Do not overwrite local modifications with updates #1463

arantius opened this Issue Nov 3, 2011 · 3 comments


None yet
2 participants

arantius commented Nov 3, 2011

Please discuss. We already know the file-modified time and the last update check time. We should be able to use these values to determine that the script has changed on disk since the last update (/install ?) and choose not to overwrite it with remote updates.

Is this important? Is this foolproof? Is this a good idea?


johan commented Nov 3, 2011

My thoughts are "yes", "workable" and "yes".

File-modified times are probably not idiot-proof, but should at least cater the 90%+ case, and likely even 99%. (I know that access-time updates is sometimes disable on file systems for disk or battery performance but haven't heard of similar things for modified-time.)


arantius commented Nov 4, 2011

Having slept on it, we need to track one more bit of information, I think. If install time (currently not tracked) === modified time, updates are always safe to install. If modified time > install time, then updates should not be installed. Last update check time doesn't really factor into things.


arantius commented Feb 6, 2012

New issue for tracking install time logged for v.Next, using that info (for updates, also pushed) pushed to v.Nextnext.

arantius closed this in ca9621f Feb 10, 2012

arantius reopened this Feb 11, 2012

arantius closed this in fb16949 Feb 16, 2012

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment