Mold 1.1.1 Released With Optimized Memory Usage, New Options

[bug77]

Semver is a gimmick for codemonkeys, forget about it in the real world, where you actually need to read the goddamn changelog.

It can still make the decisions about what to deploy/upgrade a lot easier. Not that you should ever ignore the changelog.

[onlyLinuxLuvUBack]

The name is puzzling, why didn't they go with turbo_LD ?

[Sin2x]

It can still make the decisions about what to deploy/upgrade a lot easier. Not that you should ever ignore the changelog.

There is not law requiring developers to use semver and the developer community is anarchic at its heart, so it simply doesn't work despite being out there for a long time.

The name is puzzling, why didn't they go with turbo_LD ?

Pun is fun

[arQon]

(in case anyone asks why one would run a 32bit system nowadays, I did until recently and the reason is inertia - the system was working, so why mess with it)

For developers, which is mold's actual audience, there are a LOT more 32-bit systems still in play than most of even the "informed" users (i.e. people who even know what 32-bit means :P) think, especially in embedded systems. While the "real" pipeline is done by cross-compiling, since an x86 workstation is just massively faster than whatever the anemic hardware of the target system is, every once in a while you do need a toolchain that runs on the target system, and when that only has *M*B of RAM to work with and no/slow swap, better tools like this certainly make a difference!