Lennart Spitzner

Fixed one bug i found when `indentAmount>4`, fixed the "let" issue (hopefully), cleaned up, merged into master, prepared release of `0.11`. @chreekat if you want, check that all works as...

I see, that is a bit unfortunate. However it does not seem to be connected to "multibyte" aspect, but rather to the fact that even with monospace fonts, the glyph...

Yeah, i know. Pattern layouting is not complete for any non-single-line cases. And unfortunately this has somewhat low priority for me, as such patterns usually are best avoided. Although I...

I do agree that it is a separate issue, but I cannot help but mention that I ran into (an equivalent of) this while implementing #281. Heh. I still need...

I have looked into this. There is a good amount of special syntax involved (`HsCmd`), and even if most is very close to `HsExpr` stuff it would require some time...

Thanks for the report. I can reproduce. On master this is down to a warning by the power of the errors-fall-back-on-exactprint feature, but that simply leaves it non-reformatted. It is...

There is another approach to this: For certain nodes, non-bottom-spacing modifier could be adapted to also apply a `take 1` to the spacings. If applied carefully, this won't have any...

I have added [a section to the brittany rationale](http://hexagoxel.de/postsforpublish/posts/2017-10-28-brittany-rationale.html#layout-of-module-imports) that explains the current behaviour, and highlights the trade-off of this decision. The "context-sensitive" term [is introduced above](http://hexagoxel.de/postsforpublish/posts/2017-10-28-brittany-rationale.html#context-sensitivity-versus-least-newlines). As mentioned there,...

Thanks! I agree, and your preferred formatting looks good to me. We just have not come around to implementing anything in this direction, because patterns can appear in so many...

Yeah, some buggy/far-from-optimal layouting is clearly going on there. I will ask nonetheless: What is the expected/desired layout for this example?