining. It is just different than every other alarm clock I've ever had.

[JJZolx]

Interesting. It never occurred to me that snooze should do anything other than add x minutes to the time at which you hit the snooze button. Are you saying that if you hit the snooze button it should just add x minutes to the original alarm time? So if the alarm goes off at 07:30:00 and you hit snooze at 07:33:42, the snooze should stop at 07:39:00 (assuming a 9 minute snooze)? Is this how most alarms work?

I think it is.

Also, I'm pretty sure that most will go off only at the moment when the minute changes.

I wonder if the use of timers are the reason for so many apparent problems with alarms. If you don't use a timer for the snooze function, then the whole thing is simplified to executing a singe task every minute that asks "Are there any alarms scheduled to go off now?".

[dwilliams01]

Being able to snooze in mid-snooze might be nice...but for me it'd be dangerous. Who knows how many button presses get registered when I'm fumbling in the dark...I'd snooze all the way past my express train.

I wasn't meaning that it should add 9 minutes every time you hit it, just start counting from that point. I seem to always open my eyes a minute before the thing goes off again so I just want to hit it then and have it start over at that point instead of having to wait.

[androidtopp]

OK, gotcha. That's much less dangerous! ;-) As someone with a ridiculous body clock that wakes him up about 2 minutes before the alarm, this would be kinda handy.

[max.spicer]

I wonder if the use of timers are the reason for so many apparent problems with alarms. If you don't use a timer for the snooze function, then the whole thing is simplified to executing a singe task every minute that asks "Are there any alarms scheduled to go off now?".

This isn't a problem with timers. It's simply working as its designed to do.

Max

[androidtopp]

I think JJZolx is referring to people who have their alarms go off at either a) not the correct times (meaning, middle of night, not just "off" by a few seconds like we're discussing here) or b) not at all.

Now, I've not had a bit of trouble with my Boom's alarm functionality - it's been flawless. So I can't complain personally. But it does seem like a small group of folks are seeing erratic behavior. I think it's my super sweet 4 year old server that makes everything run so smoothly. ;-)