Show HN: Send huge files, pause in the middle and continue
dropjar.com
dropjar.com
Here are some observations & Questions :
1. Looks like you decide the number of chunks the file is divided into before you start uploading. How do you decide this ? (ie) the algorithm used ..is the chunk size a function of network's bandwidth & throughput . Or is it a standard chunk size ?
flowChunkNumber:6 flowChunkSize:512000 <<===== This one here flowCurrentChunkSize:512000 flowTotalSize:5005687 flowIdentifier:5005687-emc_celera__Command_Reference_Manualpdf flowFilename:emc_celera__Command_Reference_Manual.pdf flowRelativePath:emc_celera__Command_Reference_Manual.pdf flowTotalChunks:9
2. How do you generate the random file name ? I suppose this is one the standard algorithms. Curious
3.I suppose you are using statcounter & google analytics. What are some of the parameters you are looking at your end? (like #of files uploaded etc., )
4. Even if I change the name of the file (or) move the file to another location when the upload is inflight ; it still sucessfully continues to upload the originally intended file.Are you precaching this somewhere ? Is there a copy stored in the temp folder ?
5. What are you using to compress the files before they start upload ?
1. Right now just chose something that made sense, need to test more 2. just randomly choose 7 chars a..zA..Z0..9 3. yes using both. right now still not looking into metrics 4. not sure, i'm using flowjs which is a library that supports this need to test further 5. I don't think files are compressed before upload.
I think this is far more useful for these kinds of "one-off" shares than going through a series of third-party gates like Dropbox for example, especially if the intended recipients don't have accounts on those sites.
Running Chrome 36.0.1985.143 on OS X