4 ms·
> Gentoo developer Michał Górny first suggested [dropping support for LibreSSL] at the end of December, saying that LibreSSL offers no benefit over OpenSSL at t
by kelnos 6y ago
> Gentoo developer Michał Górny first suggested [dropping support for LibreSSL] at the end of December, saying that LibreSSL offers no benefit over OpenSSL at this point while imposing a lot of costs. In particular, he complained about the large number of packages that require patches to work with LibreSSL and the constant stream of regressions that the project must deal with.
That about sums it up. The LibreSSL developers decided not to maintain even source compatibility, let alone binary compatibility. If application authors don't want the maintenance burden of supporting extra code paths for both, and distribution maintainers aren't happy keeping patchsets up to date, then that's about it.
And it sounds like OpenSSL is in much much better shape (both the code and the community around it) than when LibreSSL forked. This is a good thing! It means all the concern raised after Heartbleed actually functioned as intended, and turned OpenSSL into a healthy project.
I think it's also a good thing that LibreSSL continues to exist and that it is used in some of the BSDs. Having competing, well-maintained implementations is great, even if they're not drop-in replacements for each other. If OpenSSL falls into disrepair again, it means that an alternative is already available, even if there are switching costs.
- hda111 6y agoIt makes a lot sense that OpenBSD removed a lot of code from LibreSSL after forking it. There is only limited time and one can not maintain everything. This incompatibility is inevitable.
- j16sdiz 6y agoTo make it worse, they share lots of symbol / file path. Make it impossible to co-exist with openssl
- tottenhm 6y agoYeah, in the long run, breaking compatibility is totally understandable for a fork. And generally... power-to-them in publishing a free/open/quality SSL implementation. But... the article also says: > ... it is not possible to install both OpenSSL and LibreSSL on the same system. That combination -- not-compatible and not-coexisting -- sounds rough.
- kazinator 6y ago> Yeah, in the long run, breaking compatibility is totally understandable for a fork. Sure, in the long run, when you've overtaken the original and sent it into obsolescence. I.e. at approximately that point where even the original would have deviated from compatibility. In the short run, drop-in replace, or perish. Linux still follows POSIX. I can make a program that is source-code compatible with GNU/Linux, Solaris, Cygwin, MacOS, Android/Bionic, ... If I can't do that between two SSL implementations, something is wrong.
- mgorny 6y agoIf it only were that simple. Imagine you're maintaining some program that intensively uses OpenSSL. It's 2014-ish, and some LibreSSL people come to you and tell you that you're using deprecated API and you need to upgrade for compatibility with LibreSSL. Sure, it's an improvement and maintains compatibility with OpenSSL, so why not. Then OpenSSL 1.1 comes with some more API changes. You end up adding #if conditions to support both OpenSSL versions. And the next thing you realize, you've just broken LibreSSL because it pretends to be OpenSSL 2.0.0. And when you patch it, you realize you're adding another ticking time bomb because LibreSSL will probably support the new API at some point too and you will have to add yet another version check to the code. I am not surprised that somewhere in middle of that process upstreams stop taking LibreSSL seriously.
- tottenhm 6y agoRight, sprinkling #if's across a dozen downstream codebases sucks. Consider an analogy to shell-scripting -- there are different shells ("bash", "dash", "zsh") which have substantial overlap (insofar as they largely support traditional POSIX). Someone writing a script/program can target a specific flavor ("#!/usr/bin/env bash") and get more features, or they can write conservative code and let downstream policy determine the shell ("#!/bin/sh"). General-purpose distros (like Debian/Redhat) don't seem to have a problem with supporting them side-by-side. I'm more than a little rusty on the mechanics of dependency-management in C... but shouldn't it allow some analogous arrangement where an app either (a) signals a requirement for a specific implementation or (b) signals ambivalence/deference (and limits itself to common APIs). (This, of course, only makes sense if the developers of LibreSSL/OpenSSL and of the distros believe that it's better to coexist+compete than to consolidate. The tenor of the LWN article seems to convey a XOR mentality, but if both projects have independently healthy teams, then... maybe they prefer friendly coopetition...)
- rstuart4133 6y agoI'm not sure what LibreSSL's thinking was. During heartbleed what the world sorely lacked was a working, well maintained OpenSSL. Had LibreSSL created one of them their thoughts would have been understandable. What there was no shortage of then or now is opinionated, well maintained maintained crypto libraries. There's nothing wrong with creating a new opinionated crypto library of course, that's how progress is made, but they are like weeds. By breaking backward compatibility that is what LibreSSL created, but in doing that they threw their hat into a highly competitive ring Darwin will sort out by killing almost all of the contestants. But in the opinionated crypto library ring, LibreSSL is a weak entrant. Compared to some of it's competitors it's just a minor polish on OpenSSL, making hardly worth of effort of switching really. This outcome looked inevitable to me.