The new gui can be found here - https://github.com/Pragmatismo/pigrow_windows_remote/blob/master/test_guiWin_008.zip

This includes picam3 support with rpicam, they’re great cameras and make focusing much easer - really good for close up work where depth of field makes focal distance important.

Also Venv support which is required in Bookworm, the newest version of raspberry pi OS which is required for the picam3, pi5 and a few other new things. If using bookworm then set up the venv before running the install wizard or installing any python modules.

Note - using bookworm there may be other changes they’ve made, for example the location of the /boot/config.txt file has changed so the boot config tools in system_pnl wil not work, i’ll fix them but for now make any required changes manually (e.g. i2c baudrate, 1wire config)

The timelapse and localfiles panel have a few small updates and fixes, and there are a couple of other minor improvements.

I’ll keep testing with bookworm and fix anything that needs it but hopefully will be able to focus on moving over the graphs tab so we can finally reorder everything. I’ve got a few projects up and running that will gathering timelapse while i’m away, i’ll also be using some of the time i’m away to work on finishing some designs so hopefully i’ll have a load of new stuff to add soon.

If you're not interested in technical details don't worry I'll sort this all out and make it automatic so you can focus on growing good plants, it only applies to using the new bookworm version of the OS and i'll make every effort to ensure changes i make won't affect existing installs.

The new raspberry pi OS, version Bookworm, requires the use of venv (a Virtual Environment for python) when installing modules with pip, although some modules can be installed via apt a lot of the modules we use aren't available. Using a direct connection to the pi with a screen and keyboard it's pretty simple to use,

python3 -m venv test source test/bin/activate 

you'll now see (test) at the start of your prompt, anything you run will now use the python version and modules stored in ~/test/ which is really useful when running things with awkward version dependencies which is common with the neural nets and computer vision projects. It's also a bit awkward because if you want to run a script from cron then it'll use the system version of python and it's modules which won't include the ones you installed... likewise when you open an ssh connection (such as from the remote gui) then it'll want to use the system wide one.

We can of course explicitly tell it which version to use in cron but this would break all our tools and make it ugly, we can instead add a line like

PATH=/home/pi/test/bin:$PATH 

this puts out version of python at the start of the path list for cron so when it goes looking for which version of python to use it'll use the one in our venv rather than the system version

For ssh there's two things we can do; the first is to add a similar line to bashrc, this is a script which runs when a new connection is made or a shell opened, there are two options at the end of the line we can add one of either

export PATH="/home/pi/test/bin:$PATH" 

is the same thing as we did with cron, or

source test/bin/activate 

which does pretty much the same thing, the environment we created is enabled which means requests for python go to it instead of the system version unless explicitly stated.

I'll have to edit the gui because ssh via paramiko won't trigger the bashrc file unless a shell is created, again it's nothing huge just check to see the version of the OS it's using and enable the environment if it's required.

unfortunately these will still require me to edit pretty much every script on the pigrow because to make them executable the first line is '#!/usr/bin/python3' which is an explicit path to the system version of python... it's a fairly easy fix thankfully though instead of replacing it with a strict call to the venv which would break systems currently running without a venv or any future installs on an os other than bookworm... I'll instead make an sh script which checks which version or for the presence of a venvpigrow or whatever i name it and if present runs the script using that python, if not runs it with system python. - i've tested it and it seems to work, just got to do a clean implementation of everything and the tools to put them in (i'll add to the install dialogue wizard)

Anyone that knows about this sort of thing love to hear your opinions or suggestions, i've read way to much documentation and it's fried my brain back to not understanding anything. I also have a suspicion they'll make some changes in future versions because this seems to be pretty unpopular,

I'll have this done and a new version out this week because i'll be away for a few weeks, all my video projects have been on hold while i work on the new version but i've got a few test boxes up and running and some interesting projects ready to start running so when i get back in august i'll have loads of stuff and hopefully soon after be ready to launch the coding contest i mentioned before for image analysis tools.

To use the picam3 and other newer CSI port cameras I'm having to write a new capture script as they've switched to the module picamera2. I thought this would be a simple case of copying picamcap to picam2cap and changing a few lines, adding in some cool new functionalities and it'd be done in a weekend.

The new camera tools are better, give more control and a greater range of features which are especially useful for some things i want to play with and we'll be able to have a live preview window remotely when focusing and framing the image, so i'm happy for and fully support the change.

