Bug 426657

Summary:	[1.8] Presentation for long form of error messages
Product:	[Eclipse Project] JDT	Reporter:	Stephan Herrmann <stephan.herrmann>
Component:	UI	Assignee:	JDT-UI-Inbox <jdt-ui-inbox>
Status:	ASSIGNED ---	QA Contact:
Severity:	enhancement
Priority:	P3	CC:	daniel_megert, gautier.desaintmartinlacaze, markus.kell.r, srikanth_sankaran
Version:	4.3.1
Target Milestone:	---
Hardware:	All
OS:	All
Whiteboard:
Bug Depends on:
Bug Blocks:	404675

Description Stephan Herrmann

2014-01-25 19:50:15 EST

When I'll address bug 404675 it will probably be a good idea to offer a short and a long version of some errors. See that javac has s.t. similar with its -Xdiags:verbose switch, but I think in the IDE we can do even better :)

It's not that I have an implementation of explanatory error messages up my sleeve, yet, but the internal InferenceContext18 actually has all the information for given multi-page error messages ...

WDYT?

Comment 1 Dani Megert

2014-04-18 03:36:34 EDT

(In reply to Stephan Herrmann from comment #0)
> When I'll address bug 404675 it will probably be a good idea to offer a
> short and a long version of some errors. See that javac has s.t. similar
> with its -Xdiags:verbose switch, but I think in the IDE we can do even
> better :)

What exactly do you mean by "do even better"? Create something in the middle of short and long? ;-)

Comment 2 Stephan Herrmann

2014-04-19 12:22:11 EDT

(In reply to Dani Megert from comment #1)
> What exactly do you mean by "do even better"? Create something in the middle
> of short and long? ;-)

Create something more interactive than a long unreadable error message.
At least a tree structure that you can drill into to get more details on demand.

Why not an extra view with some sync'ed highlighting so you can select individual AST nodes and inspect what inference "thought" about them, which nodes can be typed, and where inference found a conflict etc.

The deluxe version of this could be called a typing-debugger, a tool that lets you experiment and inspect.