Moved!

to hyperdata.it

Not Climbing : Mattress Rescue

we have a problem

Now lockdown is easing in the UK my mother has starting making noises about coming over. She usually visits at least once a year, but her health’s not been great the past couple of years and, ermm, pandemic. I mentioned this to Marinella this morning, she asked, “where’s she going to sleep?”. Well, my bed, I’ll take the fold-down, as usual. I do have a spare bed on the soppalco (mezzanine above music room) and the fold-down couch in the kitchen, but my bed is the comfiest, and has a bathroom next to it, major bonus in my mother’s book.

Except it hasn’t been very comfortable of late. Ancient mattress, and when Claudiopup was a bit younger he enjoyed tearing it up. Bare springs in places, which I’ve worked around by piling on folded blankets in appropriate places. Has got a bit medieval.

Over in No.7, my previous house just across the lane (empty, awaiting letting or sale), there’s a perfectly good mattress – a futon I was given years ago, no doubt better for the spine. I’ve been putting off bringing it over because it seemed so daunting. Double mattresses are fairly heavy things, total nightmare to shift because of shape & bendiness and both houses have narrow, winding stairways. Two strong people needed, even then there’d be a lot of cursing. I’m not strong. Could ask neighbours who’d be happy to oblige, but don’t like asking…

But, in recent months I’ve been accumulating climbing gear. For other jobs around the place – fix roof; trim trees. I do fancy trying climbing for recreation again (after occasional attempts 30+ years ago), but that’s a just corollary of having spent $$$ on climbing gear, get the most out of it.

This mattress replacement, surely I, with the help of climbing gear on Newtonian physics, can do it without breaking my back. Give me a lever long enough…

[[Tyrolean traverse did cross my mind, very elegant solution, but most of this house would be in the way, and frankly, it’d be suicidal…]]

Step one : hugely encouraging.

I’ve watched lots of videos recently featuring not only people that fix roofs and rock climbers but also those of arborists (Rivendell with chainsaws). And high-building window cleaners. And the Americans that hitch themselves up in trees on a weird ladder setup to shoot things, presumably at half-term when the schools are empty. A brilliant bit of kit across them is a lanyard (longe, cordino). Shortish, extendable line, clip yourself in for backup safety/support.

The number of times I’ve been around climbers on crags in the Peak, had to look away. They’re hopping around unroped right on the edge of a potentially fatal drop. When so easy to be safe.

So I got a 3m/10′ bit of dynamic rope, a double-lock Via Ferrata-style carabiner and a Kong Slyde, a minimal little plate that allows the length to be adjusted easily (funnily enough Kong mention their use as energy dissipators, but I for one wouldn’t trust one for a Via Ferrata).

So versatile. Slapped on a pulley, hoist the mattress into a roll, lock it with (120cm) slings held by quickdraws.

At this point, 90% of the hassle of shifting a mattress has gone.

Next, set up anchors for my speedline (hah, terminology).

I put a bolt into my bedroom window ledge and attached a knotted rope some years ago. I guess around the time there were the nasty earthquakes around L’Aquila, not all that far away, bit of (rational) paranoia. That end sorted. Sling around a sturdy bush at the other side of the yard. Rope between, this one a proper dynamic climbing rope. Frictions knots and pulley to tension. Very ad hoc, but worked.

BIG POINTS IF YOU CAN NAME THE SIX KNOTS (one used twice). I only knew one of these a few weeks ago. Thank you weird YouTube people.

Here we go…

I’ve already been surprised by how stretchy dynamic rope is, but that was with my weight on it. Mattress must weigh a lot less, but still, boing! This photo is staged – I pulled in more slack.

Note the blue lanyard used again, this time to stop it swooping down to far (it did anyway).

I took it down, put some crappy nylon rope around with bowlines to keep the roll and bunged it in the cantina. Either for the end of the month to put by bins when big things are picked up or to use as a bloody awful bouldering mat.

Next, new mattress. Same tying procedure. Even though it looked like heavier material, I guess without all the iron of archaic European mattresses it was much more manageable.

Here I made A STUPID BEGINNER’S MISTAKE. For an anchor I’d looped the rope around the bricks & mortar divider between the windows, I think with a hitch so it’d dangle a bit and balance. I used a locking carabiner. DIDN’T LOCK THE CARABINER.

