It's a boy thing, I can relate 
3 Likes
Not just a boy thing 
As I branch out I just move the applications needing less power but still need a raspi or similar to another device, and upgrade those I need. I was on the waiting list for a first gen raspi B when they gave me a free upgrade to the about to release 1B+ (First batch). Still use it, though mostly for prototyping and as a headless remote camera. I know have a 3B+ that runs both node-red and Snips while Iām developing, and the plan was to move node-red to a yet-to-buy dedicated 3B+ when things were production ready. Snips runs fine of the 3B+ so far, meaning the dedicated node-red will instead become a pi 4.
But I definitely get it. Iām about to start experimenting with Pycom devices, ESP32 based and running micropython. Saw Pycom is releasing an 8 channel LoRaWAN gateway for 49 euros (excluding a LoPy, so like 75 total) and I already start drooling even though Iāve no current LoRa projects going on. I might actually use a setup like that for my wheelchair activity tracker, but thatās not in the active development pipeline yet...
1 Like
I am messing about with micropython, you will not be disappointed.
Iāve done most of my programming of the last 5 years on a Python based stack. I remember when I first had to use the arduino IDE in school, first for Processing, later for programming Unos. All the while thinking if things wouldnāt be more efficient in raw C, or in an actual higher level language . Micropython feels like a fresh breath. Iām already planning to use a bunch of regular (cheaper) WiPy devices for other hardware projects, like smart kitchen scales (MQTT controlled, hx711 connecting the load cells back), and a controller for an autonomous herb garden. As for the last one Iām not sure if Iāll manage in the 4mb available RAM and Iām not looking too forward to switch that one out for a pi instead. However if it means I can connect several more plants to the same controller I might consider it... it definitely wonāt need a pi4 though 
1 Like
One day when the kids have finally flown the nest, I'll be able to afford to retire and then I'll be able to boast of a project To Do list like this!
I aspire to it 
2 Likes
@TotallyInformation Juilan, hate to tell you but one of mine flew the nest, got married, had three kids and moved three houses up the street. Guess how many days a week we see them....
2 Likes
Iām pretty sure my way of getting that free time isnāt the preferred one either. Itās based on (lifelong) disability and depending on the equivalent of a PIP, with the constant thought/threat our version of the DWP will take it away. So my projects are all based on a shared goal of a future where Iām mostly independent and donāt have to fall back on assisted living. And it might cost a bit to build everything up, but durability is a must, and low power usage a plus. Iāve savings, which is how all of this is budgeted up front: if the costs of prototyping and building a solution, plus upkeep of it all end up outweighs having to hire people (think caretaker, nurse, cleaner), Iām going to try to use my own (created) technology instead. If I mash 3 benefits together I might just be able to afford living on my own, but Iām not expecting to ever go beyond social/council housing. Adding an extra hurdle in that everything has to be modular, being able to pack it up in boxes and move it to another similar property is a must. But as a result creative solutions happen. Just tonight I came up with yet another potential (mini) project regarding small robotics to control a bunch of servos and stepper motors, moving sets of beams, with the goal to put a duvet cover over the duvet. I can see the components moving together and the kind of sensors it might need in my head, but actually working that out might be difficult. But hey, my hands no longer shake by default when working with small electronics so thatās a plus already. (Side effects of medication causing artificial Parkinson, while doing an electronics course is a hellish affair). Instead I get full body tremors so 
itās a never ending challenge, but it also shows the reasons for my automation projects all too well.
1 Like
Urg! It's fine, we are moving countries as soon as they go! 
I had kind of guessed that might be the case and I don't wish to make light of that. I very much admire the fact that you have lots of plans and ideas no matter what the issues and struggles.
I very much hope that you will be able to find other people also wishing to extend the capabilities of those with physical issues and you will continue to innovate on behalf of everyone who has so much to deal with.
The potential for IoT and social media interactions to extend the opportunities for people with physical issues is enormous and I have no doubt needs more people who live with those issues to help work out ways to overcome them.
Thankfully, this forum is always helpful and has lots of creative, innovative people who are always ready to help others.
1 Like
As long as I can still joke about it Iāll cope. If I lose that last bit of humour too things will be a lot harder. And while Iām still capable of both designing and building solutions I hope I can keep going with that. Itās going to depend on the speed at which I deteriorate. Based on the trends Iād say Iāve (at least) a decade of clear speech left. At this point Iām the only one able to notice a difference there. I donāt think Iāll ever walk again without problems, but the last couple recoveries brought my walking range back to a couple miles. About as much as my current wheelchair range Iād say. But itās still freedom of a kind. That feeling of taking it off-road in a forested area is great, being surrounded by nature is perfection like that. And as long as I keep having the strength and control in my arms to do so itās good.
