Hi all, This 3 repo has moved successfuly to gitbox:
https://github.com/apache/geronimo-xbean https://github.com/apache/geronimo-javamail https://github.com/apache/geronimo-txmanager We can now merge the pending PRs. regards, François fpa...@apache.org Le 08/06/2021 à 14:15, Richard Zowalla a écrit : > Thx for the ticket id ! > > Am Dienstag, den 08.06.2021, 14:07 +0200 schrieb Francois Papon: >> Hi, >> >> Migration is still pending, waiting for infra: >> >> https://issues.apache.org/jira/browse/INFRA-21908 >> <https://issues.apache.org/jira/browse/INFRA-21908> >> >> regards, >> >> François >> fpa...@apache.org >> >> Le 08/06/2021 à 13:56, Richard Zowalla a écrit : >>> Hi François, >>> >>> any updates from INFRA on this? Couldnt find the ticket id anymore, >>> sry. >>> >>> Gruss >>> Richard >>> >>> Am Mittwoch, den 19.05.2021, 09:38 +0200 schrieb Francois Papon: >>>> Hi, >>>> >>>> Yes, we plan to do this just after the migration to git ;) >>>> >>>> regards, >>>> >>>> François >>>> fpa...@apache.org >>>> >>>> Le 19/05/2021 à 09:09, Zowalla, Richard a écrit : >>>>> Hi, >>>>> >>>>> thanks for your response! I think, that [1] might also affect >>>>> the >>>>> hard- >>>>> coded TLS1.0 in GERONIMO-6792 [2], so the pending patch would >>>>> be >>>>> very >>>>> appreciated. >>>>> >>>>> Maybe after the migration to git? ;) >>>>> >>>>> Gruss >>>>> Richard >>>>> >>>>> [1] https://bugs.openjdk.java.net/browse/JDK-8202343 >>>>> [2] https://issues.apache.org/jira/browse/GERONIMO-6792 >>>>> >>>>> Am Samstag, den 01.05.2021, 08:20 +0200 schrieb >>>>> fpa...@apache.org: >>>>>> Hi, >>>>>> >>>>>> I think I can take a look to the Jira and merge the PRs. >>>>>> >>>>>> regards, >>>>>> >>>>>> François >>>>>> fpa...@apache.org >>>>>> >>>>>> Le 28/04/2021 à 11:09, Zowalla, Richard a écrit : >>>>>>> Just to follow up on this thread: >>>>>>> >>>>>>> Do we have any plans for moving forward with the mail- >>>>>>> related >>>>>>> patches? >>>>>>> The hard-coded TLS config in mail is a bit "pain" ;) >>>>>>> >>>>>>> Gruss >>>>>>> Richard >>>>>>> >>>>>>> Am Dienstag, den 23.03.2021, 08:50 +0100 schrieb Romain >>>>>>> Manni- >>>>>>> Bucau: >>>>>>>> Well it can use a singleton but from a factory method. So >>>>>>>> immediate >>>>>>>> solution is to add a public static X getInstance();. >>>>>>>> But as mentionned it means, to keep the pluggability we >>>>>>>> should >>>>>>>> target >>>>>>>> with such a spi, you will enforce all other impl to use >>>>>>>> such >>>>>>>> a >>>>>>>> pattern (you cant' just switch with -D easily since >>>>>>>> adding is >>>>>>>> easy >>>>>>>> but dropping system props is almost impossible). >>>>>>>> A noarg public constructor is trivial and more natural >>>>>>>> with >>>>>>>> resources >>>>>>>> IMHO - but once again tomee can does all the work to >>>>>>>> makes it >>>>>>>> equivalent, just requires to duplicate/wrap the impls of >>>>>>>> the >>>>>>>> SPI >>>>>>>> in >>>>>>>> tomee codebase which sounds weird to me ("we have an impl >>>>>>>> but >>>>>>>> you >>>>>>>> need to use another one"). >>>>>>>> >>>>>>>> On a more personal note I think this pattern is no more >>>>>>>> relevant >>>>>>>> and >>>>>>>> has more pitfalls since you enforce a static instance as >>>>>>>> soon >>>>>>>> as >>>>>>>> the >>>>>>>> class is loaded whereas it is not needed depending the >>>>>>>> lifecycle >>>>>>>> of >>>>>>>> your main - it is not much but still, I see it as a leak >>>>>>>> in >>>>>>>> terms >>>>>>>> of >>>>>>>> design (indeed this one is not important and not a >>>>>>>> blocker >>>>>>>> but >>>>>>>> all >>>>>>>> implies to move to the noarg public constructor on my >>>>>>>> side). >>>>>>>> Maybe a habit and personal choice so would be great to >>>>>>>> have >>>>>>>> another >>>>>>>> opinion to move forward :). >>>>>>>> >>>>>>>> Le mar. 23 mars 2021 à 08:38, Zowalla, Richard < >>>>>>>> richard.zowa...@hs-heilbronn.de> a écrit : >>>>>>>>> Hi, >>>>>>>>> >>>>>>>>> I think, it is about the configuration flexibility in >>>>>>>>> tomee's >>>>>>>>> <resource> definitions, which wouldn't allow the use of >>>>>>>>> a >>>>>>>>> singleton >>>>>>>>> instance. Hence, the consuming project would need to >>>>>>>>> implement >>>>>>>>> the >>>>>>>>> interface to make it possible. But I am not that deep >>>>>>>>> as >>>>>>>>> Romain >>>>>>>>> in >>>>>>>>> the >>>>>>>>> TomEE codebase, so it is still a guess from my side. >>>>>>>>> >>>>>>>>> Gruss >>>>>>>>> Richard >>>>>>>>> >>>>>>>>> Am Montag, den 22.03.2021, 23:14 +0100 schrieb Florent >>>>>>>>> Guillaume: >>>>>>>>>> Hi, >>>>>>>>>> >>>>>>>>>> I can drop the private constructor if you want. >>>>>>>>>> I'm surprised it's needed though, as the default >>>>>>>>>> instance >>>>>>>>>> is >>>>>>>>> already >>>>>>>>>> used by the code if no value is provided for the >>>>>>>>>> timeProvider >>>>>>>>>> parameter of TransactionImpl. >>>>>>>>>> >>>>>>>>>> Florent >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> On Mon, Mar 22, 2021 at 5:49 PM Romain Manni-Bucau < >>>>>>>>>> rmannibu...@gmail.com> wrote: >>>>>>>>>>> Hi Richard, >>>>>>>>>>> >>>>>>>>>>> I still think SystemCurrentTime should have a >>>>>>>>>>> public >>>>>>>>>>> noarg >>>>>>>>>>> constructor (or just drop the private one) since it >>>>>>>>>>> will >>>>>>>>>>> enable >>>>>>>>>>> tomee to fully configure dynamically the tx mgr >>>>>>>>>>> with >>>>>>>>>>> this >>>>>>>>>>> new >>>>>>>>>>> feature but otherwise +1 to apply them all. >>>>>>>>>>> >>>>>>>>>>> Romain Manni-Bucau >>>>>>>>>>> @rmannibucau | Blog | Old Blog | Github | LinkedIn >>>>>>>>>>> | >>>>>>>>>>> Book >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> Le lun. 22 mars 2021 à 17:03, Zowalla, Richard < >>>>>>>>>>> richard.zowa...@hs-heilbronn.de> a écrit : >>>>>>>>>>>> Hi all, >>>>>>>>>>>> >>>>>>>>>>>> wanted to raise attention on this again. 6792 >>>>>>>>>>>> would >>>>>>>>>>>> be >>>>>>>>>>>> very >>>>>>>>> nice >>>>>>>>>>>> as we >>>>>>>>>>>> should allow TLS/SSL protocol versions for a >>>>>>>>>>>> given >>>>>>>>>>>> mail >>>>>>>>> server >>>>>>>>>>>> instead >>>>>>>>>>>> of falling back to some hard-coded default. >>>>>>>>>>>> >>>>>>>>>>>> Gruss >>>>>>>>>>>> Richard >>>>>>>>>>>> >>>>>>>>>>>> Am Mittwoch, den 24.02.2021, 09:33 +0100 schrieb >>>>>>>>>>>> Romain >>>>>>>>> Manni- >>>>>>>>>>>> Bucau: >>>>>>>>>>>>> Hi all, >>>>>>>>>>>>> >>>>>>>>>>>>> AFAIK we have a few pending patches to >>>>>>>>>>>>> apply/issue >>>>>>>>>>>>> to >>>>>>>>> close: >>>>>>>>>>>>> - [mail] >>>>>>>>> https://issues.apache.org/jira/browse/GERONIMO-6792: >>>>>>>>>>>> update >>>>>>>>>>>>> some defaults and config capacity >>>>>>>>>>>>> - [mail] >>>>>>>>> https://issues.apache.org/jira/browse/GERONIMO-6801 >>>>>>>>>>>> and >>>>>>>>>>>>> https://issues.apache.org/jira/browse/GERONIMO-6800 >>>>>>>>> (setText) >>>>>>>>>>>>> - [transaction-manager] >>>>>>>>>>>>> https://issues.apache.org/jira/browse/GERONIMO-6805 >>>>>>>>>>>>> : >>>>>>>>>>>>> enable >>>>>>>>> to >>>>>>>>>>>> change >>>>>>>>>>>>> the time evaluator impl >>>>>>>>>>>>> >>>>>>>>>>>>> If someone else can have a review it would be >>>>>>>>>>>>> great >>>>>>>>>>>>> (feel >>>>>>>>> free >>>>>>>>>>>> to >>>>>>>>>>>>> apply the patch or I can do it after). >>>>>>>>>>>>> >>>>>>>>>>>>> note: some of the patches are waiting for some >>>>>>>>>>>>> feedback >>>>>>>>>>>>> - >>>>>>>>> in >>>>>>>>>>>>> particular txmgr one, wonder about tomee >>>>>>>>>>>>> <resource> >>>>>>>>>>>>> usage >>>>>>>>> which >>>>>>>>>>>> can >>>>>>>>>>>>> need to remove the private constructor of the >>>>>>>>>>>>> default >>>>>>>>>>>>> impl >>>>>>>>> to >>>>>>>>>>>> enable >>>>>>>>>>>>> to configure the impl completely. >>>>>>>>>>>>> >>>>>>>>>>>>> Thanks, >>>>>>>>>>>>> Romain Manni-Bucau >>>>>>>>>>>>> @rmannibucau | Blog | Old Blog | Github | >>>>>>>>>>>>> LinkedIn >>>>>>>>>>>>> Book
