[time-nuts] Needed: The Real Serial USB Fix

[Bob Camp]

Hi

On Aug 24, 2013, at 3:27 AM, Hal Murray <hmurray at megapathdsl.net> wrote:

> 
>> If it were me, I remove the LCD display.  I don't see a need when everyone
>> today has a phone.  the little AVR chip can bit-bang a UDP packet once per
>> second and put it on your wifi router.  That cuts the parts cost by 1/3rd
>> and the display will be where you can see it, on the phone or computer.
> 
> There are still a few of us who don't have cell phones.  :(
> 
> I split the problem into two areas:
>  Collecting the data
>  Displaying it
> 
> If I'm collecting data, I want to be able to look back over hours, days, 
> months, or even years.  That means I probably want to backup/archive it too.
> 
> 100 characters of text every 10 seconds is under a megabyte per day.  That's 
> small relative to modern disks, medium relative to thumb drives or SD cards 
> and small to medium relative to RAM (or ramdisk).  [I generally chop things 
> up into a file per day.]

That's a *lot* more than a simple monitor. Once you get past "simple" you already have LH and a lot of hardware that will run it just fine.

Bob

> 
> So far, I've always had a handy Linux box running 24x7 when I wanted to 
> collect data, so I've never tried to use a smaller system.  Since the data is 
> collected on a real system, it's easy to display it any way I like.
> 
> I won't be surprised if a project comes along where I want something like a 
> small uP to grab careful timing data.  Until then, I'll keep collecting my 
> data on a real PC.
> 
> ------------
> 
> Does anybody have data on lifetime of Thumb/SD cards if you flush a log file 
> every 10 seconds?
> 
> What type of file systems are supported with whatever OS comes with small 
> uPs?  Is there any work on flash-friendly file systems for append-only log 
> files?
> 
> A year or 3 ago, I did some work on logging to ramdisk and occasionally 
> copying to hard disk.  The idea was that the hard disk would be spun-down 
> most of the time, saving a lot of power.  It didn't save much so I bailed on 
> that project.  I assume that means the power to keep the disk spinning is low 
> relative to the power to keep the electronics going.  Maybe I just botched 
> the experiment.
> 
> 
> flushing the file after each line is the simple way to make (mostly) sure 
> that your data gets to disk in case the system crashes, but it turns into 
> several disk writes each "line".  That's no big deal for a lightly loaded 
> hard drive but gets interesting in terms of total writes to flash drives.
> 
> You could, of course, fix the code to only do the flush once every N minutes, 
> or only when something interesting happens...
> 
> 
> -- 
> These are my opinions.  I hate spam.
> 
> 
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.