Re: port release candidate versioning question

2020-08-17 Thread Arno Hautala
> On 17 Aug 2020, at 17:22, Ryan Schmidt wrote: > > What is the benefit to anyone in being able to presume (but of course not to > know for certain without inspecting the git log/blame) that the epoch was > last changed in the port 15 years ago? > > I updated the guide 9 years ago with the r

Re: port release candidate versioning question

2020-08-17 Thread Ryan Schmidt
> On Aug 17, 2020, at 15:11, Arno Hautala wrote: > > >> On 17 Aug 2020, at 15:40, Ryan Schmidt wrote: >> >> We do not recommend setting the epoch to a date-like value. Simply use an >> increasing integer value. Its default value is 0. The first time you need to >> increase it, set it to 1;

Re: port release candidate versioning question

2020-08-17 Thread Arno Hautala
> On 17 Aug 2020, at 15:40, Ryan Schmidt wrote: > > We do not recommend setting the epoch to a date-like value. Simply use an > increasing integer value. Its default value is 0. The first time you need to > increase it, set it to 1; the second time, set it to 2, etc. Is there any specific re

Re: port release candidate versioning question

2020-08-17 Thread Ken Cunningham
thanks to all for the helpful responses. I will try to get it right off the mark. K > On Aug 17, 2020, at 12:40, Ryan Schmidt wrote: > > > >> On Aug 17, 2020, at 13:39, Arno Hautala wrote: >> >>> On 17 Aug 2020, at 14:35, Arno Hautala wrote: >>> >>> To verify this, and to test other cases,

Re: port release candidate versioning question

2020-08-17 Thread Ryan Schmidt
On Aug 17, 2020, at 13:39, Arno Hautala wrote: > On 17 Aug 2020, at 14:35, Arno Hautala wrote: > >> To verify this, and to test other cases, you can try the ‘vercmp’ script > > I should probably have included a link to the script, instead a test of the > comparisons… > > https://svn.macport

Re: port release candidate versioning question

2020-08-17 Thread Arno Hautala
> On 17 Aug 2020, at 14:35, Arno Hautala wrote: > > To verify this, and to test other cases, you can try the ‘vercmp’ script I should probably have included a link to the script, instead a test of the comparisons… https://svn.macports.org/repository/macports/users/ryandesign/scripts/vercmp

Re: port release candidate versioning question

2020-08-17 Thread Blair Zajac
> On Aug 17, 2020, at 11:35 AM, Arno Hautala wrote: > >> On 17 Aug 2020, at 13:28, Ken Cunningham >> wrote: >> >> ie does macports see versions like this: >> >> 11.0.0 > 11-rc3 > 11-rc2 > 11-rc1 ? > > > My understanding is that your example will indeed sort correctly. To verify > this,

Re: port release candidate versioning question

2020-08-17 Thread Arno Hautala
> On 17 Aug 2020, at 13:28, Ken Cunningham > wrote: > > ie does macports see versions like this: > > 11.0.0 > 11-rc3 > 11-rc2 > 11-rc1 ? My understanding is that your example will indeed sort correctly. To verify this, and to test other cases, you can try the ‘vercmp’ script [1] [1]: https

Re: port release candidate versioning question

2020-08-17 Thread Christopher Chavez
Sent: Monday, August 17, 2020 at 12:28 PM From: "Ken Cunningham" To: "MacPorts Developers" Subject: port release candidate versioning question > This sounds like a dumb question, that I should already know the answer to, > but apparently I don’t seem to: > I would

port release candidate versioning question

2020-08-17 Thread Ken Cunningham
This sounds like a dumb question, that I should already know the answer to, but apparently I don’t seem to: I would like to make sure this works the way I think it works before I do it. When I make a new port clang-11 with version 11-rc1 and then follow-up with versions 11-rc2, 11-rc3, etc and