Early on it crossed my mind a lot of this looked like mountain rescue stuff. Where I grew up in Derbyshire, that, and cave rescue were things. Similar terrain & issues here, but about 5km from here the Protezione Civile guys have a helicopter. Doubt they’d take a callout for a stuck mattress.

Right here you have a very bad situation. On loading, the rope etc. got messed up, jamming the anchor carabiner open and the rope stuck. Freeing the rope might have caused the casualty to fall to their doom. In this case, no human life involved, but it was lucky a lovely old planter at the top of the steps didn’t get knocked off.

I was able to hook the lanyard just below the bad carabiner, yank it from above and it clicked back properly (lanyards ftw!).

Taking it to it’s new home, hooked back onto the rope there.

Retensioned. Attached the other rope from bedroom, naively thinking haul it up.

Ok, not that naive. Did occur to me at this point that it really should be able to clear the window ledge. So stuck a couple of screw eyes into the beam in the bedroom, rigged anchor to haul it up higher.

Hmm. No clearance whatsoever.

I swore twice. Once with this realization, second time on catching myself on some dangling thorns. I tried putting another screw eye into a rafter just above, outside the window. But the wood started splitting, thought better.

Now properly stuck, in the abstract sense. Spent ages just trying to haul, (just using lanyard & the remaining carabiners with friction knots on cords).

Last night I heard mention on the radio of whalers, apparently Arthur Conan Doyle worked whaling ships to make a bob or two before doing Sherlock Holmes. On my mind, got a steel fence stake and trying hauling him aboard. No joy.

Gave up. Let Moby Dick go free. Lowered down, unclipped.

I guess because I had been pulling a lot, it didn’t know feel too heavy. Pulled it over my back, took it upstairs without any trouble, Igor stance.

Job done.

This took hours. But apart from getting the job done, was really good for learning. I need to be totally familiar with all this stuff before taking on the tasks where there will be genuine risk (however small). I came close to damaging plant life this time, def don’t want to damage mine. Some lack of foresight – making it up as I went along didn’t entirely work (though I never went to a position I couldn’t back out of). Outrageously stupid mistake of not screwing up a key carabiner. Also, you can never have too many carabiners.

Hope this thing is comfortable.

No Such Thing as a Coincidence

I’m only now making a start at preparing the orto (veg plot) for planting. My neighbour Achille did his annual favour of digging it over with his little excavator a good few weeks ago. It would have been better if I could have started say 3 weeks sooner, but the weather has been almost continuously wet until yesterday. Suddenly, hot & sunny! So yesterday I made a start on clearing the brambles around the edges.

Orto, human lawn, house – cherry tree on right

For reasons I won’t go into here, just now I wanted to measure out the main area. My workshop metal tape measure only goes to 5m but I remembered an antique surveying kind of measure my dad gave me years ago. Imperial units of course, but one side was mostly blank, so I marked metres on there.

I was pleasantly surprised that the main area was almost exactly 10 x 5m (2 x 1 poles).

Convenient units

So I put in a couple of strings to mark it out, guide me for clearing, hoeing & raking.

As I was doing this, looking around, the tidyness of the figures impressed me more. Coming from the house there’s the human lawn (dog lawn is in the yard) which demarcates one side. Looking that way, the neighbour’s field is a terrace on the right, a wire fence at the bottom of the rise on my side. Another fixed side.

But looking along, beyond is the tiny field that’s also part of this property.

Little field is up to the right

So the line there is totally arbitrary. But I wouldn’t have wanted to go much further, following the neightbour’s side there’s a big elder bush (well, remains of – Achille took his digger to it a bit while clearing) then a row of unmarked pet graves.

On the other side, there’s an area designated as being a chicken run, as and when I get around to populating the shed. Right now it’s overgrown with brambles and a cherry tree has sprung up, but essentially that side of the orto could have gone anywhere. Further along, plenty of space between the current orto and a little access track then a village lane.

House on the left is No.7, where I lived before here

So how come almost exactly 10 x 5m? Finally it clicked. Because I’d decided it should be 10 x 5 metres. I believe I initially put the fences up long before moving in here, it wasn’t an orto before. Perhaps 5 x 8 metres. I extended the length 3 or 4 years ago, obviously choosing a round number. D’oh!

