+1 on Stephen's reasons Jeff
On Thu, Feb 7, 2013 at 12:46 PM, Stephen Connolly < stephen.alan.conno...@gmail.com> wrote: > On 7 February 2013 08:58, Chris Graham <chrisgw...@gmail.com> wrote: > > > > > > > Sent from my iPhone > > > > On 07/02/2013, at 6:59 PM, Baptiste MATHUS <m...@batmat.net> wrote: > > > > > Le 7 févr. 2013 04:54, "Chris Graham" <chrisgw...@gmail.com> a écrit : > > >> > > >> Hey All. > > >> > > >> Regarding the discussions around upgrading to a minimum of Java 1.6. > > >> > > >> Whilst I understand the desire for developers to play with something > > shiny > > >> and new, > > > > > > Please. Can we stop using that kind of father-ish formulation? That's > not > > > the first time. > > > It makes it sound like "meeh, it happened again. Those /developers/ > kids > > > not getting what business is about just did it again...". > > > > > >> I do find that the current 1.5 based Maven is more than sufficient > > >> for our needs. > > >> > > >> I lot of responses about the upgrade are normally around the "just > > upgrade > > >> the java" (assuming that I am running under tomcat of similar [if only > > it > > >> was so simple as that!]). > > >> > > >> I am running Maven inside of Jenkins inside of WebSphere on AIX. I am > > >> currently hosting Jenkins under WAS 6.1 on AIX 5.3. Whilst AIX 5.3 is > > >> nearing (or may have reached it's EOS), WAS 6.1's EOS dates have *just > > > been > > >> extended* by a year! > > >> > > >> Additionally, although the Sun/Oracle Java 1.5 may have been EOS'd > > quite a > > >> while ago, the IBM Java 1.5/5 is most definately not EOS. Please see: > > >> http://www.ibm.com/developerworks/java/jdk/lifecycle/index.html. > > >> > > >> Running on AIX, we have one choice in JDK: IBM. > > > > Can you get a 1.6+ JRE on there? In other words, use 1.6 or 1.7 to run > Maven and use toolchains to fork down to the 1.5 JDK for compiling and > running tests. > > > > >> > > >> The IBM 1.5/5 EOS is Sept 2015! > > >> > > >> The reality is that we have lots of WebSphere servers running in large > > > data > > >> centres. So simply stating "just upgrade java" (assuming that it is > > > running > > >> under Tomcat or similar) is simply not an option for us. > > > > > > Twice in a row. What's the issue with products that eases your life > when > > it > > > comes to upgrading things? > > > > > > > Not at all. It's a matter of scale and support. > > > > We have one client who has 3,500 instances of WAS running and no tomcat. > > > > We are not going to retrain the (thousands) of support staff (globally) > to > > support a new product easily (or cheaply), especially when you generally > > have multiple different buckets of money for dev, support maint and BAU. > > > > > > >> > > >> Whilst I do realise that very (1.5% by your [Jenkins] figures {I > suspect > > >> that maven figures are similar]) few of us still run on 1.5, some of > us > > >> will simply not be able to upgrade to a newer version of WAS/JDK as a > > >> simple task. It's not as easy as clicking your fingers. > > > > > > I'm not sure I perfectly see your point. We also use aix and jenkins > and > > > Maven, and it's possible as anywhere else to install many JDK versions > on > > > an IBM/aix server. You're then not forced to use the same JDK version > for > > > your builds and for running your jenkins server. > > > > That may work in some cases. Agreed. But perhaps not all. Thinking ESB > and > > BPM builds here; I'll need to think about that one. > > > > Sounds like either defects in Toolchains or the plugins used for those > builds do not understand toolchains. > > Long term we need to get those issues resolved, because like it or not, at > *some stage* we need to move off of 1.5 and onwards to 1.6 or 1.7 or 1.8. > > I am against moving up "just because". However I am all in favour of moving > up "because XYZ". > > For me invalid reasons to move to as a runtime requirement 1.6 are things > like: > > * I cannot get a 1.5 JDK on my chosen development machine => configure > animal-sniffer > > * There is a compiler