The proposal does seem logical to me. I am happy to participate as a committer in whatever the result of this is, but the statement about the historical PMC membership in Flex is not true for me (and I prefer to remain as a committer for FlexJS, not a PMC member).
I was also surprised not to see Justin and Chris in the proposed PMC. And following on from both of their comments in the thread, overlooking their contributions is an omission given that the evidence is there. Justin's encouragement was the main reason I ended up contributing to FlexJS in the first place, so beyond what he added in the thread I can confirm he has also 'recruited' at least one committer for FlexJS, and I know Chris had to work hard with me (and others) to promote the benefits of maven, but I am now hooked :). On Tue, Aug 29, 2017 at 4:37 PM, Alex Harui <aha...@adobe.com.invalid> wrote: > Hi Folks, > > I (and many others on the PMC) believe there are sufficient reasons to ask > the Apache Board of Directors to create a new Apache Top-Level Project > called Apache FlexJS separate from Apache Flex. The Apache Board has > approved splitting other projects in the past and the proposal below is > based on a past proposal to split another Apache project. > > The main reason is that there seems to be a set of folks who are only > working on Apache Flex and another set who are only working on FlexJS. > Only a few have made commits to both products in the past year and the > contribution to one product or the other was minimal. Second, forking > should save time for folks only interested in Flex or FlexJS as you don't > have to filter out emails, web site and wiki information and bugs reports > that apply to the product you are not interested in. Third, forking would > give a cleaner opportunity to try out Apache's GitBox implementation that > allows us to leverage more GitHub functionality and make GitHub more > primary in our development pipeline. Fourth, many folks on the PMC have > not been active in a long time, and a new project with a smaller PMC and > committers list might make it more clear who is involved with FlexJS. > > --------- proposal ----------- > > Apache FlexJS for Apache Top-Level Project > > Abstract > > Apache FlexJS is a new implementation of the principles of Apache Flex but > designed for JavaScript runtimes instead of Adobe Flash/AIR runtimes. > Apache FlexJS was created and released as part of the Apache Flex project > and the UI framework is a full rewrite and does not leverage the code > bases donated by Adobe. Only the FlexJS compiler source was donated by > Adobe. > > Over time, the set of Apache Flex committers who are working on Apache > Flex releases have become distinct from those working on FlexJS. Add to > that the Adobe announcement of the end-of-life for Flash in the browser, > it makes sense to separate FlexJS from Flex so that followers of one side > or the other don't have to filter out emails, bugs, web pages, etc, that > they don't need. > > The proposal is to form an Apache FlexJS top-level project instead of > creating a new Incubator Podling. That is because all of the folks being > proposed to the new Apache FlexJS PMC below have been on the Apache Flex > PMC for months, if not years, and understand how to participate > productively in an Apache project. Plus, Apache FlexJS has already been > released by the top-level Apache Flex project. > > Overview of FlexJS > > Apache FlexJS is designed to improve developer productivity in creating > applications for wherever Javascript runs, including browsers as well as > Apache Cordova applications, Node, etc. FlexJS already has an ecosystem > of folks creating IDEs for FlexJS, and some folks are at various points in > the process of migrating existing Flex applications to FlexJS. > > Current Status > > Meritocracy > > Within Apache Flex, the FlexJS development has been conducted as a > meritocracy. No existing FlexJS developers came with the compiler code > base donated by Adobe nor were they on the initial committers list for > Apache Flex other than Alex Harui. Everyone else has earned their way in > by contributing and earning merit. > > Community > > Many FlexJS users are trying to migrate an existing Apache Flex app off of > Adobe Flash. As their migration progresses they will become less and less > interested in Apache Flex releases and activity. However, the Apache Flex > JIRA project contains some 30,000 Flex issues migrated from Adobe's bug > base, and the team page for Flex makes it hard to determine which > committers are active on FlexJS, and it is frequently hard to determine if > a user's question is about regular Flex or FlexJS. We are seeing an > increase in user interest with the Flash end-of-life announcement and > having separate mailing lists, web pages, and JIRA projects will help save > everyone time and energy. > > Alignment > > Apache FlexJS is already being released by Apache. > > Known Risks > > Code Extraction > > The flex-asjs, flex-tours, flex-typedefs, and flex-falcon Git repos would > be migrated to GitBox. Other Apache Flex repos would be treated as > external dependencies. This includes flex-utilities and the Apache Flex > Installer. > > Orphaned Projects > > Most FlexJS PMC members are planning to remain on the Apache Flex PMC to > help out just-in-case, but are already mostly inactive on the Flex code > bases. However, there are at least 3 Apache Flex PMC members who have not > contributed at all to FlexJS and participate in regular Flex releases, so > the expectation is that Apache Flex will remain a viable project for some > time to come. > > Inexperience with Open Source > > All of the initial team has been contributing to Apache Flex for months if > not years. > > Homogenous Developers > > No company employs more than two of the members of the initial team. > > Reliance on Salaried Developers > > Until about a year ago, only 2 of the proposed PMC members were paid to > work on FlexJS. The rest contributed in their spare time. Two more are > now paid to migrate an existing application and thus contribute to add > missing features and fix bugs. By creating a separate project, we hope to > attract more folks who may not want to be involved with the legacy Apache > Flex code. > > Relationship with Existing Apache Projects > > Apache FlexJS might use some releases from Apache Flex for some features > or for an Installer app. FlexJS can be used to create Apache Cordova > applications and we hope to find time to make the Cordova community aware > of this capability. The Apache FlexJS Compiler has dependencies on Apache > Commons, and we use Apache Ant and Apache Maven for building the various > release artifacts. > > Cryptography > > Apache FlexJS does not currently distribute cryptography code other than > an MD5 hash algorithm. > > Required Resources > > Mailing Lists > > -private@ > -dev@ > -commits@ > -users@ > -issues@ > > Version Control > > We would like to use GitBox if it is practical to convert the current git > repos listed above. We want to explore using GitHub's wiki and > documentation tools. > > Issue Tracking > > If GitBox allows us to use GitHub issues, we will use that, otherwise, we > will request a new JIRA project from Infra. > > Other Resources > > We would like to try using Git for our website version control and > whatever the publishing process is. We will probably continue to run jobs > on builds.a.o. > > Initial PMC > > Harbs > Yishay Weiss > Alex Harui > Peter Ent > Josh Tynjala > Piotr Zarzycki > Carlos Rovira > Greg Dove > Erik De Bruin > Omprakash Muppirala > Frédéric Thomas > Olaf Krueger > Michael Schmalle > Dave Fisher > > > Initial Committers > > Christofer Dutz > Andy Dufilie > Darrell Loverin > > We propose Harbs as initial PMC Chair and VP Apache FlexJS. > > Affiliations > > Alex Harui and Peter Ent are currently employees of Adobe Systems Inc. > Nobody else works for a large corporation. > > --------- end proposal ----------- > > > It isn't clear how much longer Adobe will pay Peter and I to work on Flex > and/or FlexJS. So getting rid of as much overhead as possible so we can > try to get as many folks to use FlexJS in the next few months will help > make a case to keep us on FlexJS longer. Yes there is a risk that forking > may be a distraction and cause a loss of momentum, but whether I get to > continue full-time or in my spare-time, I want to make the most efficient > use of my time. > > > I will be stepping down as Apache Flex chair since I want to spend most of > my time on FlexJS. I will remain on the Apache Flex PMC to help out in > case of emergency. > > Please provide feedback and questions ASAP. The plan is to have a vote in > a few days so we can present this proposal to the board around September 1 > so it can be on the September board agenda along with changes to the > Apache Flex chair. > > > Thanks, > -Alex > >
