Changes between Initial Version and Version 1 of Ticket #87, comment 8
- Timestamp:
- Oct 22, 2012, 2:48:20 PM (12 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
Ticket #87, comment 8
initial v1 1 Made another attempt to fix this in r . Now the old behavior mentioned in comment:7 is restored. The current logic is to only add '.EXE' if there is no (any) extension already specified, regardless of wether there is a path component or not. This in particular means that 'path/name' will be resolved to 'path/name.exe' while 'path/name.myext' will be left intact and loaded as given. This also means that if the real exe has no extension at all, you won't be able to start it with PE/PEC -- but as I've just checked, this matches the behavior of CMD.EXE on both OS/2 and Windows.1 Made another attempt to fix this in r22028. Now the old behavior mentioned in comment:7 is restored. The current logic is to only add '.EXE' if there is no (any) extension already specified, regardless of wether there is a path component or not. This in particular means that 'path/name' will be resolved to 'path/name.exe' while 'path/name.myext' will be left intact and loaded as given. This also means that if the real exe has no extension at all, you won't be able to start it with PE/PEC -- but as I've just checked, this matches the behavior of CMD.EXE on both OS/2 and Windows. 2 2 3 3 So only the low memory condition is an issue for me now. Looking at it.