Convenient though.

Line of Doody

Short post to mention the ‘bad apple’ narrative around the police.

Setting the scene : two very significant precursors of modern police were the cops employed by West Indies merchants on the Thames and slave patrols in Carolina.

There’s a dichotomy within the police between their job as crime-preventors, protectors of the public, and their role as the domestic strong arm of the state and associated (many financial) institutions.

Thing is, the typical media narrative is that of the bad apple. Even if the corruption in a drama like Line of Duty goes right to the top, it’s placed as something exceptional.

The well-documented ‘Firm within a Firm’ of the Met. CID in the 1970’s, criminals effectively in the employment of coppers of all ranks, is a classic case of widespread corruption. Win-win for those involved, so such things are presumably fairly common. But I’d suggest this narrative fails to mention the more sinister purpose of the police, as the government’s militia. A delightful phrase I found on Wikipedia is ‘the monopoly on violence’. Think about it.

I personally experienced bad apples in 1980’s Sheffield (luckily when it came to the cop I was alleged to have assaulted to give his evidence, he was suspended from duty for an unrelated assault incident so I wasn’t locked up for trying to defend a victim of police assault…).
Same period, Thatcher & co. mobilised the cops as their private army to harry the North, shut down the miner’s strikes (and by extension the power of their unions).

I get on well with the (armed!) cops that live near me now. Who knows what it would be like without them, people are generally honest and non-violent, but I’m prepared to accept life is better because they are here. Is rural, which makes a difference. But police proper is about the city (or City), check the etymology. Their main job is historically to maintain the status quo for the aristocracy and the merchants.

What I’m suggesting is that the things of concern aren’t only bad individuals, bad culture (like the institutional racism in the UK police). Looking at those issues, you’re distracted from the more fundamental fact that they are a quasi-militia that serves the state, wherever it prefers to go.

Incidentally –
‘chauvinism’ : named for Nicolas Chauvin, a legendary and excessively patriotic soldier.

Not Climbing

For a few years now I’ve had a leaky roof. It has to rain a lot to be an immediate issue (puddle on bathroom floor) but this last winter there were a couple of days with torrential rain, the ceiling is starting to look dodgy. I’ve put it off and put it off, because the only options seemed to have been to pay a company to erect scaffolding, at a cost of €1000s, or to ask a local odd job man to kill himself for about €8 an hour.

Thing is, it’s not, on the face of it, a very big deal at all. The roof here is done in the traditional local style, terracotta tiles held on by gravity alone (aided by a few rocks around the perimeter). The problem area is only a few metres away from easy ladder access.

But…on one side it’s a 5m drop onto a yard paved with flagstones, the other 6m onto a concreted track.

Hmm…

When we got this place there was a lot of work needed doing (it was basically a shell), but we didn’t need to move in immediately. So despite having only minimal skill at woodwork I decided that I’d tackle the bits that seemed manageable myself (stairs, windows…). Spend the money that would have gone to proper craftsmen on equipping a little workshop.

That worked out remarkable well. The results would not win prizes for aesthetics, but I got (mostly) functional bits in place. (Mostly, the window frames are far from perfect, bit gappy, but still adequate). This is an old house in a rural hamlet, rustic is acceptable.

So there’s my Plan C for fixing the roof. Get the necessary gear, do it myself.

I believe I have zero experience of roofing. But years ago I did wind up spending a lot of time in the company of expert climbers in a shared house in Sheffield (entertainingly documented in a short British Mountaineering Council documentary video, Statement of Youth: Hunter House Road and the Dole). I did have a go at climbing myself a few times around that period, only very easy stuff, but enough to get an idea of how it was done.

I am scared of heights, but only much bigger ones than this (that weird compulsion to jump off terrifies me). In this case, I’m scared of hurting myself.

Gear then. Wait, first, the problem…

Biggest problem is, as you can see in the photos above, virtually nothing on top that could serve as anchor points. A very old, crumbling chimney stack, that I doubt a gecko would trust.

