OK - using a downstream server/PC effectively as a backup for the logged data? But I still don’t really see the distinction in principle between eg NSLU2 and a SCC running eg WD. They’re both PCs that can be programmed to perform specific tasks. Granted, at present, the NSLU2 pretty much has to upload its data to another computer for further processing and so I suppose intrinsically performs some data backup by default almost, whereas this isn’t a standard part of how WD works (is it?). But it would be a fairly minor feature addition for WD to be able to upload some sort of data backup to a web server at regular intervals and so yield the same end result.
One other attraction of the NSLU2 approach that has been mentioned here in the past is that it provides fanless operation, whereas the current generation of SCCs still do use a fan, albeit typically a small and quiet one. I’m sure that next year we’ll see SCCs that are passively-cooled, but it is true for now that the standard SCC design is not fanless. (Actually I suspect that some VIA Nano-based SCCs will be fanless very soon but the VIA-based SCCs always tend to be significantly more expensive than the Intel ones and so aren’t as attractive. Of course there have been fanless VIA-based SCCs for some years now but these have always been relatively low-performance PCs, possibly unsuited to running WD, whereas the new Nano is a higher-performance part.)
I suppose that the other general aspect of this discussion is the extent to which people might happily accept more processing being done at the server level, ie by the ‘cloud’ as in GoogleApps etc. If this were to become more widely accepted and implemented for weather data then the role of the local logger/PC might be considerably reduced and all that would be necessary at the local level is for a device to upload data in appropriately summarised text form to a computer in the cloud. So in this scenario a suitably-programmed NSLU2 might indeed be all that’s needed.
Err, not quite what I was thinking.
I thought that you were pondering something like “why might someone see this as a useful/desirable device?”.
I was simply suggesting that one way in which non-Davis users might see it as desirable would be for some of the very same reasons that you mention in on your site on pages about the Davis datalogger, such as: http://www.weatherstations.co.uk/datalogger.htm
That clarification aside, I don’t think I have any particular disagreements or contentions with your other points and musings. It will be interesting to see how things work out…
I think the main problem hitherto has been that fanless motherboards have tended to be fairly low performance or that they’re only marginally fanless IYSWIM - ie that they run pretty hot and are best used with a case fan and so still do not yield fanless PCs. (I really don’t know why VIA haven’t made a small barebones PC with one of their fanless boards inside and with the CPU heatsink heatpiped - to coin a verb - or whatever to the case. It would seem such an obvious way of expanding their market.)
And isn’t the standard approach for fanless PSUs to have something like a picoPSU eg:
in the case plus a fanless 12v brick in the mains lead? I think the picoPSU is very efficient, but not so sure about the availability of efficient 12v bricks
I looked for, but haven’t found, efficiency numbers for those, but I doubt that a 120 watt psu is going to be very efficient running at 10% to 20% rated OP
An addendum to this topic of low-power PCs: There’s a detailed new review of the Eee Box PC at:
Well worth a read by anyone potentially interested in using a small, low-power (~20W), dedicated PC for creating weather web pages. Release date for the UK is nominally given as 21st Sept, but AIUI the unit is already shipping in the US/Canada and Far East.
