marks interesting changes over the 3.0
plan)Please send comments about this draft plan to the eclipse-dev@eclipse.org developer mailing list.
Last revised 18:15 EDT July 29, 2004 (
marks interesting changes over the 3.0
plan)
Please send comments about this draft plan to the eclipse-dev@eclipse.org
developer mailing list.
This document lays out the feature and API set for the next feature release of Eclipse after 3.0, designated release 3.1.
Plans do not materialize out of nowhere, nor are they entirely static. To ensure the planning process is transparent and open to the entire Eclipse community, we (the Eclipse PMC) post plans in an embryonic form and revise them throughout the release cycle.
The first part of the plan deals with the important matters of release deliverables, release milestones, target operating environments, and release-to-release compatibility. These are all things that need to be clear for any release, even if no features were to change.
The remainder of the plan consists of plan items for the three projects under the Eclipse top-level project. Each plan item covers a feature or API that is to be added to Eclipse, or some aspect of Eclipse that is to be improved. Each plan item has its own entry in the Eclipse bugzilla database, with a title and a concise summary (usually a single paragraph) that explains the work item at a suitably high enough level so that everyone can readily understand what the work item is without having to understand the nitty-gritty detail.
Not all plan items represent the same amount of work; some may be quite large, others, quite small. Some plan items may involve work that is localized to a single Platform component; others may involve coordinated changes to several components; other may pervade the entire Platform. Although some plan items are for work that is more pressing that others, the plan items appear in no particular order.
With the previous release as the starting point, this is the plan for how we will enhance and improve it. Fixing bugs, improving test coverage, documentation, examples, performance tuning, usability, etc. are considered routine ongoing maintenance activities and are not included in this plan unless they would also involve a significant change to the API or feature set, or involve a significant amount of work. The intent of the plan is to account for all interesting feature work.
The current status of each plan item is noted:
The release deliverables have the same form as previous releases, namely:
Release milestone occurring at roughly 6 week intervals exist to facilitate coarse-grained planning and staging. The milestones are:
The 3.1 release is targeted for 2Q2005. All release deliverables will be available for download as soon as the release has been tested and validated in the target operating configurations listed below.
In order to remain current, each Eclipse release targets reasonably current versions of the underlying operating environments.
Most of the Eclipse SDK is "pure" Java™ code and has no direct dependence on the underlying operating system. The chief dependence is therefore on the Java 2 Platform itself. The 3.1 release of the Eclipse Project is written and compiled against version 1.4 of the Java 2 Platform APIs, and targeted to run on version 1.4 of the Java 2 Runtime Environment, Standard Edition.
There are many different implementations of the Java 2 Platform running atop a variety of operating systems. We focus Eclipse testing on a handful of popular combinations of operating system and Java 2 Platform; these are our reference platforms. Eclipse undoubtedly runs fine in many operating environments beyond the reference platforms we test. However, since we do not systematically test them we cannot vouch for them. Problems encountered when running Eclipse on non-reference platform that cannot be recreated on any reference platform will be given lower priority than problems with running Eclipse on a reference platform.
