[Haskell-cafe] Platform Versioning Policy: upper bounds are not our friends

[MigMit]

On Aug 18, 2012, at 12:35 AM, "Bryan O'Sullivan" <bos at serpentine.com> wrote:

> We already have a simple versioning scheme for which, despite it being easy to grasp, we have amply demonstrated that we cannot make it work well, because it has emergent properties that cause it to not scale well across a large community.

Well, I think that the main reason for this failure is that despite being easy to grasp, this scheme doesn't really reflect the reality. It seems to be chosen arbitrarily.

> Any vastly more complicated and detailed versioning scheme has a huge burden to prove that it won't collapse dramatically more quickly. (Frankly, I think that anything involving "specify every detail of your known dependencies" is dead on arrival from a practical standpoint: it's way too much work.)

That's not true. All this work can be automated — if you're the developer, you've certainly compiled the code yourself, and the compiler knows what was imported from other packages. The only problem is to make it save this information into a specific file.