Measuring up (taking photos including a tape measure, using Gimp editor to measure other bits) –

I reckon if I put a line over the middle of the section, bolts in the wall lane side, to be decided yard side, that’ll give me an anchor point at the apex with a fair bit of clearance if I was to take a tumble.

Thing is though, this is Not Climbing. There, the typical task for a rope would be to catch you if you fell. Here, main thing is not to fall.

Now gear…

Climbing gear is so sexy. I’m tempted to get loads just to hang on the walls as ornaments, sit on my arse.

Budget is limited, so I want the minimum that will make me feel safe. Proper helmet, natch – that’s a multi-justifiable buy, I’ll be getting the bicycle out of the cantina very soon, seems daft not to wear one.

One the right, a dirt-cheap rope advertised, very irresponsibly, as being for climbing. The aforementioned gecko wouldn’t go fishing with that. The carabiners are truly awful – might be handy to use with dog leads, maybe, not even sure about that. But the rope itself is plenty strong enough to take my weight without a jerk. Will do for the High Line (not that kind). This is Not Climbing.

As a realistic concession to safety, the rope on the left is a proper UIAA-certified dynamic rope. As long as I have a reasonable amount of redundancy in the anchors, their individual strength doesn’t matter too much. Like, I will put a loop on the dodgy chimney, and bolts at the start. Effectively walk-abseil across the roof, Figure-of-8 belay device backed up by at least one Prusik.

I opted for one of the Petzl Figure-of-8s. The ‘rescue’ type, with the horns, I would have preferred, but I want that to be solid beyond question and the reputable brand ones are pricey. Nor do I want to wind up locked on a trad figure-of-8, needing to be rescued. The square-looped Petzl seemed a good compromise. (I will carry an extra loop or two for Prusiks, just in case).

There’s no rush to do this, any time before autumn rain will be be fine. There are some little crags over the back where I can practice abseiling, maybe find an easy route or two. Try and build a bit of confidence. (Also hopefully strength – I’m ultra-weedy at the mo).

It is a recurring problem, so if I have to I’ll get more gear. Not sure I’ll be thinking about this other tile slippage this year though.

PS. Oh yeah, harness. My darling Marinella has her old hang gliding harness on offer. I’ll have to have a look, attaching from the back seems a little kinky. Can get good quality climbing ones <€50, which is still within my woolly budget. Tempting, but for something I may only use 2 or 3 times (hmm, maybe annually thereafter…), hard to justify. This afternoon I had a play (following YouTube instructions) on doing a ‘Swiss Seat’ harness. The video seems to suggest serious genital disfigurement, but it did actually feel quite comfortable, definitely secure. It’s also very easy to make a chest harness with a sling (YouTube again), combined with that Seat I reckon that’ll be fine for occasional use. So have ordered a couple of decent slings, couple more crabs.

PPS. My old mucker Stephen (who tends to know about things like roofs) commented asking whether I could check the roof wood (timber, er…he uses other words I don’t know) from inside. The main beam is open, poked with screwdriver, rock hard. Same beam continues above my bed. I slept there for 5? years before noticing it has a date carved in it, I think says “A 1798 D”.

Demonic Electronic Firefly (3v Flashing Joule Thief)

A few weeks ago I gave my girlfriend a Kinder Egg. She gave me back the (turtle!) toy container, suggesting I might have a might have a use for it. Of course I had: lucciola elettronica!

Aside for having the Best Name Ever, the Joule Thief is a remarkable little circuit:

[Pic from Wikipedia by Rowland, CC BY-SA 3.0]

It’s a switching boost converter that can run from a very flat 1.5v battery and ‘constantly’ light an LED at a very low average current by delivering a chain of very brief pulses. Demonstrates how close to magic even the apparently simplest electronics can be.

A video by Eman2000 had been recommended to me for a Blinking Joule Thief. Essentially the same circuit as above, but with a biggish capacitor added to stretch the timing. Yeah, that’s what should go in the pod.

But, part of the alchemy depends on the relative voltages around the transistor & LED. The original circuits work for a <=1.5v supply. I only had a 3v (2106) battery that would fit in the Kinder pod.

Because it’s Dark Art material, I didn’t bother analysis or trying a sim in Spice (which almost certainly wouldn’t be remotely accurate), went straight to breadboard.

