upvote

points

by cocoto19 hours ago |
comments
upvoteby FabCH19 hours ago|
[-]
Comparing is slightly complicated but roughly speaking:

Difftastic only diffs code, which is only 2/3 of all diffs.

When only measured on stuff difftastic supports:

Difftastic is perfect on 64% of diffs and near-perfect on 17% more.

Omnidiff is perfect on 90% of diffs and near-perfect on 5% more.

Speed is

p50 - 35ms difftastic vs 4ms OmniDiff p99 - 637ms vs 420ms

Difftastic wins on license. MIT vs AGPL.

reply
upvoteby LVB17 hours ago|
parent|
[-]
> Difftastic wins on license. MIT vs AGPL.

I'm a little confused on that, since in other comments you've mentioned that AGPL was a deliberate choice for <reasons>. So, what's the definitions of "winning" here?

reply
upvoteby FabCH17 hours ago|
parent|
next
[-]
It was a tongue-in-cheek joke :)

But it was worth pointing out the difference in license, it is one of the ways in which the projects differ.

reply
upvoteby folmar11 hours ago|
parent|
prev|
[-]
The more permissive MIT is often an advantage if you want to build it into something bigger, which is reasonable for diff tool.

Discrete users would arguably win more with integrity of AGPL, but developers have reasons to prefer MIT.

reply
upvoteby fishgoesblub9 hours ago|
parent|
next
[-]
> but developers have reasons to prefer MIT

Hi! as a developer, I personally prefer it when the libraries/tools used in my projects are GPL or another Copyleft license. The drove of MIT projects from the past decade has been depressing.

reply
upvoteby lejalv3 hours ago|
parent|
prev|
[-]
Servitude to their masters
reply