| Summary: | [launcher] Try an alternate guess at default ws | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | [Eclipse Project] Equinox | Reporter: | Andrew Niefer <aniefer> | ||||||
| Component: | Framework | Assignee: | Andrew Niefer <aniefer> | ||||||
| Status: | RESOLVED FIXED | QA Contact: | |||||||
| Severity: | normal | ||||||||
| Priority: | P3 | CC: | tjwatson | ||||||
| Version: | unspecified | ||||||||
| Target Milestone: | 3.6 M2 | ||||||||
| Hardware: | PC | ||||||||
| OS: | Windows XP | ||||||||
| Whiteboard: | |||||||||
| Attachments: |
|
||||||||
|
Description
Andrew Niefer
Created attachment 144316 [details]
patch
Patch also checks osgi.ws system property instead of just -ws. We only set the alternate ws to be used if we actually find a launcher fragment for it.
Thanks Andrew. I'll do some testing on the mac. Lets plan to release for M2. Created attachment 144435 [details] updated patch (In reply to comment #1) > Patch also checks osgi.ws system property instead of just -ws. Shouldn't we do this for -os and -arch also? This is an updated patch that checks for osgi.os and osgi.arch properties also. I tested this on a Mac cocoa eclipse installation and it correctly uses the alternate "cocoa" value for osgi.ws. (In reply to comment #3) > Shouldn't we do this for -os and -arch also? This is an updated patch that > checks for osgi.os and osgi.arch properties also. I didn't do os & arch because there is presumably only one possible answer for those. However, upon reflection, it is good to check them as it is conceivable for someone to port to some platform we don't know about where Main.getOS()/getArch() would have returned OS_UNKNOWN I have tried this with linux.motif and win32.wpf, both of which failed to launch in 3.5 with "java -jar" due to the wrong ws. They now start correctly. I have released this change to 3.6 head. |