Snipped a little toroid off an old burnt-out PSU, wound on 2×25 turns of thin wire (Wikipedia suggests 20t, but because I was a bit sloppy threw in an extra 5 for good measure). Went for the trial & error school of design. Starting from Eman2000‘s 1.5v blinker I had a good play.

So…first observations : the capacitor/resistor sizes bear little relation to the timing compared you might expect, seriously non-linear. At 3v the LED appears permanently lit (I think most likely blinking at a high frequency, but I didn’t bother to check).

Semi-logically, I started by tweaking the LED (the loading is significant) – variations on colour, series resistance, series diodes for voltage drop etc. Two series red LEDs seemed most promising. Scientifically changing several variables at once, I thought maybe reducing the current from the battery might help. Yes & no. It made quite a difference but only approached what I was after when backed up by a capacitor. I tried a couple of other transistors, it didn’t seem at all sensitive, probably any small-signal bipolar would work. I didn’t bother playing with the transformer, too fiddly.

This is the circuit I settled on:

It flashes a couple of times a second with either 3v or 1.5v. I don’t believe the measured current because it was a very intermittent draw (I know, should’ve got it on the scope, but I just wanted to get it done). Probably more like 3 x 25uA, which would suggest 1000 hours before the battery reaches 2v. That’s a guess. But given that it can work below 1.5v, it should last a month or two, if not a year or two.

Soldered up on a tiny scrap of stripoard:

Bit of hot glue & tape:

Then it crossed my mind, where’s she going to put this? Fridge magnet! So I hot-glued 3 tiny magnets into the Kinder pod.

Turns out there are much more hi tech approaches to similar problems, ATtiny45 EverLED and TritiLED both use microcontrollers for very long life little lights.

It is a little creepy in the dark. Leave the radio on.

Bitcoin, Ponzi & Capitalism

Just some passing thoughts I felt like spitting out. Coincidentally I watched another episode of Adam Curtis‘ excellent Can’t Get You Out of My Head documentary series last night.

I’ve seen talk of Bitcoin as being a kind of con, because it’s all virtual. But virtually all money is virtual since Nixon dropped the tie to the gold reserve.

A situation that Curtis brought up was that of the Appalachian coal miners in days of yore. They worked for the coal company, got paid in the coal company’s own currency which they had to spend at the company store.

Nowadays, the company isn’t so easy to demarcate. But a miniscule percentage of people control the vast majority of wealth. It may be someone middle-class that owns the store, but they are still company employees. It’s the company‘s definition of money everyone has to use. Hard to point the finger very precisely, but a vague direction might be bankers, the uber-rich and most politicians. Oh yeah, and the management of Google, Facebook, Amazon & co. It’s popular to blame the mass media for a lot of things – but I reckon they’re more like symbiots, creatures down the scale that live on the pickings of the bigger fish.

That may sound verging on conspiracy theory, deep state etc. But it seems to run deeper than that, and isn’t even hidden, not even in plain sight, it’s apparent all around.

Such a system might be considered just the way things are, the lesser of various evils (fascism, USSR-style communism), at least for folks in wealthy Western nations.

But a fundamental flaw in this setup is that it’s inherently unsustainable. The generation of wealth currently comes from the exploitation of natural resources, and the way that’s currently done, they are finite. A side effect, of greater concern, is that the climate is getting messed up.

A crude umbrella term for this system is Capitalism.

I’m no political theorist, not sure I’ve read more than a page of Marx or a paragraph of Adam Smith. But as a student I was briefly drawn into a pyramid scheme (not long enough to lose more than a tenner, but still). Is a well-known company, I forget it’s name. Cleaning products mostly, you make your money by getting other people to sell the product for you. But these systems collapse because there isn’t an infinite tree of sellers/customers. Well not collapse exactly, the money gets funneled uphill.

Seems to me the current use of the planet has a lot in common. The resources of people further down the tree, in time, are being funneled into the profit now. People in the past have being sold a future based on the wealth of people now, and in turn we’re sold of that of those not even born yet. In this context wealth equates to livable conditions and breathable air.

Count as a rant? Whatever. Penned it, can get back to consuming.