It's super awkward and annoying though, i won't bore you with details but so far i've found about ten ways to totally lock the system when taking pics (to the point of needing to power cycle the pi to get ssh back) and a dozen ways of getting an uncatchable error when setting controls or config - also the controls all need to be set in their own special way as ints, floats, tuples, tuples of 4, using an imported enumerator... all of which makes it so much more complex than just letting me give them a string. Fswebcam handles a much more complex situation far more elegantly, but it's nothing unsolvable - i was trying to make it flexible so when they inevitably bring out a camera with a new feature which requires a new control i wouldn't need to change anything to use it but there have ended up being so many special case lookup tables it'll better if start fresh and just use a table from the start instead of polling for settings and things. I still might check for extra setting and note them to the user or something but i'm just going to focus on getting it working with the cameras i've got - not even tried it with the ardu cam yet,,,

So yeah long story short the picam3 is the camera i'm building into sagemoss so i need to get this sorted before i can get that set up and i really want to get that set up it's had far too many delays already.

oh and the other bookworm stuff like the pip issue will be fixed when i release the version of the gui with the picam3 update in, for now just use the prior version of the OS unless you want to manually install pip packages.

a lot of the python packages are installed via apt so it won't break everything, if you see the error message 'This environment is externally managed' then you'll have to ssh in or use a screen and keyboard to install the package you require.

i've also ordered a Picam 3 which can't use picam cap - i'll be testing that (i imagine it works with fswebcam but don't know) and updating the libcam capture tool to properly make use of it's autofocus and etc. I think they've finished the python integration for it so i'll look into that and try to find the best options for controlling it.

I know a lot of pigrow users in areas where it's legal to do so love growing that beautiful flower they just legalised, hopefully the rest of Europe will follow soon.

https://github.com/Pragmatismo/pigrow_windows_remote/blob/master/test_guiWin_007.zip

You can copy your gui settings file and frompigrow folder to carry everything over. To make and play timelapse you'll need MPV installed and in your $PATH (or with the mpv executable in the same folder as the gui exe).

It doesn't currently have the overlay log feature as that will be added when the Graphing is (because it uses the same log loading mechanism).

As it's still too cold to paint or pour resin I still can't finish my growbox which means i'll probably have the user logs and graphs tabs ported into the new gui soon - also i'll be making a video explaining how to make tools for the timelapse tool.

Here's an example of one of the new analyse tools; https://imgur.com/a/UtOPgz4 it makes it very easy to see missing images, light level changes, etc. Depending on your computer it may take a long time with bigger image sets.

new graph too, https://imgur.com/a/uPmVQrT it's seconds since the first image was taken, a smooth ramp means even spacing and any sudden steps suggest a missing period.

I'll be adding a few more upgrades and new tools, any suggestions or ideas i'd love to hear them.

There are a couple of cool features that i'll be testing out at some point, they now have two dual use camera and display ports which will make it much easier to make a camera that measures photosynthesis (using a regular and IR camera next to each other) - you might have seen be doing this before using two zeros which worked but was awkward, this should make it much easier and maybe we can film a spring timelapse of a landscape changing.

the new PCIe access is likely to bring some interesting new additions to the Pi ecosystem, while a lot of focus is likely to be on crypto we'll no doubt see a lot of GPU/Compute modules tailed to AI which might be fun to play with plus some of the performance upgrades already help that a lot which hopefully will allow us to start building some interesting Computer Vision tools into the pigrow, maybe even experiment with some robotics control if and when cheap and useful tools make that fun.

I can't see any fundamental changes which will require changes to the Pigrow, i assume the new dual cam ports will work with the new libcam tools so that shouldn't be a problem but of course until i'm able to test it i won't be able to say for sure and as it's expensive i doubt i'll purchase one until i have a justified use case (probably the stereo camera)

Pigrow Development News: currently filming testing and set up of water sensors for sagemoss testbox, will be making a datawall script for that and checking through everything for improvements in how it handles a cycling water system (as used in hydro grows) so hopefully that video will be out soon and close followed by others finalizing my sagemos testbox.

I’m trying to design a feature i’ve wanted for a while, it tracks the progress of your grow to make it easy to see things like ‘these said they take 12 weeks from flower to finishing, yours have been in 13 weeks now…’ as well as compare them to prior grows and get info like ‘you’ve grown these seeds 6 times and with the new lights it’s growing much faster’ or ‘this variety is growing faster than any other you’ve tried but the yield is lower’ as well as things like ‘grows where the temperature dipped below 10C took 12% longer on average to flower’ (all those are random numbers btw).

I’d like to have it so the log and gui tools are generated based on a template file, there’d be an online repo which you can select templates from or upload your own to for others to use, hopefully we’d get a good selection of useful options that way and it could be great for doing citizen science community projects like all recording the same log info then comparing at the end.

Here’s how i see it working for users;

User presses a button to start a new plant log, they select a template based on the plant type (preloaded, downloaded or custom) this loads a blank form with the initial questions ready to be filled in (appropriate type for the field - dropdown, date picker, text input, check boxes).

