restarting download

[ulu]

Is it possible to restart a download without losing what's already downloaded?

I'm downloading software for a midi device I just bought-internet connection here is terrible so it took 8hrs to download 4GB . Got to 95% downloaded, then the connection was lost.

Connection started up again 10 minutes later, but the download is stuck at 95%. I can start new downloads, but this one's not moving.

Is there a way to get it started again so it will finish?

windows 7

 

[High Fidelity]

It used to be possible I think, but if you've restarted then it's no doubt been removed from temporary storage and you'll need to start again.

Sorry to say, but either way, it's looking like you'll need to do it over again.

 

[dysert]

Since it looks like you'll be starting over, I suggest using a download tool like BitTorrent. That way, if it fails partway through you will be able to pick up where it left off.

 

[paul1149]

I don't know about rescuing what you've already downloaded, as you'd have to get your download prog to recognize it before it would resume. But starting afresh, any download manager should Resume a stalled download. I use Opera 12, which has one built in. For large downloads, I wouldn't operate without a download manager.

 

[Qyöt27]

Since it looks like you'll be starting over, I suggest using a download tool like BitTorrent. That way, if it fails partway through you will be able to pick up where it left off.

BitTorrent gets away with being able to resume at will because that functionality is part of the protocol itself (and because BT isn't a 'download tool' as such, it's an entirely separate download network), and the blocks are SHA-1 hashed for integrity. Also that BT clients aren't supposed to be used as sequential downloaders, so the downloaded blocks are filling in many disparate parts of the file.

HTTP is usually not resumable like that without the download client including special support; most browser-based ones are notoriously fragile about this, but some clients (like wget and curl) can do it. It may also be a case that the server itself needs to support the functionality, and there's also the threat that even if you can manage to resume, the join point might have been corrupted (if the download has an associated CRC32, MD5, or SHA-1 hash provided, then the user can check the integrity themselves and make sure nothing weird happened, but not everyone provides those).


The simplest trick if it's a matter of network congestion is to pause the halted download and then try to unpause it. But if this doesn't work, you're out of luck and need to start over again; or if they provide non-HTTP download options, use those instead.

 

[EphesiaNZ]

Never had trouble with wget on http/ftp downloads and it's usually my preferred method for regular file downloads using the command line.

Uget is better if you require a windowed GUI interface and it's also available for Win, Mac and Linux.

 

[Sketcher]

It's got to be built into the application you're using to download (i.e. Steam, your web browser, etc). How well it works depends on a number of factors beyond what you can control. Compare the file sizes of the downloaded file and the claimed size of the file you are downloading. If md5sums are available for the file you are downloading, check the md5sum of that file to see if it matches (winmd5sum will do that on Windows).