EDT:0.7.0 Planning:Iteration 6
Focus for Innovate
- showcasing the extensibility concepts
- relaunching our EDT website presence, including collateral
- basic content that shows we are a live site - banner, doc, samples, etc.
- extensibility paper (Jeff) - extending gen/run, and plugging it into the IDE
- planning information - share with our open community, get some discussions going to shown signs of life
- how do we get tied into the overall Eclipse communications
- talk to eclipse to understand better how we should hook in so that non-EGL eclipse users would hear about us.
- building 3rd party community
- ASIST is working toward a contribution to the IDE - even if they can't get the contribution together by June, we should be showing the discussion about it in the mailing list, so other people can see that this is going on, and that non-IBM contributions really will be made. We need to get our talks with ASIST out of emails, and onto the website (once the website is working correctly). Tim has contacted them.
- ClearBlade - EGLdoc contribution - get this conversation going on the mailing list. Will has contacted Aaron
- we should initiate mailing list entries about the additional code that we are contributing. IDE code - explain about the extensibility, and any other differences for RBD. Debug from SMAP - explain how that works
- enabling business partners - we haven't been hearing requests, but we should contact them this week to see what they are working toward for Innovate, and if it requires anything additional from us in the driver.
- ASIST - Tim
- ClearBlade - Will
- Xact - Tim - Lars asked for some more function, which Paul has completed.
- Synobsys if we get the RUI widgets out there
- demo - desire is still for a thin path for RUI calling services that do database access. Don't necessarily need the editor, more JS gen/run. But the message is more important than the running driver.
- do we need to have a frozen beta driver for anybody who wants to download a driver, or just have the daily builds. This would be an integration build that would be frozen on some known level that we have run on.
Also need to be planning toward EDT 0.7.0 - end of August - we need to be working toward that now, because it will take us longer than just June-Aug to get what we want there. We will just continue building up our functional driver toward 0.7.0 in August. Whatever we have in June is what we will show at Innovate
- IDE - continue on current path - finish preferences / properties, wizard to create EGL project - start simple, just implementing what the current action does that converts a java project to EGL. Then we can decide what additional functionality is desired. Getting the project explorer to be EGL-aware. Also need perspective.
- Compiler - front end needs to be designed - refactoring, validation framework, etc. and then tasked. This needs to be prioritized against the indexing work. For Paul, this is higher. Also the utility to create IRs - Tim needs to poll the BPs to see whether this is urgent in this time frame, or have they already worked around it. If not urgent, then Paul should focus on the front end redesign. Tim says it doesn't seem to be urgent, so defer to later implementation.
- Core - indexing to include annotations & stereotypes, smarter generation - understanding structurally changed. Paul & Justin with Tim's help for investigation.
- debug - continue on current path as a lower priority than core work, if time allows.
- Builds - the additional code being contributed needs to be added to the build. Matt, but not until after Eclipse approves our contribution. Should just be an hour or so of work.
- Services - work in the priority shown in RTC
- data base I/O - needs to be designed, including how JPA will play in, vs straight SQL. In I6, we will target having the direction / design set.
- Language test framework - generation of the program based on library is finished, in command line mode. Now working toward automation. Then need to handle the results, to get a report out - build a tool to read the output and display it. Also start writing the test cases. Kathy will be full time, plus whatever time Jing can give. This might drive some requirements on the EDT language gen/run implementation.
- New thought -tooling explicitly for extending generators - an EGL generator project. Tim needs to spec out what this would look like. This might be more important for Innovate than other incremental IDE work, but it needs to get designed first.
- Exploration & planning for how we will do open development, to be implemented in I7 - project plans, milestone plans, iteration plans, using bugzilla for stories, line items, & tasks, and how the wiki plays in. Jon & Lisa