May appens that an ebuild has a broken digest (some examples: Bug 147382, Bug 162872, Bug 162922); in this case, emerge delete suspicious file and try to download it again. It's frustrating (at least for a user with slow internet connection) to have to re-download a correct source file, so should be nice to have a new option (or a new feature) that stops emerge w/out deleting "suspicious" file and lets user to check if the file is corrupted or not.
As further suggestion, when a digest verification failure occurs, emerge can go on as usual, but instead of deleting suspicious file, it can move it to a default directory.
This kind of suggestions completely defeats the whole purpose of manifests. If this happens to you, then file a bug and get the digest fixed. How will the user "check if the file is corrupted or not"? That's what the checksum does.
fixed in sys-apps/portage-2.1.3_rc9 ChangeLog pym/portage.py 7322 zmedico If the fetcher reported success and the file is too small, don't attempt to resume. Upstream probably changed the distfile and we need to show a digest verification failure so the user gets a clue. (trunk r7321) 7326 zmedico When digest verification of a distfile fails, rename instead of unlinking it in case it's a large download and the user wants to salvage it due to a bad digest. (trunk r7325) thanks zmedico :)
