commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From sebb <seb...@gmail.com>
Subject Re: [VOTE] Release of commons-email-1.3 based on RC5
Date Wed, 12 Dec 2012 19:47:07 GMT
On 12 December 2012 19:39, Siegfried Goeschl <sgoeschl@gmx.at> wrote:
> Hi folks,
>
> to avoid the full circles here
>
> * I hopefully fixed the binary compatibility issue, wrote test and also
> tested it with my production code
>
> * I moved the constants to EmailConstants because there are many constants
> and using an non-trivial implementation class (Email) to access a few
> constants is also questionable in my opinion

A better solution might have been to use a constant class rather than
an interface.

The Email class can then reference the constant class:

public static final String XYZ = EmailConstants.XYZ;

Other classes can then reference EmailConstants directly.

> * The design of commons-email is flawed in a few ways and there is not a lot
> you can fix in a major version. So I suggest that we focus on delivered
> value instead
>
> Cheers,
>
> Siegfried Goeschl
>
>
>
> On 12.12.12 15:06, sebb wrote:
>>
>> On 12 December 2012 13:17, Gary Gregory <garydgregory@gmail.com> wrote:
>>>
>>> On Wed, Dec 12, 2012 at 3:59 AM, Thomas Neidhart
>>> <thomas.neidhart@gmail.com>wrote:
>>>
>>>> On Wed, Dec 12, 2012 at 4:58 AM, Gary Gregory <garydgregory@gmail.com
>>>>>
>>>>> wrote:
>>>>
>>>>
>>>>> Thank you for doing another RC.
>>>>>
>>>>> While I was digging for a justification of the Clirr errors, I found
>>>>> this
>>>>> in the release notes: "Clirr reports several errors for this release
>>>>> due
>>>>
>>>> to
>>>>>
>>>>> moving constants from the Email class to the newly introduced
>>>>> EmailConstants interface. These changes are guaranteed to be binary
>>>>> compatible."
>>>>>
>>>>> Is it really binary compatible? What if I use reflection to access the
>>>>> constant on Email, will the reflection call be redirected to
>>>>> EmailConstants? There's unit test for ya ;)
>>>>>
>>>>> Using an interface to define constants is a no-no in my book. I've seen
>>>>> this discussed before in other places and for a long time, but to
>>>>> summarize, I see an interface as defining a contract for a class to
>>>>> implement. A constant does not fit.
>>>>>
>>>>> Constants in interface feels like a hack to provide the short hand of
a
>>>>> class implementing an interface just to be able to access the constants
>>>>> without qualifying them with a type. Not nice design IMO and a dubious
>>>>> us
>>>>> of an interface, very Java 1.0. It seems that static imports is another
>>>>> attempt to solve this desire for a short hand to use constants.
>>>>>
>>>>> What to do? Move the constants back to their 1.2? What's so bad about
>>>>
>>>> that?
>>>>>
>>>>> Hm...
>>>>>
>>>>> Make the EmailConstants a class instead of an interface? If binary
>>>>> compatible is broken, the constants have to move back, and you can
>>>>> still
>>>>> have a new EmailConstants class and deprecate the old constants to
>>>>> point
>>>>
>>>> to
>>>>>
>>>>> the new class.
>>>>>
>>>>> Maybe I'll see this more clearly in the AM...
>>>>>
>>>>> Interested in you all's feedback.
>>>>>
>>>>
>>>> Hi Gary,
>>>>
>>>> well, I think we go in circles with this change ;-).
>>>>
>>>> I assumed that this topic is settled after reading the comment from sebb
>>>> in
>>>> the RC2 thread (see http://markmail.org/message/svrb7nf3ocz7lgmd).
>>>>
>>>> Otoh, it's the first time I see constants in an interface and would be
>>>> in
>>>> favor of reverting to the previous version (also because I do not fully
>>>> understand the rationale behind the change, some of the constants are
>>>> not
