On 30 January 2013 13:41, Fred Cooke <[email protected]> wrote:

> Mark,
>
> There is no need to educate me on SVN semantics, I used it for quite a few
> years before I found git and abandoned it entirely. This is not about that.
>

You do know that GIT tags are mutable too ;-)


>
> There was an artifact with a name on a public repository free to be
> distributed as the public see fit under a specific name. Now there is a
> second one with the same name out there, in the wild. MUCH better to skip
> 31 and roll with 32 than to have two artifacts with different behaviours
> under the same exact name. Granted the chances of someone nabbing a copy in
> the mean time are low, but still, it's an unacceptable situation in my
> perfect little world, especially when it can be avoided simply by skipping
> 31 and using 32. It's not like the String the version is stored in has any
> limitations such as 9223372036854775807 for the signed 64 bit long.
> Although I doubt you'd run out even if you skipped 10, used one and used an
> unsigned short ;-)
>
> As for deleting tags from SVN being "perfectly fine", I couldn't disagree
> more. Not even slightly more.


The convention here is that until the artifacts have been pushed to
central, delete as you see fit.

Subversion tagging is just a convention.... your world has a different
convention than ours... horses for carts


> Any commit to an SVN tag is forbidden in my perfect little (dream) world
> where a name means exactly one thing and nothing more. Not even slightly
> more. SVN tags are a primitive thing simply meant to make it easier for the
> dev to remember from whence a version came, should they need to branch from
> it and do a fix, for example. There's really no point to SVN tagging at
> all, save that, as the mvn release plugin is quite happy to permanently
> brand your history with the fact that it did what it did, and that alone,
> with some patience (SVN is the definition of slow), and simple grep use is
> enough to expose where the releases happened over time. From those
> numerical revisions, you could then easily branch, dispensing with the
> expensive SVN tag operation entirely.
>
> However, this isn't the place for such a discussion, hence my not replying
> to Tom's previous comments. Bait taken, this time, though. I'm sure you
> didn't mean to patronise me, though, right? :-)
>
> Regards,
>
> Fred.
>
>
> On Wed, Jan 30, 2013 at 2:11 PM, Mark Struberg <[email protected]> wrote:
>
>>
>>
>> Hi Fred!
>>
>> This is absolutely no problem in SVN. Even deleting a TAG will not remove
>> it from the history! It is still there and can be checked out with the
>> revision. It's just not listed in the directory anymore...
>>
>> This is of course not the case for some other SCMs like e.g. CVS where
>> you would loose history, but removing a tag in SVN is perfectly fine!
>>
>> LieGrue,
>> strub
>>
>>
>>
>>
>>
>> >________________________________
>> > From: Fred Cooke <[email protected]>
>> >To: [email protected]
>> >Sent: Wednesday, January 30, 2013 7:54 AM
>> >Subject: Re: [mojo-dev] [VOTE] Release mojo-parent version 31 and
>> mojo-sandbox-parent version 12 (take 2)
>> >
>> >
>> >Wait a second, did you guys delete the tag and re-tag? (or modify tag?)
>> :-o Is this normal codehaus practice? Or is it because I just crawled out
>> of bed and am very wrong/confused?
>> >
>> >
>> >On Tue, Jan 29, 2013 at 11:51 PM, Tony Chemit <[email protected]>
>> wrote:
>> >
>> >On Tue, 29 Jan 2013 23:50:34 +0100
>> >>Tony Chemit <[email protected]> wrote:
>> >>
>> >>my +1
>> >>
>> >>tony.
>> >>
>> >>
>> >>> Hi,
>> >>>
>> >>> I'd like to release version 31 of the mojo-parent and version 12 of
>> the mojo-sandbox-parent
>> >>>
>> >>> Diff between last release of mojo-parent:
>> >>>
>> https://fisheye.codehaus.org/browse/mojo/trunk/mojo/mojo-parent/pom.xml?r1=16245&r2=17891
>> >>>
>> >>> m-surefire-p plugin and report version are now synch:) (thanks Anders
>> for this one!)
>> >>>
>> >>> A lots of updates were done (
>> https://jira.codehaus.org/browse/MOJO-1898).
>> >>>
>> >>> There is also https://jira.codehaus.org/browse/MOJO-1722 fixed and
>> >>> https://jira.codehaus.org/browse/MOJO-1807.
>> >>>
>> >>> Note that m-compiler-p stays on version 2.5.1 since we do not need at
>> the moment incremental compilation stuff (our project are not multi-module).
>> >>>
>> >>> Surefire
>> >>> Staging Repositories:
>> >>> General:  https://nexus.codehaus.org/content/groups/staging/
>> >>> Exclusive:
>> https://nexus.codehaus.org/content/repositories/orgcodehausmojo-009/
>> >>> (Staging) Site:
>> >>> not available
>> >>> SCM Tag:
>> >>> http://svn.codehaus.org/mojo/tags/mojo-parent-31
>> >>> http://svn.codehaus.org/mojo/tags/mojo-sandbox-parent-12
>> >>>
>> >>> [ ] +1
>> >>> [ ] +0
>> >>> [ ] -1
>> >>>
>> >>> The vote is open for 72 hours and will succeed by lazy consensus.
>> >>>
>> >>>
>> >>> Cheers,
>> >>>
>> >>> tony.
>> >>
>> >>---------------------------------------------------------------------
>> >>To unsubscribe from this list, please visit:
>> >>
>> >>    http://xircles.codehaus.org/manage_email
>> >>
>> >>
>> >>
>> >
>> >
>> >
>>
>> ---------------------------------------------------------------------
>> To unsubscribe from this list, please visit:
>>
>>     http://xircles.codehaus.org/manage_email
>>
>>
>>
>

Reply via email to