Print the most recent available version of packages being upgraded.

Print the most recent available version of packages being upgraded.

If a local dependency isn't upgraded to the most recent available
version, pub will now indicate what that most recent version is in its
"pub upgrade" output.

R=rnystrom@google.com
BUG=16020
Committed: https://code.google.com/p/dart/source/detail?r=37924

[Bob Nystrom, 2014-07-01 17:04:41]

https://codereview.chromium.org/366513002/diff/1/sdk/lib/_internal/pub/lib/sr...
File sdk/lib/_internal/pub/lib/src/solver/solve_report.dart (right):

https://codereview.chromium.org/366513002/diff/1/sdk/lib/_internal/pub/lib/sr...
sdk/lib/_internal/pub/lib/src/solver/solve_report.dart:234:
"${pluralize('version', newerUnstable)} available through $latest)";
I find the "through ___" hard to make sense of. How about:

1. If there's one newer version available (stable or unstable) show: "(1.2.3
available)"
2. Otherwise, show: "(1.2.3 and 2 other newer [unstable] versions available)"

[nweiz, 2014-07-01 21:47:42]

https://codereview.chromium.org/366513002/diff/1/sdk/lib/_internal/pub/lib/sr...
File sdk/lib/_internal/pub/lib/src/solver/solve_report.dart (right):

https://codereview.chromium.org/366513002/diff/1/sdk/lib/_internal/pub/lib/sr...
sdk/lib/_internal/pub/lib/src/solver/solve_report.dart:234:
"${pluralize('version', newerUnstable)} available through $latest)";
On 2014/07/01 17:04:41, Bob Nystrom wrote:
> I find the "through ___" hard to make sense of. How about:
> 
> 1. If there's one newer version available (stable or unstable) show: "(1.2.3
> available)"
> 2. Otherwise, show: "(1.2.3 and 2 other newer [unstable] versions available)"

That reads like you're saying there are 2 versions available that are newer than
1.2.3.

Do we need to include the number of newer versions at all? It doesn't seem super
relevant. The distance between the locked version and the latest version
provides a better metric of how behind the locked version is anyway.

[Bob Nystrom, 2014-07-01 22:01:56]

https://codereview.chromium.org/366513002/diff/1/sdk/lib/_internal/pub/lib/sr...
File sdk/lib/_internal/pub/lib/src/solver/solve_report.dart (right):

https://codereview.chromium.org/366513002/diff/1/sdk/lib/_internal/pub/lib/sr...
sdk/lib/_internal/pub/lib/src/solver/solve_report.dart:234:
"${pluralize('version', newerUnstable)} available through $latest)";
On 2014/07/01 21:47:41, nweiz wrote:
> On 2014/07/01 17:04:41, Bob Nystrom wrote:
> > I find the "through ___" hard to make sense of. How about:
> > 
> > 1. If there's one newer version available (stable or unstable) show: "(1.2.3
> > available)"
> > 2. Otherwise, show: "(1.2.3 and 2 other newer [unstable] versions
available)"
> 
> That reads like you're saying there are 2 versions available that are newer
than
> 1.2.3.

Hmm, good point.

> Do we need to include the number of newer versions at all? It doesn't seem
super
> relevant. The distance between the locked version and the latest version
> provides a better metric of how behind the locked version is anyway.

That's true. I'd be fine with leaving off the count.

What do you think about: "(behind 1.2.3 and 3 other versions)"?

[nweiz, 2014-07-01 22:25:43]

code review

[nweiz, 2014-07-01 22:25:58]

https://codereview.chromium.org/366513002/diff/1/sdk/lib/_internal/pub/lib/sr...
File sdk/lib/_internal/pub/lib/src/solver/solve_report.dart (right):

https://codereview.chromium.org/366513002/diff/1/sdk/lib/_internal/pub/lib/sr...
sdk/lib/_internal/pub/lib/src/solver/solve_report.dart:234:
"${pluralize('version', newerUnstable)} available through $latest)";
On 2014/07/01 22:01:56, Bob Nystrom wrote:
> On 2014/07/01 21:47:41, nweiz wrote:
> > On 2014/07/01 17:04:41, Bob Nystrom wrote:
> > > I find the "through ___" hard to make sense of. How about:
> > > 
> > > 1. If there's one newer version available (stable or unstable) show:
"(1.2.3
> > > available)"
> > > 2. Otherwise, show: "(1.2.3 and 2 other newer [unstable] versions
> available)"
> > 
> > That reads like you're saying there are 2 versions available that are newer
> than
> > 1.2.3.
> 
> Hmm, good point.
> 
> > Do we need to include the number of newer versions at all? It doesn't seem
> super
> > relevant. The distance between the locked version and the latest version
> > provides a better metric of how behind the locked version is anyway.
> 
> That's true. I'd be fine with leaving off the count.
> 
> What do you think about: "(behind 1.2.3 and 3 other versions)"?

I like it, but the more I think about it the more I like the terseness of just
leaving off the number entirely.

[Bob Nystrom, 2014-07-01 23:45:35]

On 2014/07/01 22:25:58, nweiz wrote:
>
https://codereview.chromium.org/366513002/diff/1/sdk/lib/_internal/pub/lib/sr...
> File sdk/lib/_internal/pub/lib/src/solver/solve_report.dart (right):
> 
>
https://codereview.chromium.org/366513002/diff/1/sdk/lib/_internal/pub/lib/sr...
> sdk/lib/_internal/pub/lib/src/solver/solve_report.dart:234:
> "${pluralize('version', newerUnstable)} available through $latest)";
> On 2014/07/01 22:01:56, Bob Nystrom wrote:
> > On 2014/07/01 21:47:41, nweiz wrote:
> > > On 2014/07/01 17:04:41, Bob Nystrom wrote:
> > > > I find the "through ___" hard to make sense of. How about:
> > > > 
> > > > 1. If there's one newer version available (stable or unstable) show:
> "(1.2.3
> > > > available)"
> > > > 2. Otherwise, show: "(1.2.3 and 2 other newer [unstable] versions
> > available)"
> > > 
> > > That reads like you're saying there are 2 versions available that are
newer
> > than
> > > 1.2.3.
> > 
> > Hmm, good point.
> > 
> > > Do we need to include the number of newer versions at all? It doesn't seem
> > super
> > > relevant. The distance between the locked version and the latest version
> > > provides a better metric of how behind the locked version is anyway.
> > 
> > That's true. I'd be fine with leaving off the count.
> > 
> > What do you think about: "(behind 1.2.3 and 3 other versions)"?
> 
> I like it, but the more I think about it the more I like the terseness of just
> leaving off the number entirely.

I think the number of other versions is also a useful signal (since different
packages tend to rev their versions differently), but do what you like.

[nweiz, 2014-07-02 00:13:25]

Committed patchset #2 manually as r37924 (presubmit successful).