Open Source Week 9

Lefkowitz really doesn't seem to like RMS. The "Free" here is specifically "Free/Libre". He regards the Gnu approach as fundamentally political.

But then Lefkowitz goes off the rails a bit, with the idea that software publishers should be liable for errors, and that should apply to FSF. Holding open-source contributors liable for bugs would have dire consequences.

And then Lefkowitz suggests that software privacy and security would be better addressed by government regulation. There is no way that is going to end well.

"The point being: large corporations and government agencies have always had the ability to get access to the source code of commercial software, and modify it if they wish. And if they have a good reason, they will do so. But usually, they don’t wish to."

Lefkowitz' idea about "public software" is, well, bizarre. Software is not like books. There's no point in a library-like check-out-and-return policy.

Yes. The business model is "open core". But thank you for yet another tedious reminder.

Earlier I raised the issue of clang vs gcc, suggesting that maybe plugins to allow gcc to interface with non-GPL IDEs was the issue. Here's another article on that: https://lwn.net/Articles/629259.

It took many years before the GNU Compiler Collection (GCC) changed its runtime library exemption in a way that allowed for GCC plugins, largely because of fears that companies might distribute proprietary, closed-source plugins.

Stallman is concerned that proprietary backends could take the [gcc abstract syntax tree] output and generate code from it.

By forbidding the editor from having advanced knowledge of the parse tree, you are intentionally crippling the ability of smart people to build better and better free software tools. You are preventing smart hackers from going in and making the system as good as they can make it, from expressing their creativity in such a way as to build the best development environment they know how.

There is also this quote from David Engster:

For almost five years now (since gcc 4.5 introduced plugins), access to GCC's AST is wide open for everyone. However, in all that time (and to my knowledge) no one has used that to feed non-free backends, and that is in my opinion enough evidence that your worries are unfounded. They might have been valid in the past, but not since LLVM and clang have joined the scene.

An interesting view of Torvald's no-regression policy

blog.farhan.codes/2018/06/25/linux-maintains-bugs-the-real-reason-ifconfig-on-linux-is-deprecated.

Specific system calls are not the only mechanism by which applications interact with the Linux kernel. There are also applications that get the information they need by using generic methods to access kernel memory data structures. When these data structures are updated, things break.

One example is the network program ifconfig.

From the article:

Linux’s solution to this problem was to create a policy of never breaking userland applications. This means userland interfaces to the Linux kernel never change under any circumstances, even if they malfunction and have known bugs. That is worth reiterating. Linux maintains known bugs – and actively refuses to fix them. In fact, if you attempt to fix them, Linus will curse at you, as manifest by this email.

Ifconfig gets its network information from kernel tables. Ifconfig is stuck on the model that each interface has only one IP address. This hasn't been true of IPv4 for a long time, and was never true for IPv6.

Ifconfig itself could have been updated. But it has not been; instead, it has been replaced by the newer ip command. The FreeBSD approach is to update the kernel and associated ifconfig-like commands simultaneously; the Linux no-regression policy doesn't allow this as then the kernel interface would have to change.

Open Source Computing

Cathedral v Bazaar

"Free Software": An idea whose time has passed [?]

Open source is NOT a business model