For the last five years Iāve been preparing for a future where I will lose my eyesight. It includes a rather unsuccessful experiment with teaching my computer how to program using speech to text. It worked, ish, but getting the speech to text to understand what kind of code I was looking for was much harder. Itās why my current setup is going for an NLP based, dialogue shaped approach. So I can instead request the assistant to turn on/off notifications, or refill medications (guess what happened in the week where I wrote and tested most of that input: I had 2 empty pill boxes lying around thinking they were filled, and that hard name with several trained synonyms, yes thatās the one I kept testing with and managed to run out off). All MQTT based, so requests are subscribed to in NR as to handle the task, then finish the results and report back, again over MQTT. Can I just say I love that protocol.
For now the focus is on my own QoL, and while Iāve told myself I donāt ever want to go back to a job in IT, creating smart mobility aids (supporting the user in more ways than just the regular one through additional sensors and/or motors) is something Iāve been thinking about for several years.
This is something I very much hope to accomplish in the future. Here in the village theyāre currently updating the pavement to increase accessibility, and making it more future proof, and they have a funny way of showing that. I put a fluorescent jacket on the back of my wheelchair to ensure cars see me when I have to use the main roads instead of the pavement, while I wait for the first (lethal?) accident to happen because a person with a walker is going to fall there while using the elevating entries. Once again clearly designed without actual users in mind while calling it future proof. I mentioned going off-road, well from experience forest paths are usually safer (and when compared to the old pavement more of a level surface too) than the pavement in the village here. Things you learn in practice: which mailboxes are accessible, which walkways risk causing a fall or a flat tire, where will you get stuck and will need help to get loose again (spoiler: the crossroads two streets away, because of the new pavement). Technology based things Iām hoping to accomplish here are gyroscope based checks for stability, and trying to get the wheels that hang free in the air, out of control, back towards the ground again, or extend a tiny motorised wheel to get it out of the situation.
And while my own progress might be slow, thatās what I hope to do in between, sharing my knowledge where applicable and helping others where possible.
1 Like
Back to the pi4 topic though, has anyone here by chance attempted to run TesseractOCR on it already? Iām hoping to run it on the future server-pi that will also run the main node-red instance, to OCR mono spaced, serif typed refill notes from the pharmacy to get a couple codes out, and if itās able to run better on a pi4 than a pi3, running it there would be preferred. So far Iām using it on my desktop Linux system without problems but it has monstrous specs so Iām not surprised it does fine there.
Not tried it for a long time & don't think I ever bothered on a Pi. I don't know how much it has changed but it always used to be trounced in accuracy by commercial tools, particularly the Microsoft Office ones.
If you have a version of Office, Microsoft Lens (mobile) with OneNote gives you very good results indeed.
A Pi4 4GB version shouldn't really have any issues running Tesseract I wouldn't think given that it pretty much matches the performance (excepting the SD-Card & possibly the thermal restrictions) of a low-end laptop as far as I can tell.
You might want to check out Andreas Spiess's post entitled "Raspberry Pi 4: Test and Weaknesses." He does some testing and looks at the highs and lows of the RPi 4. The bottom line is that the RPi 4 looks great on paper, but the reality is that it doesn't always measure up to the specs.
https://www.youtube.com/watch?v=0diTHAmVbhc
Cheers.
2 Likes
I think Andreas' review is really less about the hardware specs and more about the current (pre)release state of Raspbian Buster. The giveaway is his conclusion that "They will fix the issues!" If the problems he notes were the result of serious design flaws or compromises in the hardware, he could hardly be so optimistic. Other reviews have questioned the decision to implement 4K video or dual displays (because we can?) or to advertise a "desktop" experience that may be a distraction from the original goals of the Pi or not useful to the maker community. My feeling is that people in this group working with video, image processing or other AI will see real improvements.
I must say,,,I like him, Andreas, in the youtube link, honest and clear
Yes, he has some good videos.
I would tend to agree.
Having recently read up on USB-C specs, I think that the reason they have done this is that it is part of the spec. 2x4k @ 30kHz and 1x4k @ 60kHz seems to be part of the standard spec. So they probably get this pretty much for "free" simply by choosing a standard chipset.
Yes, I think 4K is "free." Dual-monitor capability means buying micro-HDMI cables or adapters even for just one monitor --> complaints from the same crowd that winges whenever Apple changes connectors 
True, but it's not proprietary & they aren't trying to get a cut from anyone making cables more expensive, that's what bugs people about apples tactics.
1 Like
Right. Just a feeble attempt at humor.
One valid point raised is that the Pi Foundation may be catering to vendors or users who buy in large volume for commercial displays. I wouldn't really have a problem with this if it helps keep the price down.
I've actually always wondered why they haven't produced a true headless Pi. Seems like an obvious option given how useful Pi's turn out to be for servers.
2 Likes
They do, they have compute modules.
