|
From: | Mike Miller |
Subject: | [Octave-bug-tracker] [bug #57202] missing bit-compatible MSVCR100.dll reports as 'java not found' |
Date: | Thu, 14 Nov 2019 20:05:29 -0500 (EST) |
User-agent: | Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/78.0.3904.97 Safari/537.36 |
Follow-up Comment #5, bug #57202 (project octave): Right, it should be possible to improve octave_w32_shlib to test for the existence of the file separately from the call to LoadLibrary, and return a more accurate error message to indicate whether the file is truly not found or whether it could not be loaded for another reason. I think that would help improve the ambiguity of the current error message you get from javaMethod: "could not find library or dependencies". That's the first point. I agree that javachk should not be extended. Calling it with the feature name "jvm" is the simplest documented method to test for whether Java exists and works or not. It's possible we could add more expressive error messages to that function as well. That's the second point, how to best propagate error messages that are useful for the user who calls "version -java" and "javachk" and wants to know how to diagnose it when Java isn't found. _______________________________________________________ Reply to this item at: <https://savannah.gnu.org/bugs/?57202> _______________________________________________ Message sent via Savannah https://savannah.gnu.org/
[Prev in Thread] | Current Thread | [Next in Thread] |