When i look at the libary i see the tree to my shared files or folders from my /home up. This kinda confuse me from which point on in the tree it is shared. I could look in advanced share but maybe it will be possible to color taint the folder name which is the shared folder or the beginning of a shared tree.
Copy the search topic from old search tabs to clipboard, when i want to repeat a search with more parameters.
Collections with a tree structure (I could only do flat collections with no depth)
I hope this will not break anything, but maybe it a possibility for a collection tree : to have collections in a collection
Autodownload option/flag in search for 'search by hash'
You mentioned that by downloading a file has/gets information about other sources, otherwise maybe a per file option for "find mor peers" is helpful.
Repeat search after x sc/min. This could make sense for:
finding new results ?
finding more sources for a filtered search ?
autoupdate collections
Some thoughts about autoupdate collections:
The biggest problem couldd be that the hash of the collections is a hash over the hashes of the files. So when they change the hash changes then the file is for muwire not the same anymore. I can't search for the hash repeatly and at the same time it gets filled with more files and still has the same hash. It would only download the old collection and had no clue about another. I hope i not talking trash here, but that's the way i undersstand it.
The collections have disscriptions. We could use the discrption to save a distinct string - a hash. We can search for this hash (maybe explicit in the discription). Let's say we created a collectins with n files. We put in the discription our hash. We search for this hash and get one collection of n files. Now the uploader put new files into his shared folder and creates another collection, maybe same name, but in the discription the same hash as in the collection before. We repaet the search. Now we will get 2 results, because both collections have the same hash in the discription. We have to find out which one is the new somehow. Because i don't think there is some date in the files themselves/search involved. But maybe there is some way or just destroy the old collection. Then download the new one/files automatic.
Maybe this could work.
It is even more thinkable.
This is kind of versioning. If i put in the search hash plus the hash of the collection in the discription , we could do more things. The next collection has the search hash, his own hash and after that the hash of collection before. We could repeat this for every new update (as long as there is space in the discription). first version of the collection is search hash and 1 hash of its own. Latest has search hash + highest number of hash entries of search results. You can even put i a date into it for every hash and the latest one would always had the information abut all his predecessors.
Thank you for your time and thanks for this great program to play with.