Eclipse SDK 3.1 is tested and validated on the following reference platforms (this list is updated over the course of the release cycle):
|
Eclipse Reference Platforms
|
|||
|---|---|---|---|
| Operating system | Processor architecture | Window system | Java 2 Platform |
| Microsoft Windows XP | Intel x86 | Win32 | Sun Java 2 Standard Edition, version 1.4.2_03 for Microsoft Windows |
| Microsoft Windows XP | Intel x86 | Win32 |
IBM 32-bit SDK for Windows, Java 2 Technology Edition, Version 1.4.1 |
|
Microsoft Windows XP |
Intel x86 | Win32 | Sun Java 2 Standard Edition 5.0 for Microsoft Windows |
| Red Hat Enterprise Linux WS 3 | Intel x86 | GTK | Sun Java 2 Standard Edition, 1.4.2_03 for Linux x86 |
| Red Hat Enterprise Linux WS 3 | Intel x86 | GTK | IBM 32-bit SDK for Linux on Intel architecture, Java 2 Technology Edition, Version 1.4.1 |
| SuSE Linux 8.2 | Intel x86 | GTK | Sun Java 2 Standard Edition, version 1.4.2_03 for Linux x86 |
| SuSE Linux 8.2 | Intel x86 | GTK | IBM 32-bit SDK for Linux on Intel architecture, Java 2 Technology Edition, Version 1.4.1 |
| Sun Solaris 8 | SPARC | Motif | Sun Java 2 SDK, Standard Edition, 1.4.2_03 for Solaris SPARC |
| HP HP-UX 11i | hp9000 PA-RISC |
Motif | HP-UX SDK for the Java 2 platform, version 1.4.2.00 for hp9000 PA-RISC |
| IBM AIX 5L Version 5.2 | PowerPC | Motif |
IBM 32-bit SDK for AIX, Java 2 Technology Edition, Version 1.4.1 |
| Apple Mac OS X 10.3 | PowerPC | Carbon | Java 2 Standard Edition
1.4.2 for Mac OS X |
Although untested, Eclipse should work fine on other OSes that support the same window system. For Win32: Windows 98, ME, NT, 2000, and Server 2003; SWT HTML viewer requires Internet Explorer 5 (or higher). For GTK on other Linux systems: version 2.2.1 of the GTK+ widget toolkit and associated librares (GLib, Pango); SWT HTML viewer requires Mozilla 1.4GTK2. For Motif on other Linux systems: Open Motif 2.1 (included); SWT HTML viewer requires Mozilla 1.4GTK2.
An early access version of Eclipse is also available for 64-bit Linux GTK. Testing has been limited to early access 64-bit J2SEs running on AMD64 processors.
SWT is also supported on the QNX Neutrino operating system, x86 processor, Photon window system, and IBM J9 VM version 2.0. Eclipse 3.1 on Windows or Linux can be used cross develop QNX applications. (Eclipse 3.1 is unavailable on QNX because there is currently no 1.4 J2SE for QNX.)
The Eclipse Platform is designed as the basis for internationalized products. The user interface elements provided by the Eclipse SDK components, including dialogs and error messages, are externalized. The English strings are provided as the default resource bundles.
Latin-1 locales are supported by the Eclipse SDK on all of the above operating environments; DBCS locales are supported by the Eclipse SDK on the Windows, GTK, and Motif window systems; BIDI locales are supported by the Eclipse SDK only on Windows operating environments.
The Eclipse SDK supports GB 18030, the new Chinese code page standard, on Windows XP and 2000, and Linux.
German and Japanese locales are tested.
SWT fully supports BIDI on Windows (only). On Linux GTK, SWT supports entering and displaying BIDI text.
The Eclipse SDK is a development environment targeted at technical professionals - not an end user application. However, the Eclipse SDK tools will permit technical professionals who are working in English to build Hebrew/Arabic end user Java programs which are themselves not based on the Eclipse SDK. The BIDI support in the Eclipse SDK allows a Java programmer to work with BIDI strings, code comments, etc. but the Eclipse SDK itself is not designed to be localized for BIDI locales and its widget orientation can not be changed.
Eclipse 3.1 will be compatible with Eclipse 3.0.
API Contract Compatibility: Eclipse SDK 3.1 will be upwards contract-compatible with Eclipse SDK 3.0 except in those areas noted in the Eclipse 3.1 Plug-in Migration Guide. Programs that use affected APIs and extension points will need to be ported to Eclipse SDK 3.1 APIs. Downward contract compatibility is not supported. There is no guarantee that compliance with Eclipse SDK 3.1 APIs would ensure compliance with Eclipse SDK 3.0 APIs. Refer to Evolving Java-based APIs for a discussion of the kinds of API changes that maintain contract compatibility.
Binary (plug-in) Compatibility: Eclipse SDK 3.1 will be upwards binary-compatible with Eclipse SDK 3.0 except in those areas noted in the Eclipse 3.1 Plug-in Migration Guide. Downward plug-in compatibility is not supported. Plug-ins for Eclipse SDK 3.1 will not be usable in Eclipse SDK 3.0. Refer to Evolving Java-based APIs for a discussion of the kinds of API changes that maintain binary compatibility.
