Local Time Zone Doesn't Affect Start of New Day

[youda]

The system counts a new day as starting three hours after midnight, as this is a time most people are likely to have finished their day's studying.

I'm in UTC-7, and the new day seems to occur at 4:00 PM my local time.

Using browser developer tools, I did notice a POST to https://www.maorma.net/Practice.aspx/UpdateTimeZone with my time zone, so I assume the intention is to account for local time zone.

Additionally, if I happen to be practicing through the start of a new day, the previous day's practice time is carried over into the new day, essentially double-counting the time from the previous day.

None of this really matters besides making my statistics inaccurate, but I thought I'd mention it here in case it was just some typo in whatever is handling the UpdateTimeZone URL.

Of course, thanks for creating and continuing to run this site. It's been part of my daily routine for over a year now.

[Programmer]

Hi Youda, thanks for the feedback. Yes, that definitely sounds like a bug. My reasoning for the three a.m. thing was that there might well be people burning the midnight oil and studying past midnight, and it would be confusing if the statistics suddenly got reset while they were still at it. I settled on 3.00 am as a time when practically everyone is going to be asleep. The night owls have probably all gone to bed by then and the earliest of the early risers still haven't got up. But obviously something is going wrong there, and as you say it is not just creating confusion but also affecting the statistics. I will have to take a look and see if I can find the problem.

While we are on the subject of the statistics, I am aware that the statistic module really isn't very helpful at the moment and I am working on an overhaul of it. The reason I say that is that it tracks "the combinations you find hardest", but really the website is just tracking for example how many times you get the combination "34" wrong. That doesn't actually contain that much information because it is relatively easy to differentiate between "34" and "44", but very much harder to differentiate between "34" and "24" (at least for me). So my idea is to change it so that the statistics really track the combinations of combinations, i.e. each tone pair combination compared with every other tone pair combination. This kind of change has to happen in two stages. I have done the first part, which is that for the last six week or so I have been logging both pieces of data in one table, i.e. the combination the user was asked and also the combination they thought they heard. Hopefully in a month or so I will be able to do the second part which is to change the way the statistics are presented to show this. But there is no point in doing that until I have collected enough data about my own mistakes to have something to work with.

Returning to the point though, I will take a look at the time zone thing when I can and post back on this thread if I come up with a potential fix.

[youda]

Thanks for your reply.

Yeah, 3:00 am seems too early to be waking up, and too late to be still awake. I can't imagine many people will be studying through that time.

Not that anyone would, but if someone practiced 25 hours straight beginning from the start of the day, they would see 24 hours in that day, and 25 hours in the next day, despite really only having practiced just 1 hour into the next day. If they practiced again later in that day they could have over 30 hours for that day. Actually, if each day they only practiced a few minutes during the day change, they could keep carrying over the previous day's total into the new day, despite having only spent minutes practicing. The new day changing at 4:00 pm my time wouldn't be much of a problem for me if the practice time wasn't double counted when I practice during the day changing.

Oh, great! I was considering making a feature request for something along those lines, so that I could focus on just the "combinations of combinations" that I have particular trouble with. Thank you for clarifying that the table doesn't currently take into account what the combination was confused with. I had never really been sure what was going on, I was mostly just tracking my practice time.

[Programmer]

I think I have been able to figure out what was going wrong, and I think I have been able to fix it. I can't really say that 100% though because my own time zone is very close to UTC which makes it a bit difficult to tell. It is also possible that I missed somewhere the fix should have been applied, as the problem I identified was in more than one place, so if you see anything date/time related that still doesn't make sense please let me know!