whitelist as alternative to /home tree traversal #45

New Issue

cmccabe · 2019-12-05T20:42:27Z

cmccabe commented

2019-12-05 20:42:27 +00:00

This follows from issue #25.

On very large systems, it is too slow for linkulator to traverse home directory trees. An alternative is to allow users to maintain a list of the fellow users that they want to use linkulator with. So, rather than looping through all the linkulator.data files found through globbing, it linkulator would just loop through addresses in the whitelist file.

The risk to this approach is that each person may only know a few other linkulator

users on a system, and their whitelist will be incomplete. This will mean that link/reply threads will have gaps. The solution to this is for linkulator to not only search for linkulator.data files, but also to scoop up data from friends' whitelist files and to update ones own file with that data. There would be a bit of recursion in adding locations to the whitelist, and then adding locations found from within those new locations, but this is doable.

And if we ever want to get really ambitious, a future approach could be to allow urls from other servers in the whitelist file, so that you would aggregate link/reply threads from across the internet using urllib2.urlopen().

This follows from issue #25. On very large systems, it is too slow for linkulator to traverse home directory trees. An alternative is to allow users to maintain a list of the fellow users that they want to use linkulator with. So, rather than looping through all the linkulator.data files found through globbing, it linkulator would just loop through addresses in the whitelist file. The risk to this approach is that each person may only know a few other linkulator users on a system, and their whitelist will be incomplete. This will mean that link/reply threads will have gaps. The solution to this is for linkulator to not only search for linkulator.data files, but also to scoop up data from friends' whitelist files and to update ones own file with that data. There would be a bit of recursion in adding locations to the whitelist, and then adding locations found from within those new locations, but this is doable. And if we ever want to get really ambitious, a future approach could be to allow urls from other servers in the whitelist file, so that you would aggregate link/reply threads from across the internet using urllib2.urlopen().

cmccabe added the

future release

label 2019-12-05 20:42:40 +00:00

cmccabe referenced this issue

2019-12-06 02:22:52 +00:00

Allow alternate /home paths #25

sloum added the

enhancement

label 2020-01-16 19:46:32 +00:00

Sign in to join this conversation.