Cancelled. Anthropic can get bent if it wants to market itself as the ethical AI choice and at the same time lies in bed with those openly pushing for a surveillance state.
I've started experimenting with claude to run end-to-end tests for an emacs package I'm developing. It's incredible.
The way it works:
1. You start emacs in daemon (server) mode.
2. You prompt claude to instantiate an emacs client and write/run tests.
That's it.
Claude will then "pilot" an emacs instance where you can visibly see it running tests. Since almost everything is a first-class function in emacs, and emacs enables almost complete introspection, claude can debug the code in the execution environment. You can also just look at the piloted emacs instance and prompt the agent on what's wrong in the running application state.
This is much more thorough then just having claude write unit tests because many of the issues you might encounter are visual/gui things - which, again, because emacs allows so much introspection, can be examined by looking at the current application state.
Isn't this just a Bambu marketing point? I have a Prusa MK4S and a Centauri Carbon - they both print without any fuss and can be operated without any deep technical understanding.
I'm not sure what went wrong exactly, but a common issue people run into with lofts is caused by the fact that FreeCAD tries to match vertices to vertices. If you use the split edge tool on your square to create 8 edges out of the existing 4 (maintaining the same square shape), it should loft smoothly to the octagon.
If the loft appears twisted it only means you need to change the order in which you created the geometry. FC in addition to matching vertex to vertex will also match them in the order they were created. This allows you to purposely create twisted shapes, but sometimes can happen inadvertently also.
I can't tell for sure without knowing exactly what's happening, but one reason is that if you create a sketch from the Sketcher workbench, it will not be added to the active body.
If you create the sketch from the Part Design workbench, then it will be added to the active body.
A Body is specifically a Part Design concept, and FC doesn't presume you'll be working in PD, so this makes sense in a way - it works on the presumption that the Sketcher workbench works with other workbenches and not just PD specifically.
One thing to note is that creating sketches from Sketcher and PD is different. Sketcher offers attachment options to faces, edges, etc., while PD only offers to attach the sketch to base plane (XY, XZ, or YZ).
There is a good reason for this also. The reason is that in designing parts, especially complex parts, it is highly discouraged to use faces or edges (i.e., features) as attachment points because it makes your model very brittle against changes.
This is more of a general CAD philosophy than a FC thing. It's better to set where a sketch attaches based on variable values. For example, if you have a cylinder with a height 20 and you want to attach a box to the top of the cylinder - rather than attaching to the top face of the cylinder, it's better to create a variable h=20, and set the cylinder to height h, and set the box's z-value also to h.
In FC, I use VarSets for this. I used to use the Spreadsheet workbench, but found it clumsy.
As an opposing viewpoint, I also use FC extensively for designing moderately complex parts (fully parametrically constrained assemblies, dozens of parts per assembly, mechanical components involving motion).
I've also extended the functionality with python, and have heavily customized the theme and shortcuts to fit my personal taste.
I not only tolerate the software, but enjoy using it, and am quite proficient at it.
I would recommend FreeCAD to others, but with some caveats. The most important being that they need to be willing to tolerate a few hours of introductory material, and second that they are serious about using the software long-term.
Otherwise, I'd probably just recommend Onshape. But, for many others, FC is fully viable.
FreeCAD is still in active development and there was a major release (1.0) recently. There will be a feature freeze soon where the devs will focus primarily on stability and eliminating bugs.
That said, I use FC as my main CAD driver and, not only tolerate it, but enjoy using it. I had to watch several hours of introductory videos to get the hang of things initially, but now I'm quite fast and proficient.
The initial pains and common complaints about UI and such, are basically non-issues for me now and when I model, my cognitive energy is basically devoted to the design problem itself rather than issues with UI or the behavior of the software.
It's necessary to put the time into learning it, but it's worth it.
Sorry about the bad experience. If you decide to give it another try, it's worth spending a few hours on some intro tutorials first. I recommend MangoJelly on youtube.
FC is not a program you can just open and start using, especially if you have zero experience with parametric modeling.
If you're serious about design, modeling, engineering, etc., and want to own your own data, it's worth investing the hours to learn it starting from the very basics.
Some of the basics aren't immediately obvious or even hinted at very well for new users, but the "problems" that come up are consistent with its own editing model.
> complaining that the sketch doesn't belong to any object
The sketch is by default attached to the "active body". Active Body is a simple, but important concept to understand. Any operation you do, including adding a sketch, is applied to what is designated as the active body. You designate the active body by right-clicking on the desired body in the object pane.
> It suffers from too many "workbenches"
Another understandably common source of confusion. There's the ever-confusing Part and Part Design workbenches.
I think it's best to just ignore Part and use Part Design whenever possible. Part lets you do operations at a more granular level, but Part Design provides a lot more QOL enhancements and is more intuitive. For the vast majority of things, Part Design is more than capable. I would only use Part workbench when absolutely necessary.
You probably understand all of this already. It's directed more towards the reader. I feel the need to defend FC when certain accusations are brought up. It's immensely powerful, capable, and usable. In my case, I can work very rapidly with it - though it's taken some time to arrive here. The project deserves more than just aspersions.
Productive FreeCAD user chiming in here. I understand the frustration. It has a moderate learning curve, but the editing model forms an intuitive picture once you learn the basics and start using it.
Also second the MangoJelly tutorials. You will have a much better time if you walk through a few lessons first as opposed to just winging it and expecting to understand how everything works immediately.
I recommend starting with vanilla Emacs and just adding things as you find the need for them. Emacs comes with a lot of things OOTB. After a decade, my only essential package addon-ons are magit and yasnippet.
I have other packages installed, but they're esoteric for my own purposes.
Okay, but that is a complete misrepresentation of Prusa machines. You can buy a Prusa today and it'll spit out prints like any other modern printer. And this has been the case for a long time now. Only a small minority of people actually tinker with their machines. It seem like you're attacking a complete projection.
Are you confusing Prusa with older Enders? If you buy assembled, they just work like any other modern printer. In fact they're known for long-term reliability. If there's any detraction, it's on the basis of printing speed, price, and availability, but not on fragility.
It's not a lot of work, but it takes a little bit of existing Elisp knowledge. You can still evaluate plain SQL code in a buffer (sql-mode) and get most of the benefits I described, you just won't get the convenience of evaluating Lisp forms.
It might be time for me to start looking into Chinese models or purchasing hardware for local llms, even if the cost amounts to 5-10k.