509191 - Resuming downloads relies on the cache, not on the download files

Status: UNCONFIRMED

(Toolkit :: Downloads API, defect)

Description

[Rolf Bode-Meyer]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; de; rv:1.9.1.3pre) Gecko/20090804 SeaMonkey/2.0b2pre
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; de; rv:1.9.1.3pre) Gecko/20090804 SeaMonkey/2.0b2pre

After a download is interrupted, resume completely relies on the cache. If cache was cleared in the meantime or the particular entry was removed due to cache full, download starts again at 0 bytes.
Shouldn't resume use the two download files for where to continue?

Reproducible: Always

Steps to Reproduce:
1. Download a file (e.g. a nightly from http://ftp.eu.mozilla.org/pub/mozilla.org/seamonkey/nightly/latest-comm-1.9.1/). Two files are created, the download file (.zip) itself and an additional .part file.
2. During download: .zip 0 MB, .zip.part x MB
3. After cutting connection using personal firewall: .zip 0 MB, .zip.part 3.9 MB
4. Clear cache
5. Resume download
Actual Results:  
Download starts again at 0 bytes

Expected Results:  
Download starts at 3.9 MB.