PS. Bum. I used Bitcoin as clickbait, didn’t really follow through. It’s just an extreme form of same-as. Extreme in the fact that it cosnciously seeks to waste natural resources to make (mine) money. If an alien wanted to turn this planet into a place more hospitable to sulphur-breathing hell beasts, perfect invention.

Incidentally, long before I knew anything about Bitcoin, my old mate Reto came up with an eco-alternative : Greuro

Headless Chatterbox

I wanted to jot down some notes about where I’m at with this thing, on writing the title realised a peculiar coincidence. Bit of tech stuff below, but first…

For a kind-of fork of my own project, Chatterbox (a speech-like synth) I am making another version of the hardware without all the manual controls. It’s common to call software without a GUI ‘headless’, near enough.

But…

There’s a pub about 5 miles from my mother’s house, the Quiet Woman in Earl Sterndale, Derbyshire, England, with a gruesome legend attached. The landlord’s wife used to nag him too much, so he cut her head off. Er, yeah, that’s basically it. Except he wasn’t held to account for his actions, instead praised by the pubgoers who had a whip-round for her headstone. Traditional values, eh.

Decapitation figures in another local legend to which I’m personally attached. Tidza Sawyeds is the sobriquet given to people like me that come from Tideswell. The story there goes that a cow got it’s head stuck in a gate so the farmers sawed the cow’s head off.

There have been arguments that it wasn’t foolish, that a gate might have been more valuable than a gate at the time. Just nah. It’s a smear by people of Bradda (Bradwell), about 5 miles away. In another reflection of traditional values, I grew up calling anyone that came from that far away a foreigner. Even in my generation there were differences between the dialects/accents between villages over that kind of distance. But you knew they were foreigners anyway, because you weren’t closely related. (Image above lifted from Phil’s old blog, er, cousin Phil).

Anyway, one Wakes (carnival) a Bradda woman was noted for lifting her pig up on the wall so it could see the brass band go by. [That does actually speak of more humane animal husbandry down there, hey ho].

Anyway

I am daft. Spent over a year on the Chatterbox project (based around an ESP32 microcontroller) without really getting tooled up properly. I started knowing virtually zero C++, still very vague, so quick debugging would be at a premium. Even just a computer better than the minimal laptop I’ve been plodding by on would have been a huge time saver.

I’m finally getting there. Best : a new computer, flies. Flows! But there’s a lot more.

I was sussed enough to start using PlatformIO (on VSCode on Ubuntu) rather than the Arduino IDE early on. (That’s excellent for small projects, but as the codebase increases, the productivity plummets). Only recently found out there are things a step better than debugging over serial [can’t see link right now – but there are], also found out the JTAG interface cards are really cheap.

Ok, if you’ve read this far you’ll know this, but just in case. The processor in the microcontroller ain’t the same as the one in the computer that’s doing the compilation. There’s no direct way of examining what’s going on inside. I’m not aware of any ESP32 emulator on a x64 PC (unlike with Android devices).

But, there’s an interface called JTAG that can provide this, via a cheap little card:

USB connector on the right to the host computer, hook up a couple of jump leads from the sockets on the left to the microcontroller, you have full debugging facilities. Can halt a program mid-run and examine the variables.

Only snag for me, all the IO on the Chatterbox ESP32 is spoken for:

GPIO 12-15, that JTAG use, are already taken up by some of the switches.

Now, I had thought since before I started that what I was doing was making a microcontroller-based synth. All the human control was novelty, everything could be controlled by MIDI. The ESP32 is plenty fast enough to do DSP on audio, all you really need is a sensible DAC for output, and they are cheap & easy to code in.

Also, although my main motivation is to have something I can play with (and learn all this shit), it would be good if other people could play with one too.

My brother Nigel would be very much up for having a machine that makes sounds like no other. But I haven’t the will to make another machine like the one above. But everything can be controlled by MIDI, hence the notion of a Headless Chatterbox. An unquiet woman.

All that’s really needed is :

From left, clockwise: ESP32 module, UDA1334A DAC module, stereo jack out, DIN MIDI in, power regulator.

Also the 3 or 4 components for wired MIDI in (just needs an opto-isolator and a couple of resistors) and a box.

(MIDI over BLE is definitely doable, but I haven’t played yet).