When the initial information is input the program generates a simple graphic display showing progress, as time progresses the user reports events (either through the GUI, messaging apps, or etc) which get added to the log and displayed in the graphic - also optionally included is log information from the growbox, high-low temps, humidity, soil moisture level, watering times, etc.

There will also be optional features to have periodic visual inspections, it will message the user through the selected means (in app notifications, push notifications, message to reddit/lemmy user, LED activation, etc) the user will then input the required information - for example; leaf / node count, height, general health / appearance, etc. This allows creation of more interesting data-points to graph and can hopefully be integrated with tools like CV node counting at a future date. The idea is to log things which can help compare future grows of the same seed type or to compare variants.

When the grow is complete the user will select the option to end grow (selecting if it was harvested, moved, failed optionally selection a rating) the data is packaged with other appropriate data and stored. When the user starts the same type of grow again they will be able to show comparisons with other grows - e.g. drawing out markers when prior grows were switched to flower and how long after that until they were harvested. This should make it much easier to notice what negatively and positively affects growing, what certain traits are associated with and seeing how new equipment (especially lights) changes the growth rate and yield. example

Plant log

plant type : Cannabis - photoperiod (name of the form preset used) plant name / ID : sum23cheese1 (unique plant identifier, default suggestion is autogen from date, type and numeric) seed origin : named seedbank seed type : the strain name, feminised notes: a text string used to put any info about the plant worth noting, copy paste of seed blurb or important details

grow medium : soil

expected growth duration : seed - 1-2 weeks veg - 2 months flower - 2 months

seed planting date : 14th June 2023

stuff logged through duration –

date transfer into current box : [cultivation, veg, flower]

prune log : date + what was done [trim, top, tie] + notes

date light duration change : aug302023 - 12:12 new ratio (on:off)

observation log : [problem noticed [bugs, discolouration, mold], new features [first true leaves, first flowers]

user measured : [every week] height = 45cm

harvested date : [date of actual harvest] opinion based rating : 7/10 final plant weight : optional dried yield weight : optional

crossposted from https://slrpnk.net/post/1249169 i'll be checking both so reply in either

I'll be duplicating posts here that i make on Lemmy and checking comments and messages regularly so there's no need to change, but if you're moving away from reddit then come join us here;

https://slrpnk.net/c/pigrow

As you may know i have a few tools in the pigrow that integrate with reddit, allowing you to send notifications to your reddit account is probably the most useful but the wiki updating bot was great while i was using it and i had been hoping to improve it to work with datawalls as they become more established however reddit is not the site it used to be, way back when they did the original redesign they messed up wikis so a lot of users couldn't see them and they were never clear or accurate in their statements about this which is why the wiki updating stuff got sidelined, it was unclear if everything was about to change or not. The current API pricing changes would not affect most the bots that i have written for reddit but it's increasingly obvious that the owners aren't interested in anything but money so not is the future of any work uncertain but it makes me reluctant to support them.

The functionality to receive alerts through reddit will stay as long as it's possible but as it stands i'm going to be diverting my efforts to platforms that i can see growing beyond themselves in the future - i'm a strong believer in open-source and the power of community (this is what brought me to reddit in the first place) so i will be trying out and working to improve more open communities such as Lemmy. (though i'm only starting to explore and experiment with it so there may be better options i switch and would love to hear any ideas or comments on where to go)

For more information on what lemmy is visit - https://join-lemmy.org/

I've set up pigrow on the solarpunk instance mostly because it seems to fit well, i don't know if it'll stay there or if the server will stay up so things might change in the future - if you're on a different server then you can still view and post to pigrow.

From what i've seen Lemmy seems like a really promising and interesting platform, it's a little nicer to use than reddit both on desktop and mobile, as of yet there's not many communities and it's a bit quiet in many areas but it's gaining users fast and i can really see a lot of scope for growth both technically and socially far beyond what reddit is capable of. Give it a go and lets build an internet for users not corporations.

I've finally finished the new install system, i've done a lot of testing and it seems to all work properly though of course any bugs or issues you run into please let me know as soon as possible.

My final testing step was recording the install process for a video i'll now start making, this will be a quick introduction to my new timelapse environment and demonstrate the new install plus setting up switches and LEDs

I'll also start working on writing the new install instructions on the github wiki, updating the reddit wiki to point to the new instructions and rechecking everything.

https://github.com/Pragmatismo/pigrow_windows_remote/blob/master/test_guiWin_006.zip

Hopefully when all this is done i can get back to adding more interesting and useful features, extra sensor support, and etc.

I now 'just' need to bring over the timelapse tab, graph tab, user log tab, and a couple of other small bits, finish updating all the existing scripts, rework a few things to work in line with how everything now works and then we'll be able to get rid of the old gui forever and focus entirely on cool new stuff.