Cache Folder To Store The *.!ut Files In. Once Complete, Move To Correct Location.

[AetherMichael]

I would be interested in a feature that would allow me to set a download location maybe specifically or some "Default" folder in the OS to download files to....  then as the downloading files complete,  it moves them to the folder structure in the place they should be. 

ISSUE #1
There are a couple issues being addressed here....  One is that "IF I see the file,  its complete."  I have had a couple instances where it was done syncing,  and just hasnt reindexed yet,  and had not completed each file download. Meaning I dont take my laptop to the office thinking its done syncing only to find out there is a file there,  its just a MYFILENAME.!ut file and not yet complete functional file.    I want to be able to,  "IF I see the file there, its complete"

ISSUE #2
The second issue is INDEXING.   I have indexed huge file structures in the past,  in the process one PC indexed files that were already *.!ut files as files that were not supposed to be indexed. These files were still in the process of being downloaded and were incomplete.  I am unsure why,  There are several fixes for this.  For example,  You can use only one index file located on one PC for all sharing PCs to copy from.  This can help solve the problem,   You can possibly just make a default exclusion for all software created files.  For example, Make a rule to never index *.!ut and any other software fabricated file ending.
I would not mind personally having a cache file that files just download into,  so that the *.!ut files are not located inside the actual Indexable file structure.  Keeping the created files separate from the Indexable files will guarantee that none of my files will be damaged in the future.  due to incorrect indexing.

ISSUE 3#
Fragmentation -  Just want the files to be one contiginous file on disk.  That is all.
(maybe just reserve space on disk ahead of time?)

So to solve all issues with one stone,  I think a cache folder would do the trick best!  This will let me know for sure that if I see it there,  its complete.  and also, this will keep non-bittorrent related files out of my file structure preventing fatal data errors and losses, and finally,  the move of one complete file into its folder can if done right handle the fragmentation issue as well.



Mike!

Edited February 11, 2015 by AetherMichael

[Moe]

ISSUE #1

There are a couple issues being addressed here....  One is that "IF I see the file,  its complete."  I have had a couple instances where it was done syncing,  and just hasnt reindexed yet,  and had not completed each file download. Meaning I dont take my laptop to the office thinking its done syncing only to find out there is a file there,  its just a MYFILENAME.!ut file and not yet complete functional file.    I want to be able to,  "IF I see the file there, its complete"

 

If you get into the office and your file has not been fully synced it would just be starting to sync as you connect to the internet again. I don't see the point in having a "cache" or more like a "incomplete" folder.

Are you experiencing these issues a lot that BitTorrent Sync GUI says it's done syncing and the actual file is not being present in said share? Maybe a debug log sent to the devs could help fix an issue if there is one.

 

The rest sounds very like specific requests. The only way I can think of that this would make it in BitTorrent Sync is when running with an API key in a config file. Normal user GUI mode - I am not so sure about that.

[AetherMichael]

Ya its like I just said in the above post.  The file was there.  It was just a FILENAME.!ut    Meaning it was only partially downloaded.  Was an incomplete file.   I didnt indicate anywhere in that post that the actual file is not being present in said share.  I have no idea where you got that from at all.

One of the issues I have is timing.    The software itself has timing issues I Think.   PC1 Gets a whole lot of new files in the folder.  It then indexes and takes a while doing it. Once done,  it begins distributing the files, 
PC2  Is downloading normally.  PC2 needs a restart.  Once booted back up,  PC2 Begins INDEXING again. it then indexes and has some incomplete files that were originally named FILE1.ISO and FILE2.IMG but now are named FILE1.!ut and FILE2.!ut and so on in them.  But the second pc, PC2 just indexes them as if they are supposed to be called that.   The data is also incomplete in them.   PC2 HAVING THE NEWER INDEX LIST then shares the broken files named FILE1.!ut and FILE2.!ut back to PC1,  Corrupting the data completely and destroying the files forever.  This happened multiple times to me already.

My problem is that Bittorrent Sync is RENAMING files while it downloads them.  For example, I have a file named Windows.ISO   This file is 700mb.
I start downloading it.  The file is not named Windows.ISO,  but instead Windows.!bt and put in the folder it needs to be in.  The problem lies when the second PC is in the middle of the download.  If it is shutdown.  Restarted,  then it comes back up and starts Re-Indexing.   It will index the file named Windows.!bt and because its newer index list is newer than the first PC,  It will send the Corrupted incomplete file back to the first PC and destroy the original as well.  Meaning the entire file is lost forever from both PCs.

The reason I suggested a cache folder,  is because I personally want to start maintaining the actual file structure integrity.  I am not ok with large files that are syncing being Renamed into sometihng with the WRONG file extension and run the risk of it being indexed that way by the second computer again.   Cache folder will make sure that the data being downloaded with those bittorrent specific endings DO NOT END UP IN MY FILE STRUCTURE AND GET INDEXED by other PCs.  That is the point.    

This seems to be only a big issue when syncing 670,000 files over 937GB of data.   Syncing is a process that occours over the next week or so this way.