You will also note that I’m only going to need a couple of GPIO pins on the ESP32 for I2S to run the DAC, another for the MIDI in. Which means I can hook up the JTAG debugger, code a lot faster!!

Oh yeah, forgot to mention, instead of the switches & pots for control, could also have a browser-based UI. I’ve got proof-of-concept very much working. Only hassle there is the thing of IP addresses. Ideally this will run just as another host on your local network. But giving it the creds is a pain. Can be configured as an access point, but when I’m doing stuff I want to have an external Web connection not just to this lump.

That part might take some time.

But I reckon, with the aid of the JTAG card, if I focus on the headless version for a bit, can find the answers relatively quickly, and find the DSP algorithms I really want much faster.

Getting the headless thing together as a little module would be very sweet, might even have commercial potential. I do need to sort out all that ^^ first, but an obvious step after would be to make up a PCB, they’re so cheap to have made now (and even soldered up).

State of Play

These are the projects I’m working on. All are open source, most code etc. is at https://github.com/danja. My Home Page is at http://hyperdata.it.

This doc is a work in progress.

• HKMS – Hyperdata Knowledge Management System
• Chatterbox
• About

Hyperdata Knowledge Management System

Status : subsystems in different stages

HKMS is the umbrella name I’m using for a variety of applications on a variety of platforms. All these applications will use a common SPARQL store as database. In effect, the applications are different views of the underlying data. Facilities I’m after are :

• notetaking (FooWiki, Thiki)
• bookmarking (Foolicious)
• project management (Trellis)
• automated information gathering (NewsMonitor)
• Web publication (Seki, Clonio)

The motivation is that while there are very sophisticated and useful tools available

Chatterbox

Status : hardware v1 done, software mostly in place, close to v1

Chatterbox

About

I just got a new computer which has come like a shot in the arm. After a couple of years using an ultra-cheap laptop, it now feels like a lot of friction has dropped away. Motivation has rebounded, feeling a lot less overwhelmed. But I’ve maybe a dozen projects, hardware (analog & digital) and software (Web server, client, desktop, mobile, embedded) all ongoing, even if most have these have spent a lot of time on the shelf.

Time to make a list.

Chatterbox, next

My main reason for Chatterbox was to have something with a well-defined target that I could focus on for a few weeks and get my head away from all the open-ended projects I was being overwhelmed by. About a year on I realise it wasn’t the best choice. Making noises with microcontrollers turns out to also be open-ended. But I have at least got the hardware done.

But I’m still not remotely happy with the DSP algorithms I have in place and there’s still a huge amount to do when it comes to MIDI & Web interface.

So I wasn’t stuck in my own little bubble it occurred to me that making a version my brother could use would be a good idea. He’s tech-savvy, does music but isn’t a coder. But because he lives on the other side of the world (Lostwithiel) a stripped-down version of the hardware seemed a good idea. Basically:

An ESP32 module, a reasonable DAC, MIDI in and audio out.

If I can get something that’s useful/fun out of this that Nigel can use then I’ll have fulfilled my imaginary remit. I don’t know, maybe even consider making it a commercial product (it’ll all remain open source whatever).

A major pain with coding for microcontrollers that, without assistance, you can’t properly debug because that device’s system is a different processor than the machine you’re coding on. But those delightful Espressif folks already have it covered.

What’s prompted me now to go in this direction is hearing about JTAG and just receiving an ESP-Prog module:

It is such a misery trying to debug audio stuff via printing to a serial console. You want things to happen around 20k times a second, serial takes a lot of ticks. Added to which is the drag of the edit/compile/upload cycle. Can really be plodding through mud.

On the plus side, it has forced me to understand better what the code is actually doing, C++ is effectively new to me. But there are limits to my patience.

This module, apparently, attached to the appropriate I/O on the ESP32 allows proper debugging – pause, examine value of variables etc. – within PlatformIO (which I use on VSCode on Ubuntu).

Now funnily enough, building on impulse, I’ve used every available GPIO line on the ESP32 for controls (+DAC +MIDI) for Chatterbox. The pins JTAG uses are attached to switches. But, a headless Chatterbox doesn’t have that issue. So I can focus on the other bits, try and get good sounds out of it, then dump the exact same code down to the version with knobs on.

Haven’t tried it yet, so fingers crossed.