Source Compatibility: Eclipse SDK 3.1 will be upwards source-compatible with Eclipse SDK 3.0 except in the areas noted in the Eclipse 3.1 Plug-in Migration Guide. This means that source files written to use Eclipse SDK 3.0 APIs might successfully compile and run against Eclipse SDK 3.1 APIs, although this is not guaranteed. Downward source compatibility is not supported. If source files use new Eclipse SDK APIs, they will not be usable with an earlier version of the Eclipse SDK.
Workspace Compatibility: Eclipse SDK 3.1 will be upwards workspace-compatible with Eclipse SDK 3.0 unless noted. This means that workspaces and projects created with Eclipse SDK 3.0 can be successfully opened by Eclipse SDK 3.1 and upgraded to a 3.1 workspace. This includes both hidden metadata, which is localized to a particular workspace, as well as metadata files found within a workspace project (e.g., the .project file), which may propagate between workspaces via file copying or team repositories. Individual plug-ins developed for Eclipse SDK 3.1 should provide similar upwards compatibility for their hidden and visible workspace metadata created by earlier versions; 3.1 plug-in developers are responsible for ensuring that their plug-ins recognize 3.1, 3.0, 2.1, and 2.0 metadata and process it appropriately. User interface session state may be discarded when a workspace is upgraded. Downward workspace compatibility is not supported. A workspace created (or opened) by a product based on Eclipse 3.1 will be unusable with a product based an earlier version of Eclipse. Visible metadata files created (or overwritten) by Eclipse 3.1 will generally be unusable with earlier versions of Eclipse.
Non-compliant usage of API's: All non-API methods and classes, and certainly everything in a package with "internal" in its name, are considered implementation details which may vary between operating environment and are subject to change without notice. Client plug-ins that directly depend on anything other than what is specified in the Eclipse SDK API are inherently unsupportable and receive no guarantees about compatibility within a single release much less with an earlier releases. Refer to How to Use the Eclipse API for information about how to write compliant plug-ins.
The changes under consideration for the next release of Eclipse Platform, JDT,
and PDE address a few major themes:
Since each theme has a number of items, the committed, proposed, and deferred plan items are grouped in sections by theme. In addition, there are a few important Eclipse Platform improvements that do not naturally fit into any of the above themes; these are listed separately.
Each of the 3 projects under the Eclipse top-level project is covered in its own section:For each project, the items listed reflect new features of Eclipse or areas where existing features will be significantly reworked. Each item indicates the components likely affected by that work item (many items involve coordinated changes to several components). Numbers in parentheses link to bugzilla problem reports for that plan item.
The Eclipse Platform provides the most fundamental building blocks. Plan items reflect new features of the Eclipse Platform, or areas where existing features will be significantly reworked.
( Note: Unlike plans for
earlier releases, this plan is starting off with only a small number of items.
Additional plan items will definitely be added in subsequent plan revisions as further
requirements come in from the community.)
None at this time.
Process improvements wrt performance. For 3.1 we should get a firm grip on speed and space performance. To do this we should (1) identify which aspects are critically important performance-wise; (2) establish benchmarks to quantify performance; and (3) look at ways to improve performance. By its nature, it's hard to say in advance what kind of improvements we'll be able to achieve, making it difficult to make performance improvements a release deliverable. Instead, what we need to do for 3.1 is commit to track performance from this point on. We should carry out performance reviews (1) and create benchmarks (2) for every component, and for any critical aspects that span components. Even if we cannot find ways to make significant performance gains (3), the benchmarks are still critical to Eclipse's success to prevent backsliding in the future. The work item is to improve our process with respect to performance. The deliverables are the automated performance tests, the build process that collects and publishes performance data covering all critical aspects of Eclipse, and the committment to leverage this data to monitor speed and space performance as part and parcel of developing Eclipse. [All Eclipse components] (71123) [Theme: Built to last] [Note: To get things started, the PMC has asked each component team to create their first performance test by M1.]
