1 |
William Hubbs posted on Fri, 31 Jul 2015 16:43:48 -0500 as excerpted: |
2 |
|
3 |
> On Fri, Jul 31, 2015 at 10:27:02PM +0200, eroen wrote: |
4 |
>> On Fri, Jul 31, 2015 at 02:49:08PM -0500, William Hubbs wrote: |
5 |
>> > I'm still looking for ideas of ways to avoid making this a breaking |
6 |
>> > change. If it isn't possible to make it a non-breaking change, this |
7 |
>> > will force openrc-1.0 to be the first release where we can do this |
8 |
>> > (going from 0.x to 1.x signifies breaking changes in semantic |
9 |
>> > versioning [1]). |
10 |
>> |
11 |
>> > [1] http://www.semver.org |
12 |
>> |
13 |
>> This is a valid and possible versioning policy, but it is not semver. |
14 |
>> Your [1] states: |
15 |
>> |
16 |
>> 4. Major version zero (0.y.z) is for initial development. Anything may |
17 |
>> change at any time. The public API should not be considered stable. |
18 |
> |
19 |
> The problem is that version 0 hit stable before I knew about semver, so |
20 |
> I have to treat version 0 as version 1. |
21 |
> |
22 |
> I can't just randomly break things from 0.17 to 0.18 for example. |
23 |
|
24 |
I've seen others in that predicament do a much larger jump, say to 0.80+ |
25 |
or adding a digit. The atypically large jump seems to signify a broken |
26 |
API well enough. |
27 |
|
28 |
So what about jumping from 0.17 to 0.100, or possibly to 0.80 or 0.90, if |
29 |
you intend to jump to 1.0 reasonably soon, since x.8y and x.9y are |
30 |
reasonably often used for betas and rcs, respectively. |
31 |
|
32 |
-- |
33 |
Duncan - List replies preferred. No HTML msgs. |
34 |
"Every nonfree program has a lord, a master -- |
35 |
and if you use the program, he is your master." Richard Stallman |