search

haskell-streaming / streaming-bytestring

effectful sequences of bytes; an alternative no-lazy-io implementation of Data.ByteString.Lazy
BSD 3-Clause "New" or "Revised" License
16 stars 11 forks source link

next version must be a majour version bump #35

Closed chessai closed 3 years ago

chessai commented 3 years ago

because we are deprecating Data.ByteString. and the ByteString type, the next version bump must be majour, meaning we can no longer stay on 0.1.. Then, after a suitable deprecation period, there would be another majour version bump which removes the deprecated modules.

vdukhovni commented 3 years ago

If this a major version bump, perhaps some outstanding nits that require interface changes to fix can be addressed? If so, this is a good time to go over the API and hunt down the warts...

fosskers commented 3 years ago

I wouldn't mind, if they are small things.

vdukhovni commented 3 years ago

I wouldn't mind, if they are small things.

I'll do my best to identify the appropriate candidates. The first step might be #39 (not even a compatibility issue, since we're making the input type more polymorphic).

vdukhovni commented 3 years ago

But do you have any comments about #40? That might require an API change if the idea is to make the Char8 and Word8 interfaces consistent...

vdukhovni commented 3 years ago

OK, #42 is my take on the requisite changes. Please review.

But #42 does not subsume #40, or the other issues I opened, so feedback on those would be great. (But it does take care of #37).

vdukhovni commented 3 years ago

Before the major version bump, it may make sense to release 0.1.8 with #43 (sorry about the bug, one edge case the tests did not catch). Code coverage does not complain about incorrectly ignored _ variables... :-(

fosskers commented 3 years ago

I think it would be okay to just go ahead with 0.2 when your fix is included, since apparently it was my mistake to have only bumped the minor version.

vdukhovni commented 3 years ago

Yeah, I wondered about that. Works for me at this point.

fosskers commented 3 years ago

Done.

In 6 - 12 months when we dump the deprecations, I suggest we move to 1.0.0 to celebrate the overhauled module paths.