Show more

Interesting. Temperature doesn't seem to matter much. Running at 205 instead of 215 showed a minimal improvement but not by much.

Printing slower seems to do the trick. I printed the test piece with 20mm/s (1/3 of the original speed) and I almost got through the whole piece, maxing out at roughly 65-70° overhang. Printing at 40mm (the first test) already got me from the baseline to about 60° which is probably good enough for most things.

Show thread

Trying to approach my overhang problems (3D-Printing) systematically after franticly trying to adjust all variables at once.

First test is speed and slowing down a print definitely helps. Now trying to figure out how temperature affects it.

Finally my 3D printer is printing again. It now runs from an SKR 1.3 and that took me quite a while to configure into a state that's actually working. A ton of stuff I can tweak, but you can already hear and see the improvements by using Aluminium extrusions instead of cheap acrylic :)

First improvement that's needed is a proper housing for the new board.

And then I can finally test all those great new filaments.

The mere fact that you can often fix a problem in JavaScript-land by deleting node_modules is making me irrationally angry.

This is not acceptable. Neither npm nor yarn are doing their job properly when it comes to this. yet, here we are and obviously not enough people care for it to change.

Damn. Lots of great filament arrived today from Das Filament, but I completely frizzed my controller board yesterday and replacement won't arrive before tomorrow.

Still no idea how that happened, but I managed to brick two Megas, probably two RAMPS (not sure about that), and a couple of SilentStepSticks. Including letting out some magic smoke.

Makes me want to get a second printer as a backup. Which is probably the worst urge to have :)

First used up spool of filament since I bought the printer. Of course happened in mid print so for some tense minutes I was juggling two spools and manually feeding the second filament (same brand, different colour) in until the extruder grabbed it. Not a method I would recommend if you can't help it but I seriously underestimated the amount of filament left on that spool. On the other hand: No stupid 5 meters of leftover.

It seems like I'm collecting all "stupidest reasons for 3d printer malfunctions" badges now. Had really weird print failures and z-shifting recently. Turns out I had the printer (which sits on an inverted carrying box) too close to my recliner chair and the x carriage bumped into it regularly. I lost at least three prints to that today before I noticed. 🤦‍♂️

Ah, nothing to start the year like a full out war between USA and Iran. Because those are always fun and not destabilising at all.

I thought we all agreed to only produce good news in 2020? 🤦‍♂️

The reason why I didn't want to create complex XMLs by hand is because I thought those Java libraries would include some sort of business rule validation.

Turns out, not so much. The validation is all left to the post production stage where the XML is validated against schemas and schematrons. The good thing is that I can 1:1 replicate that with ruby (the schematron validator is comically slow, though). And I can add my own pre-production business rule validations in a language that I like.

Show thread

I spent way too much time trying to get a couple of libraries working to create so-called ZUGFeRD invoices (essentially PDFs with embedded XML to allow for non-magic machine reading) in Java or rather Kotlin. Yesterday I gave up and built a library for this from scratch in Ruby in a day.

This is not a diss on Java or the libraries. It's the realisation that trying something tricky in a non well known ecosystem is not really the best of ideas.

Did a stupid mistake earlier today and broke my z-sensor holder while printing. A couple of hours later (with some quite frustrating moments of trying to manually level the printer which I haven't done in 18 months or so) my now auto bed levelled printer prints a spare sensor holder so that I will never experience this again (hopefully).

Which means I should be ready for the E3D Hemera conversion tomorrow. Finally, a good extruder and hotend, able to print flexibles.

half/byte boosted

I think the three R's (Reduce, Reuse, Recycle) are missing a critical fourth R. (Reduce, Reuse, Repair, Recycle)

I'm pretty sure "Resilience" as a word will make a huge comeback in the coming years, btw.

We need to build resilient infrastructure, resilient societies and communities, resilient food sources/agriculture.

In many cases this means decentralisation, de-scaling and rediscovering things from the past.

The concept of "Too big to fail" will be seen as an expression of a clueless past.

I want to write about all of this so badly to sum up my thoughts, but I'm often reaching my own boundaries of knowledge quite quickly (and also it's very exhausting) and so I have trouble putting my thoughts into coherent streams of words.

The only conclusion is that I should read more from people who are able to do that and share my findings instead of trying and failing to do it on my own :)

Show thread

The thing is, "reducing CO2 emissions to zero or negative numbers as quickly as possible" is only the first step in a looooong list of things we need to get in order if we don't want to end this in a complete collapse of human societies as we know them.

We're already failing badly at that one.

At the same time, current trends in politics towards authoritarian policies and dismantling democratic safeguards makes our societies a lot more brittle and less resilient against disruptive events.

half/byte boosted

Der rechtsextreme Matteo Salvini hat einen neuen Gegner: Die Sardinen-Bewegung protestiert friedlich gegen Hetze. Sie verbreitet sich rasend schnell. taz.de/Italiens-Zivilgesellsch

Today's episode of "everything is broken" is brought to you by my Linux desktop and Google Chrome:

On XFCE, Google Chrome reports weird values for computed style properties (that we use in a couple of tests), for example 1.994px for a border that is definitely set to 2px in the css.

On GNOME desktop, it reports proper 2px.

I don't even know where to start to look for what I've configured wrong for this to happen.

Installed Catalina on my poor MBP to be able to test some Notarization issues we're running in to.

This is the Apple version of the Windows Vista debacle, isn't it? I don't have enough hands to click away all those "X requested Y, allow it?" windows.

Also, it just asked if it is okay for Terminal to access my Documents folder. Really?

Somehow ended up in a rabbit hole called . Torn between building the (quite possibly) first ruby library for generating the XML or instead undusting my Java and using one of the available open source solutions.

In any case this looks not like a fun weekend hacking project.

(In case you're wondering, ZUGFeRD is a standard to embed XML data into invoices to make machine readability better)

First real setback on my Linux journey: After updating to Ubuntu 19.10 my desktop kernel panics after waking up. No idea what could cause it, searching the interwebs does not yield help.

Show more
Ruby.social

A Mastodon instance for Rubyists & friends