>>>> even used and thus have been deprecated).
>>>>
>>> --
>>>
>>>>
>>>> Maybe we should postpone this kind of refactoring to 2.0 and do it then
>>>> in
>>>> a proper way. Introducing this interface just created headaches and I
>>>> also
>>>> had to disable some checks (e.g. InterfaceIsAType in checkstyle) because
>>>> of
>>>> it.
>>>>
>>>
>>> That sounds like a good way to go to get 1.3 out the door.
>>
>>
>> Agreed.
>>
>>> Gary
>>>
>>>
>>>>
>>>> Thomas
>>>>
>>>>
>>>>> On Tue, Dec 11, 2012 at 5:24 PM, Thomas Neidhart
>>>>> <thomas.neidhart@gmail.com>wrote:
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> I would like to call a vote from commons-email-1.3 based on RC5.
>>>>>>
>>>>>> This release candidate has the following changes compared to RC4
>>>>>>
>>>>>> +) update index and building page with correct information wrt Java
>>>>>>     compatibility
>>>>>> +) update release notes with info on Java compatibility and Clirr
>>>>
>>>> errors
>>>>>>
>>>>>> +) fix svn:keywords for all source files and remove use of $Date$
tags
>>>>>> +) add $Id$ tags for all newly introduced source files in 1.3
>>>>>> +) update javax.mail.mail dependency to 1.4.5
>>>>>> +) fix PMD warnings and add NOPMD comment for false positives
>>>>>> +) added findbugs exclude filter for false positives
>>>>>> +) fix release date in changes.xml
>>>>>> +) correctly removed *.asc.[md5,sha1] files from Nexus staging area
>>>>>>
>>>>>> The files:
>>>>>>
>>>>>> The artifacts are deployed to Nexus:
>>>>>>
>>>> https://repository.apache.org/content/repositories/orgapachecommons-137/
>>>>>>
>>>>>>
>>>>>> The tag:
>>>>>>
>>>>>
>>>>
>>>> https://svn.apache.org/repos/asf/commons/proper/email/tags/EMAIL_1_3_RC5/
>>>>>>
>>>>>>
>>>>>> The site:
>>>>>> http://people.apache.org/builds/commons/email/1.3/RC5/
>>>>>>
>>>>>> Additional Notes:
>>>>>>
>>>>>> o the download page and api links to older releases only work on
>>>>>>    the published site and will be corrected after release.
>>>>>>
>>>>>> Please take a look at the commons-email-1.3 artifacts and vote!
>>>>>>
>>>>>> ------------------------------------------------
>>>>>> [ ] +1 release it
>>>>>> [ ] +0 go ahead I don't care
>>>>>> [ ] -1 no, do not release it because
>>>>>> ------------------------------------------------
>>>>>>
>>>>>> Vote will remain open for at least 72 hours.
>>>>>>
>>>>>> Thanks in advance,
>>>>>>
>>>>>> Thomas
>>>>>>
>>>>>> ---------------------------------------------------------------------
>>>>>> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
>>>>>> For additional commands, e-mail: dev-help@commons.apache.org
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> E-Mail: garydgregory@gmail.com | ggregory@apache.org
>>>>> JUnit in Action, 2nd Ed: <http://goog_1249600977>http://bit.ly/ECvg0
>>>>> Spring Batch in Action: <http://s.apache.org/HOq>http://bit.ly/bqpbCK
>>>>> Blog: http://garygregory.wordpress.com
>>>>> Home: http://garygregory.com/
>>>>> Tweet! http://twitter.com/GaryGregory
>>>>>
>>>>
>>>
>>>
>>>
>>> --
>>> E-Mail: garydgregory@gmail.com | ggregory@apache.org
>>> JUnit in Action, 2nd Ed: <http://goog_1249600977>http://bit.ly/ECvg0
>>> Spring Batch in Action: <http://s.apache.org/HOq>http://bit.ly/bqpbCK
>>> Blog: http://garygregory.wordpress.com
>>> Home: http://garygregory.com/
>>> Tweet! http://twitter.com/GaryGregory
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
>> For additional commands, e-mail: dev-help@commons.apache.org
>>
>>
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
> For additional commands, e-mail: dev-help@commons.apache.org
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org


Mime
View raw message