@pepa65 commented on GitHub:
I think having the uploader specify a secret at upload-time is best. Uploaders that think they might want to delete a file at some point need to do that. If it was an…
@Anachron commented on GitHub:
Ping @aspacca I've heard that you are the new owner of this repo.
@pepa65 commented on GitHub:
Using the Max-Days header is a good idea. But sometimes if you change your mind, it is good to have the option to delete a file (in a way that others can't delete…
@monkz commented on GitHub:
I would recommend a second url for deletion (independent of the download link), otherwise the -X DELETE variant would be accessible for annoying people, that delete…
@monkz commented on GitHub:
And if you return two urls - it would break the current API.
@rdeavila commented on GitHub:
I agree with @monkz. When we send a file, the site returns the file URL like https://transfer.sh/<random_code>/<filename>, right?
Why not return two URLs, one…
@nl5887 commented on GitHub:
Exactly why we haven't implemented it yet. I can imagine returning an extra delete header, but that won't be visible for most people. