+1 for remove <https://maas.mail.163.com/dashi-web-extend/html/proSignature.html?ftlId=1&name=yokka&uid=iyokka%40outlook.com&iconUrl=https%3A%2F%2Fmail-online.nosdn.127.net%2Fqiyelogo%2FdefaultAvatar.png&items=%5B%22iyokka%40outlook.com%22%5D> [https://mail-online.nosdn.127.net/qiyelogo/defaultAvatar.png] yokka
iyo...@outlook.com 签名由 网易邮箱大师<http://mail.163.com/dashi/> 定制 On 08/17/2018 17:11，G C<cg.f...@gmail.com><mailto:cg.f...@gmail.com> wrote： +1 Deep agree -- Best Regards！ cgfork On Fri, Aug 17, 2018 at 4:53 PM yuhang xiu <carry...@gmail.com> wrote: Deep agree with this proposal :) Jerrick Zhu <jerr...@apache.org> 于2018年8月17日周五 下午1:39写道： I agree with keeping the following modules: * dubbo-registry-multicast, * dubbo-rpc-injvm, * dubbo-serialization-fastjson * dubbo-serialization-jdk They're useful for new users and develop experience. On Fri, Aug 17, 2018 at 11:10 AM Ian Luo <ian....@gmail.com> wrote: wow, this is a big step but I have somewhat opposite opinion against Jerrick's proposal. Instead of keeping recommended extensions only but moving out others into eco system, I would like to revise it to keeping most used extensions, use Jerrick's list as an example: * dubbo-registry-multicast <- should keep, it's good for dev experience * dubbo-registry-redis <- should move, since not everyone needs this feature ... * dubbo-remoting-grizzly <- should move, ditto * dubbo-remoting-p2p <- should move, ditto * dubbo-remoting-zookeeper <- should move, ditto * dubbo-remoting-mina <- should move, ditto ... * dubbo-rpc-injvm <- should keep, since local-call is one necessary feature, think about EJB's local call. * dubbo-rpc-memcached <- should move * dubbo-rpc-redis <- should move * dubbo-rpc-thrift <- should move * dubbo-rpc-webservice < - should move ... * dubbo-serialization-fastjson <- should keep, very popular json library * dubbo-serialization-fst <- should move * dubbo-serialization-jdk <- should keep, last resort when other serialization fails to work. Just my two cents, -Ian. On Thu, Aug 16, 2018 at 4:01 PM Jerrick Zhu <jerr...@apache.org> wrote: Hi, community As you can see, Dubbo now has an ecosystem: https://github.com/dubbo .There has a lot of interesting projects there, such as: * node, go, python and php implementations * dubbo rpc, serialization and registry extensions * dubbo samples and dubbo useful plugins and tools Now Dubbo core which is http://github.com/apache/incubator-dubbo , it's too big. As a result, it takes nearly 30~40mins to finish travis CI. And also, there are a lot of sub modules have never been modified almost, such as: * dubbo-registry-multicast * dubbo-registry-redis ... * dubbo-remoting-grizzly * dubbo-remoting-p2p * dubbo-remoting-zookeeper * dubbo-remoting-mina ... * dubbo-rpc-injvm * dubbo-rpc-memcached * dubbo-rpc-redis * dubbo-rpc-thrift * dubbo-rpc-webservice ... * dubbo-serialization-fastjson * dubbo-serialization-fst * dubbo-serialization-jdk ... So I suggest, move the above sub modules to ecosystem, each of them as a single project, such as https://github.com/dubbo/dubbo-rpc-native-thrift Also, dubbo-demo also needs to move to ecosystem. What do u guys think? Sincerely. Jerrick
