Mailing-List: contact issues-help@maven.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@maven.apache.org
Received-SPF: pass (nike.apache.org: local policy)
Date: Wed, 21 Nov 2012 01:07:13 -0600 (CST)
From: "Herve Boutemy (JIRA)" <jira@codehaus.org>
To: issues@maven.apache.org
Message-ID: 
 <1835706059.12117.1353481633444.JavaMail.j2ee-jira@codehaus01.managed.contegix.com>
In-Reply-To: 
 <450871798.4197.1349051376217.JavaMail.j2ee-jira@codehaus01.managed.contegix.com>
Subject: [jira] (MNG-5353) Ignore pre-releases in exclusive upper bound
 [lw,up)
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit


    [ https://jira.codehaus.org/browse/MNG-5353?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=314069#comment-314069 ] 

Herve Boutemy edited comment on MNG-5353 at 11/21/12 1:05 AM:
--------------------------------------------------------------

IMHO, not really a filter, since a filter can let "holes" in a mathematical range, but here, there won't be any holes: just the upper bound that is lower than previously

for the purpose of demonstration, I'll note [lw,up( the new notion, to look different from [lw,up)
given the comparisaon algorithm, [lw,up( = [lw,up-alpha-alpha-alpha-alpha-...infinite...)
Since nobody creates the alpha of an alpha, [lw,up( is not that far from [lw,up-alpha-alpha)

IMHO, if you simply add "-a-a" to upper bound version and let actual range calculation algorithm, you have the solution for every real world situation: only people knowing the approximation will try with "up-a-a-1" or "up-a-rc-1" and show you the problem (you can add "-a-a-a" to be even more tricky to get fooled)
                
      was (Author: hboutemy):
    IMHO, not really a filter, since a filter can let "holes" in a mathematical range, but here, there won't be holes: just the upper bound that is lower than previously

for the purpose of demonstration, I'll not [lw,up( the new notion, to look different from [lw,up)
givne the comparisaon algorithm, [lw,up( = [lw,up-alpha-alpha-alpha-alpha-...)
Since nobody creates the alpha of an alpha, [lw,up( is not that far from [lw,up-alpha-alpha)

IMHO, if you simply add "-a-a" to upper bound version and let actual range calculation algorithm, you have the solution for every real world situation: only people knowing the approximation will try with "up-a-a-1" or "up-a-rc-1" and show you the problem (or take "-a-a-a" to be even more tricky to get fooled)
                  
> Ignore pre-releases in exclusive upper bound [lw,up)
> ----------------------------------------------------
>
>                 Key: MNG-5353
>                 URL: https://jira.codehaus.org/browse/MNG-5353
>             Project: Maven 2 & 3
>          Issue Type: Wish
>          Components: Dependencies
>    Affects Versions: 2.0.11, 2.2.1, 3.0.4
>            Reporter: Jesse Long
>            Assignee: Jason van Zyl
>             Fix For: 3.1.1
>
>
> Please change the behaviour of exclusive upper bounds to the following:
> In a version range, like [1.7.0,1.8.0), 1.8.0-alpha1 should not be included in the range. 1.8.0* should not be included in the range.
> This allows for us configure natural ranges for projects using semantic versioning.
> Please see:
> http://markmail.org/message/4tubas4uok6ahbcp
> http://markmail.org/message/s2ry2uru4ibub43q

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://jira.codehaus.org/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira