Metacello version comparison

> I would like your take on the following question about version comparison. I don't think the following issues  actually affect any existing configurations.
>
> Today, when comparing a version string, the version '1.0' compares less than '1.0.0'. I have explicit tests, so at one point in time I thought that that made sense, but today I think that version '1.0' and version '1.0.0' should compare equal.
>
> By extension version '1' and and version '1.0' are equal.
>
> I think that trailing 0 terms in a version sequence should be collapsed before comparison.
>
> A sequence of 0's that ends in a numeric version element should not be collapsed. So version '1.0' is still less than '1.0.1'. No 0 collapsing is done for version '1.0.1'.
>
> A sequence of 0's that ends in a string version element should be collapsed. So version '1.0.0-beta.1' is equal to version '1-beta.1'.
>
> Thoughts ... do you care?:)
>
> Dale

> On Oct 6, 2010, at 11:13 AM, Dale Henrichs wrote:
>
>> I would like your take on the following question about version comparison. I don't think the following issues  actually affect any existing configurations.
>>
>> Today, when comparing a version string, the version '1.0' compares less than '1.0.0'. I have explicit tests, so at one point in time I thought that that made sense, but today I think that version '1.0' and version '1.0.0' should compare equal.
>>
>> By extension version '1' and and version '1.0' are equal.
>>
>> I think that trailing 0 terms in a version sequence should be collapsed before comparison.
>>
>> A sequence of 0's that ends in a numeric version element should not be collapsed. So version '1.0' is still less than '1.0.1'. No 0 collapsing is done for version '1.0.1'.
>>
>> A sequence of 0's that ends in a string version element should be collapsed. So version '1.0.0-beta.1' is equal to version '1-beta.1'.
>>
>> Thoughts ... do you care?:)
>>
>> Dale
>
> Makes sense to me.
>
> Pat

> I would like your take on the following question about version
> comparison. I don't think the following issues  actually affect any
> existing configurations.
>
> Today, when comparing a version string, the version '1.0' compares less
> than '1.0.0'. I have explicit tests, so at one point in time I thought
> that that made sense, but today I think that version '1.0' and version
> '1.0.0' should compare equal.
>
> By extension version '1' and and version '1.0' are equal.
>
> I think that trailing 0 terms in a version sequence should be collapsed
> before comparison.
>
> A sequence of 0's that ends in a numeric version element should not be
> collapsed. So version '1.0' is still less than '1.0.1'. No 0 collapsing
> is done for version '1.0.1'.
>
> A sequence of 0's that ends in a string version element should be
> collapsed. So version '1.0.0-beta.1' is equal to version '1-beta.1'.
>
> Thoughts ... do you care?:)
>
> Dale

> I would like your take on the following question about version
> comparison. I don't think the following issues  actually affect any
> existing configurations.
>
> Today, when comparing a version string, the version '1.0' compares less
> than '1.0.0'. I have explicit tests, so at one point in time I thought
> that that made sense, but today I think that version '1.0' and version
> '1.0.0' should compare equal.
>
> By extension version '1' and and version '1.0' are equal.
>
> I think that trailing 0 terms in a version sequence should be collapsed
> before comparison.
>
> A sequence of 0's that ends in a numeric version element should not be
> collapsed. So version '1.0' is still less than '1.0.1'. No 0 collapsing
> is done for version '1.0.1'.
>
> A sequence of 0's that ends in a string version element should be
> collapsed. So version '1.0.0-beta.1' is equal to version '1-beta.1'.
>
> Thoughts ... do you care?:)
>
> Dale