The issue I've had is not just the hundred upgrades, but also that as time passes, it may not even be feasible to solve the dependency constraints for bumping a single package without a huge amount of archaeological work.
For example, you might have lib A and lib B that need to have version bumps more or less in lockstep, and if you're lucky they even make major-version releases to alert you to this. If those are *recent* changes, you can figure this out very quickly. But if there have been 30 releases of everything since then, and the same thing happens again and again with various values for "A" and "B", you get this dependency hell situation.
BUT, if you could say "just upgrade everything to latest as of 2022-04-07", then you would not only have smaller steps to take, but also a vastly reduced conflict set to solve. Then just repeat with time increments as appropriate.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
(no subject)
Date: 2025-01-05 05:46 pm (UTC)For example, you might have lib A and lib B that need to have version bumps more or less in lockstep, and if you're lucky they even make major-version releases to alert you to this. If those are *recent* changes, you can figure this out very quickly. But if there have been 30 releases of everything since then, and the same thing happens again and again with various values for "A" and "B", you get this dependency hell situation.
BUT, if you could say "just upgrade everything to latest as of 2022-04-07", then you would not only have smaller steps to take, but also a vastly reduced conflict set to solve. Then just repeat with time increments as appropriate.