Hi!

I think of `update-patches' as a good way to produce the necessary set
of patches for a certain package to integrate into buildroot. 
From a more abstract point of view, it's helpful when creating patches
for a single purpose, but nothing more. When patches for different
reasons come together (the kernel and busybox patches are just an
example), it's getting more and more important that those changes stay
together in their logical context. Nobody likes a single Big Patch for
everything, and having all changes to a single file separated is the
same as cutting the Big Patch around the diff-markers.

In my opinion it depends strongly on the packet which is the rightest
way to handle patches. In general, I would decide according to the
single purpose / multiple purpose system explained above.

Handling patches manually produces more overhead on one hand, but on the
other is definitely the better (as easier to understand) way of doing
it, so I hope it's clear that yet existing (not auto generated)
patchsets should not be converted to the "machine friendly" system.

Greetings, Phil

Attachment: pgptus984rcMq.pgp
Description: PGP signature

_______________________________________________
freewrt-developers mailing list
[email protected]
https://www.freewrt.org/lists/listinfo/freewrt-developers

Reply via email to