Now that is different.

Access gets used for a shared DB and that is quite easy to corrupt. It is much more cost effective to have that in a proper central database (I supse SQLLite is better here as well)

Excel is also a shared DB: it has supported multiple concurrent users accessing and modifying the same spreadsheet for decades.

ANd that gets rather looked on here as the authors being deceitful and not really Open Source doing a bait and switch.

I've been working on a software package I'm hoping to release in a few months... I'm really torn on either split FLOSS with commercial extensions, or just going fully private... I was planning on a pretty generous free tier, but hoping to make a bit on the side from commercial customers.

It's a bit of a niche as it is, so that's going to be rough in any kind of pricing model, as a large part of that niche is either homebrew types and the other commercial industry that will likely require some more integrations and customization.

You could dual license as well, so it’s GPL or AGPL for personal, OSS, or academic use, but requires a paid for commercial license for commercial use.

I suggest GPL or AGPL because their copyleft clauses make them hostile towards platform providers who might otherwise seek to profit from your work without paying.

Platform providers can take (A)GPL code as-is and totally profit from it without paying.

As long as they keep releasing sources of their own modifications -if they ever do any-, the rest is fair play.

Yeah, but the copyleft makes anything they build around it a derivative work that they also have to release sources for - especially with AGPL. Most don’t want to do that because that’s where their IP lives.

Not all open source licenses are copyleft licenses (e.g., MIT very much isn’t), but at the very least copyleft licenses make it much harder to exploit open source code commercially without giving back in some way, whether that’s code, or cash for a commercial license.

Not perfect, by any means, but definitely an improvement over more permissive licenses.

I am aware of how much I’m starting to sound a bit like RMS in my old age.

I wholeheartedly agree. Licensing is a complex topic of which I've read a good deal, and even within the Open Source communities there are usually a lot of misconceptions, so I like chiming in with less commonly pointed but very practical effects of it all, in case it helps someone to learn a tiny bit that day.

In this case the provider would of course have to comply with the AGPL and release their modifications as you mention, but it's important to note that No FOSS license protects at all against, for example, just offering the code as a service. It's the exact reason why Mongodb changed licenses and then a stream of commercial products started to change into "Source-Available" licenses in the recent past.

It would be dual license effectively... the base version AGPL and the Commercial version with additional functionality. Though I'd considered BSL and alternatives... and as mentioned, just closed/commercial only.

Or for Lisp you might as well start with Emacs Lisp - you are going to use it for a decent environment unless you have the Common Lisp IDEs which you have to pay for or Racket.

Eh, no. You have Elisp+cl-lib but SBCL too, and you can use Sly wth SBCL.

Of Lem with SBCL+Quicklisp:

https://lem-project.github.io/usage/common_lisp/

Huge tip: if you use MCCLIM, install Ultralisp first and (ql-quickload 'mcclim) later: it will give you a big speed boost. Big, not as the ones from Phoronix. Actually big. From 'I can almost see redrawing on a really old ass netbook' to 'snappy as TCL/Tk' under SBCL.

https://ultralisp.org/

As you can see, you don't need to pay thousands of dollars.

For Scheme, S9 just targets R4RS but as a start it's more than enough, and for SICP you can install Emacs+Geiser+chicken Scheme and from any Linux/BSD: distro command prompt, you run:

       sudo  chicken-install srfi-203
       sudo chicken-install srfi-216

And, as a ~/.csirc file:

      (import scheme)
      (import (srfi 203))
      (import (srfi 216))

To run SCM stuff for SICP:

       csi yourfile.scm 

or

       chicken-csi yourfile.scm

Done. Get the SICP PDF and start doing SICP. You can use Emacs+Geiser with

         M-x install RET geiser-chicken

 if you are lazy. You can install the SICP book with

        M-xpackage-install RET sicp

and read it from

       M-x info RET

and do it everything from withing Emacs by running

      M-x geiser 

(pick chicken as the interpreter). Save your Emacs settings. Done.

The problem is that the social media companies have not been dealing with abusive posts of various sources. Governments can't take action against the bad posters are they are from another Government (and in some cases are employed by that government to cause trouble). Thus Governments have to take actions which they can control, unfortunately these actions will affect more than the bad abusers.

You assume your premise. No the government actually doesn't 'have to' take action about mean things on the internet. The UK has such an obsession with regulating what is, essentially, politeness.

While I don't particularly care for the UK's approach to these things, I can't help but be shocked at how many governments seem to all of a sudden have dreamed up the same idea. Independently, I'm sure.

I suppose the US is the unique one really, when it comes to a history of protecting certain types of speech. They've never really regulated (what I would call) politeness between people in any form.

The UK, and I assume much of Europe, criminalizes truly petty levels of speech. For example, it's illegal to insult someone and cause them 'alarm' or 'distress' in the street.

Thus the non-technical populace see rudeness on the internet as the result of some kind of wild west situation that the government needs to control, to bring it in line with the rest of the public realm.

This should be made a problem for the social media companies (which it largely has, hence all the age verification fiasco), not absolutely everyone on the internet.

MacOS requires there to be an App as First Responder - what app should be that if you quit all apps?

The nearest that makes sense is Log Out.

Yes but only comp sci ones do - and if you are doing that then you need more than a base computer.

Students in non STEM areas will not usually need more than writing and reading tools.

And you get the under reporting of COVID information in the UK as they passed around CSV files with too many rows for the tools they used.

An interchnage format needs to include information showing that you have all the data - e.g. a hash or the number of rows - or JSON/XML/s-expressions having closing symbols to match the start.

If this is on Homebrew then it is on macOS.

Why not just use TRime Machine?

The article explains that.

Busess are improving many of them are now fully electric.

20% of London's busses are zero emission - agree that London is dense enough for this to work - long didtance busses still have to be diesel, although Tfl have some 15 mile routes that are electric

But when you drive you can't read.

Thus you get more done when using public transport. Nowadays with phones and portables you can even read your email and work rather than justb read the newspaper as commuterts did 20 years ago.

Yes, and that's a good thing. Because if I'm on transit and I read, I miss stops, sometimes as many as four or five. Then my day is really fucked. I literally have to sit there and count off every single stop. With driving, I don't have that problem. I have internal mapping and external GPS to remind me what I need to do next.

As for getting work done, back when I was an employee and using transit, two factors kept me from doing work on the transit system. First, my employer already got enough uncompensated labor from me. I wasn't going to give them any more. Second, I use speech recognition, and dictating company confidential information in a public setting is unwise at best.