tal

If an instance isn't defederated with another instance, it can talk to it.

You can see which instances an instance defederates with yourself. For lemmy instances, it's at /instances. Just check each end.

So, for example, I'm on lemmy.today. http://lemmy.today/instances doesn't have lemmygrad.ml in its Blocked Instances list (it doesn't defederate from anything, as a matter of policy, in fact).

https://lemmygrad.ml/instances doesn't have lemmy.today in its Blocked Instance, so it isn't defederated on their end either.

Ergo, they can communicate.

Pretty easy to check a pair of Lemmy instances for that.

All that being said, though, if you want to create a series of throwaway accounts to argue with them without them banning you, I think that both you and they are going to be happier if you two stay away from each other. It's just not worth your time, and I think that the chances of there being a productive outcome for you or them isn't very high.

I mean, it's the hiring company's job to vet you, not yours. The requirements are to provide you with some guidelines to avoid having you waste time. If you think you can do the job, I'd go ahead and apply. They're gonna try and get the best fit candidate from those that apply, regardless. If they had more-specific requirements, like knowledge of some specific software package, they could have included it in the job requirements. I wouldn't over-analyze it.

If you're concerned about it, every place I've ever interviewed at has had someone who is supposed to take questions from the candidate at the end of the interview. You can probably ask them there if there's a specific set of things on Linux that it'd be useful to know.

EDIT: And as someone who has done plenty of software development work, if someone just put down "Linux proficiency" and expected it to be interpreted without additional context as having some specific background in software development, I'd be surprised. But my larger point is that I don't think that I'd fret about it.

I mean, you can build a render farm on a single Raspberry Pi if you want, technically.

But the the requirements for a server that “does it all” remains a mystery to me.

"All" can include anything. I mean, you can include a home parallel compute render farm that will cost millions of dollars.

You're going to have to narrow it a bit down. You can have people maybe suggest some of the things that they use their systems for. Maybe it's hosting services for a cell phone that some people use cloud-based services for. Maybe it's home automation. Maybe it's a webserver. Maybe it's AI image generation.

EDIT: To put it another way, a self-hosted server is just a computer, often without a monitor and keyboard directly attached, that you have in your physical possession. The range of things that that might be used for and capabilities it might have is really broad. It's like saying "I want a vehicle. What is a vehicle that can do everything?" I mean, that might be a bicycle or a three-trailer road train, depending upon what you're going for.

FreeCAD exposes a Python console as an end-user feature. It has a macro recording system for automating repetitive tasks, much like MS Office does, it uses Python as a scripting language. Can you show me an API reference for this feature?

kagis

https://freecad.github.io/SourceDoc/modules.html

I want to write a macro that will insert some text into the cell of a spreadsheet I have selected. Click a cell, click the macro button, and it puts some text into that cell. It can do this. There are macros published that do this kind of thing. Show me where in their published documentation the functions necessary to do that are described.

I don't use FreeCAD, don't have any familiarity with this spreadsheet functionality, but let's look.

kagis

They appear to have a Doxygen API reference for their spreadsheets here:

https://freecad.github.io/SourceDoc/d0/da8/classSpreadsheet_1_1Sheet.html

setCell() looks like it sets a cell value to me.

That appears to take a CellAddress, which it looks like is obtained via getCellAddress(). As I said, I haven't used FreeCAD's spreadsheets, but I expect that it has some cell-addressing syntax akin to spreadsheets that I've used, and that one passes the name in to that, same as one would if referencing the cell in a formula to compute another cell's value. I've no idea if FreeCAD's syntax is the same as Excel's, or if it differs, though, and I'm not going to look that up; that shouldn't be an API-specific issue.

It also looks like it has a macro-recording feature. It looks like, to my quick skim, that natively generates Python:

https://wiki.freecad.org/Macros

You can also directly copy/paste python code into a macro, without recording GUI action.

And looking at the source of an arbitrarily-chosen macro, it appears to be in Python, rather than some app-specific macro language:

https://wiki.freecad.org/Macro_Rotate_View

So I expect that you can most-likely just record yourself performing the operation and the macro-recording functionality will give you the code without you needing to write something.

EDIT: It sounds like you want to get the selected cells rather than specifying the to-be-modified cell by name, which it looks like SheetTableView::selectedRanges() provides you with; Range objects appear to provide for a selection including many cells, but if you only want, say, the starting cell (which would presumably be the case for a single cell selection), then it looks like Range::from() provides that, since the starting cell would be the same cell as the only cell in a selection if there's a single-cell selection. That returns a CellAddress as well.

I mean, they may not have as much as you want, but they do appear to have documentation.

https://wiki.freecad.org/

That has a manual in two ebook formats, PDF, and a French and Italian translation.

It has three help sections on the wiki, for each of users, power users, and developers.

And it apparently has some in-application help functionality.

And there's a help forum.

EDIT: It also looks like some people have written books about using FreeCAD.

Some of the "old" devices people are listing here with an integrated battery may no longer be functional, if they've been left unused and discharged long enough.

LLMs have no more beliefs than a parrot does.

Less. A parrot can believe that it's going to get a cracker.

You could make an AI that had that belief too, and an LLM might be a component of such a system, but our existing systems don't do anything like that.

Solar and wind, sorry.

Second, yes they are intermittent but that's not an argument in favour of nuclear.

Sure it is. It's just not an argument in favor of using nuclear as a peaking source to fill in the gaps for solar and wind intermittency.

Air and wind are inexpensive insofar as they have a low LCOE, but are intermittent, so require being coupled with energy storage, and that is not inexpensive.

If you're talking hydropower or geothermal, then they don't have the intermittency issue (well, hydro does, but to a far lesser degree), but both are subject to the geography of the area. They aren't available to everyone.

EDIT: And in the case of hydropower, there are also some environmentalists unhappy about the impact on river systems, since dams inevitably have at least some impact on river ecosystems, even if you build those fish channels.

EDIT2: "Fishway" or "fish ladder".

EDIT3: In fairness, for some uses, intermittency isn't such a big issue. That is, you may have an industrial process that you can only run when energy is available. So, for example, the Netherlands used to do this (sans electricity) with their windpumps in the process of poldering. That's not free

if you want your pumps to run only a third of the time on average, then you need triple the pumping capacity

but for some things like that, where the process is basically the pumping side of pumped hydrostorage, it might be cheaper than providing constant operation with a non-intermittent power source.

But for an awful lot of uses, people just want electricity to be available when they flip the switch.

Yeah, I kind of think that that should go into core curriculum in school, because it's such a mechanical process, yet people just need to figure it out on their own.

At least in a business context, the vast majority of emails that I see sent out are mostly useless fluff. Many of them don't need to be sent, and the ones that do are rarely concise or structured to summarize what they are saying up top, then later go into detail for people who might need more detail.

Time is a finite resource consumed by this, and there's no penalty for using someone else's. Businesses don't, say, try to assess the business cost imposed by an employee's sent emails when reviewing that employee's performance.

I think that users attempt to compensate by committing less time to reading them. Doing ever-more-perfunctory skims in an attempt to limit how much of their time gets consumed by email that isn't worthwhile.

And that tends to encourage not fully reading emails.