bug in generics that means that some of the generics > I want to introduce into Maven's APIs will not compile on JDK 1.5 => that > is a valid reason to require 1.6 as a build maven requirement, but > animal-sniffer can let us keep 1.5 as a run-time requirement. > > Valid reasons to move to 1.6 are things like: > > * This 3rd party dependency we use has a bug that needs fixing, they have > fixed it in version V.W but that artifact is using 1.6 bytecode so we > either keep the bug or upgrade the dependency and consequently upgrade the > minimum required JVM to run Maven. > > * There is a big feature that I want to implement and it will be 10 times > easier to write and maintain with the language features in 1.6 > > My point is I am 100% fine with us upping the requirement to 1.6 or 1.7. I > just want a *good* reason... doesn't have to be a *big* reason... just a > good one. > > -STephen > > > > >> > > >> I also do wonder how many installations are in said data centres and > are > > >> unable to report their presence. So I do believe that the 1.5% figure > > > would > > >> be low, but certainly within an order of magnitude. > > > > > > I'm thinking that'd certainly be averaged with versions bigger than > JDK5 > > > also not reporting. > > > > > >> > > >> So, please do not cut us off from future updates. > > >> > > >> Thanks, > > >> > > >> -Chris > > >> > > >> > > >> > > >> On Thu, Feb 7, 2013 at 12:05 PM, Manfred Moser <manf...@mosabuam.com> > > > wrote: > > >> > > >>> Totally agree... if people really need to build for older Java > runtimes > > >>> they still can.. but if Oracle thinks they dont want to support JDK > < > > > 1.7 > > >>> without getting paid why would the Maven project do it ;-) > > >>> > > >>> For now 1.6 seems just fine and I would even say a jump to 1.7 in the > > > next > > >>> year or three (;-)) would be reasonable.. > > >>> > > >>> manfred > > >>> > > >>>> +1, bump to JDK6 minimum for Maven. > > >>>> > > >>>> > > >>>> 2013/2/6 Stephen Connolly <stephen.alan.conno...@gmail.com> > > >>>> > > >>>>> I think we should at least have a minor version bump on core to > > >>>>> co-incide... Though I think calling it maven 4.0 might be better > > > (that > > >>>>> way > > >>>>> we catch up with the model version ;-) > > >>>>> > > >>>>> On Wednesday, 6 February 2013, Olivier Lamy wrote: > > >>>>> > > >>>>>> Hi, > > >>>>>> As we are now in 2013, it's probably time to think about that > > > (again). > > >>>>>> > > >>>>>> Reading [1], even 1.6 won't be anymore updated after feb 2013. > > >>>>>> > > >>>>>> Can we say we are safe to go to 1.6 as minimum required ? > > >>>>>> > > >>>>>> NOTE: That will probably need a vote. So depending on how the > > > thread > > >>>>>> move, I will start a vote (or not). > > >>>>>> > > >>>>>> Thanks > > >>>>>> -- > > >>>>>> Olivier Lamy > > >>>>>> Talend: http://coders.talend.com > > >>>>>> http://twitter.com/olamy | http://linkedin.com/in/olamy > > >>>>>> > > >>>>>> [1] http://www.oracle.com/technetwork/java/eol-135779.html > > >>>>>> > > >>>>>> > > > --------------------------------------------------------------------- > > >>>>>> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org > > >>>>> <javascript:;> > > >>>>>> For additional commands, e-mail: > > >>>>> dev-h...@maven.apache.org<javascript:;> > > >>>>>> > > >>>>>> > > >>>>> > > >>>>> -- > > >>>>> Baptiste <Batmat> MATHUS - http://batmat.net > > >>>>> Sauvez un arbre, > > >>>>> Mangez un castor ! nbsp;! > > >>>> > > >>> > > >>> > > >>> --------------------------------------------------------------------- > > >>> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org > > >>> For additional commands, e-mail: dev-h...@maven.apache.org > > >>> > > >>> > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org > > For additional commands, e-mail: dev-h...@maven.apache.org > > > > > -- Jeff MAURY "Legacy code" often differs from its suggested alternative by actually working and scaling. - Bjarne Stroustrup http://www.jeffmaury.com http://riadiscuss.jeffmaury.com http://www.twitter.com/jeffmaury
