WEEK THOUGHTS
I haven't been finding much time for phlog posts this month. Partly
it's just that the cold winder weather, relenting a bit now, has
put me off spending long at the desk where this old computer lives
(somehow I can't get into writing these posts on a laptop). Plus
there's my usual issue of too many computer-based projects which
means I want to use my remaining time in less computational
company. I'm really not doing very well at finishing those things
either. So, while the washing's going, here's another one of my
condensed weekly thoughts run downs.
DOCUMENTATION
The Raspberry Pi Pico 2 was released recently and its datasheet is
reportedly 1347 pages, up from 644 pages with the first Pico. This
length isn't that unusual, and hard to compare fairly since many
datasheets for microcontrollers/SoCs point at lots of other
documents for details on specifics. But the trouble with this is
that it's barely practical to even skim over the contents. Contents
pages and indexes are often misleading, probably because the length
makes them too hard to compile as well, what I think everyone does
is just read a few likely bits and word-search for things they
don't see. That's fine if it works, but when it doesn't you're left
with wondering which of the 993 pages that you didn't look at has
the info you missed, and how do you find it without spending days
actually reading the thing properly?
It's the same with standards. The Pico datasheet is now between the
USB 2.0 standard and the current Bluetooth standard in length. All
the standards for the Web now are endless. If you see someone
mention that in a forum when they're trying to develop something
then the typical response is "well of course you don't have to read
all of it, just look at the bits on x, y, and z, starting on page
1825". But then in that it refers to something on page 1053, and
you misinterpret that because you didn't read page 671, and you
don't know about the simpler alternative system on page 284 which
does all you need anyway. Oh, except in the example on page 2934
it's noted how that system is never used in practice anymore.
Also since the writers then attempt to avoid all that confusion by
summarising things, you end up with the document growing
exponentially as each section begins with a summary of info in
previous sections in case you didn't bother to read them. The only
limit on this used to be the physical cost of printing these
documents out, but since computers have eliminated that restriction
it's able to spiral completely out of control. I predict a future
where engineers all just ask AI what the standards say because
nobody would live long enough to read them all, and entire
hardware/software infrastructures are built wrongly based on the
AI's misinterpretations, then nobody can make it work anymore when
the AI is changed and starts interpreting correctly/differently. In
fact that's probably already beginning.
COMPUTER POWER
The power consumption of the cloud has become a concern even
outside of technology circles. I'm not sure how this can be so
singled out though. It may be a source of growth in energy
consumption, but theoretically physical resource efficiency should
be one of the few real features cloud computing has going for it.
People share hardware instead of each having their own. That's why
it's cheap, no? If the VPS I use for my website is using tons of
electricity, then the $1/month I'm paying for it must be making
them a big loss after counting their electricity costs. Same with
the disks storing its automated backups for free on Oracle Cloud.
OK so it all adds up, and it's true that speculation about AI has
funded a new surge of expansion, but surely the things that
actually cost all of us money are going to be bigger net energy
consumers than these cheap/free computer services? They can't
really be the place to start saving the planet. More likely they're
just something conveniently out of sight for individuals to point a
finger at while still buying a new smartphones, cars, and overseas
holidays every few years.
TONER SCAVENGING
I've been 'collecting' used toner carts, which I generate a fair
few of through my business activities, for years. They have
potentially useful things in them which I hate to throw away, not
least the remaining toner inside since you can never use them up
completely. A couple of weeks ago my power was off for the day so I
finally pulled them all apart on my verandah. I tried not to make a
mess, but it didn't work, and the rain hasn't really washed it away
since, oh well. Pics of the result are here:
gopher://aussies.space/1/~freet/photos/teardowns/toner_carts/
It took me one or two of each type of cart before I got the method
figured out - the easy way often isn't obvious. My 3D printed COVID
mask came back into service to stop me breathing in the toner that
wafts out. Last time I tried to do it indoors, which whas a mistake
- better to have a light breeze to blow it away. I think I got
about enough toner to refill three of the smaller toner carts that
I use, and which are easier to refil than the awkward big HP ones
like the one shown in the first pic (I tried refilling one of those
once and failed miserably/messily).
The gears are cool, the metal rods/tubes should be useful, the
scrws maybe handy, the other metal bits probably not so much but
easy to store, the plastic shells... well they're a problem. While
the toner cartridges were intact I could drop them off at a
cartridge recycling bin. What do they actually reuse? I don't know,
but _maybe_ they reuse the plastic. Now they're apart though, can I
recycle them? The council website says they accept "rigid plastics"
at one tip location (about 40min drive away, and shut at most times
convenient to me), but I think they just mean HDPE, PET, etc.
plastic bottles. These don't have any recycling code on them, so I
don't know the plastic. They're obviously a thermoplastic so they
_could_ be recycled, but it looks like I'd have to put them in the
rubbish. Or I could resurrect my unfinished 3D printing filament
plastic extruder project, but I never really ended up doing enough
3D printing to justify that, and there's too much big competition
now to make a business from it (yes, this was another one of my
aborted business plans, back when home 3D printing was new).
So, environmentally, was this good recycling or bad recycling?
Maybe if I'd dumped them all in a cartridge recycling bin the
plastic would be reused, but I wouldn't have been able to use the
toner and other parts. Maybe they would have just got out the toner
and sent the rest to the tip anyway? Fact is that the toner saves
me $60 in new carts (somehow I've never been able to find a
supplier selling toner for refills cheaper than new toner carts
themselves), plus the extra bits, so I figure the recycling method
that saves me money comes first. It still doesn't quite feel right
though.
- The Free Thinker