Overhaul preferences. The way that Eclipse deals with preferences should be improved. The workbench presentation of preferences generally uses a component-oriented hierarchy, which is not always the best choice for users. The UI should present preferences in such a way that users can readily discover them. Some preference settings have scopes (e.g., workspace-wide vs. per project). The UI should help the user to visualize these scopes. For team-wide uniformity, some preferences settings (e.g., compiler options) need to be shared (or imposed) across teams. The UI should also facilitate this. [Platform Core; Platform UI; Text; JDT UI; JDT Core] (71124) [Theme: Built to last; simple to use]
Logical resources. The Eclipse Platform supports a strong physical view of projects, files, and folders in the workspace. However, there are many situations where a physical view is not the most salient or useful for many purposes. In some cases, multiple distinct objects happen to be stored in a single file, like an archive. Conversely, in other cases, something that is logically a single object is stored across multiple files. This discrepancy between logical and physical creates problems for common operations such as searching, comparing, and versioning, which need to work in the physical realm. Eclipse should support some way of mapping between a logical view and the physical organization of files on disk. [Platform Core, Platform UI, Team] (37723) [Theme: Built to last; large-scale development]
Embedding editors and views. The current compare infrastructure requires a plug-in to create a separate text editor for viewing changes between files. This is due a limitation in the workbench that prevents editors from being embedded inside views or other controls. As a consequence, the compare view's editors don't support the full set of features provided by the corresponding real content-type-specific editor, and represents a major usability problem for users switching to the real editor to make changes and then returns to the compare view to review the changes. Eclipse should be more flexible in the ways various editors can be used and reused. [Platform UI, Compare, Text] (71125) [Theme: Simple to use]
Team policies. Teams of developers working together have special needs for ensuring the integrity and consistency of their collective work. For example, they may have policies for things like source file naming, code formatting, or bug report number for change tracking. Eclipse should enable tools that help achieve that integrity and consistency. Eclipse should provide more access points so that complex policies can be suggested, checked, or enforced. [Platform Core; Team] (71126) [Theme: Large-scale development]
Large-scale workspaces. Some large customers have workspaces containing several hundred projects. Scoped builds (added in 3.0) were a step in the right direction towards making these workable. Eclipse needs to provide appropriate support for large workspaces and working sets, and ensure that everything scales properly. [Platform Core; Platform UI] (71128) [Theme: Large-scale development]
None at this time.
(End of items for Eclipse Platform project.)
Java development tools (JDT) implements a Java IDE based on the Eclipse Platform. The following work items reflect new features of JDT, or areas where existing features will be significantly reworked.
Add support for J2SE 5 features. J2SE 5 (also known as JDK 1.5 "Tiger") is targeted to ship in the second half of 2004. This release contains new Java language features, including generic types, typesafe enumerations, autoboxing, enhanced for loops, static imports, varargs, and annotations. Supporting these new language and compiler features requires major changes to the Eclipse Java compiler, to JDT Core APIs, to JDT UI, and to JDT Debug. Eclipse should contain full support for developing Java programs that use the new language features, including code assist, quick fixes, new wizards, source actions, and refactorings. [JDT Core, JDT UI, JDT Text, JDT Debug] (36938) [Theme: J2SE 5 support] [Note: Since the code for Eclipse 3.1 will be based on J2SE 1.4, we will be unable to "eat our own dog food" in regards to the new J2SE 5 features. This means we'll need to consider how best to achieve high quality. For instance, we might want to develop an example plug-in written using J2SE 5 features.]
Improve program manipulation infrastructure. Parts of the Java model are still implemented in terms of the deprecated JDOM, an early precursor of the AST facility. JDT should move to an AST-based implementation of the Java model. All source and refactoring operations should be rewritten to use common program manipulation infrastructure based on ASTs and AST rewriting. AST creation in the presence of syntax errors should be improved to include partial bindings for the code that lexically precede the syntax error. The performance of AST creation should be improved by using a special parser and eliminating one internal layer of ASTs. There should also be better support for navigating between a node in an AST (or an AST binding) and the corresponding element in the Java model (which would benefit views that present type hierarchies). [JDT Core, JDT UI] (71129) [Theme: Built to last]
Snippet-based searching. Allow search patterns to contain code snippets that are to be further matched against the occurrence. For example allow searching for all method calls to IRunnableContext#run where the first argument is boolean 'false'. [JDT Core, JDT UI] (71130) [Theme: none]
None at this time.
(End of items for Eclipse JDT project.)
None at this time.
RCP support. PDE currently supports developing Eclipse plug-ins, but is not expressly geared for the specific needs of clients developing applications based on the RCP. PDE should facilitate developing and deploying RCP-based applications. (71131) [Theme: RCP]
None at this time.
(End of items for Eclipse PDE project.)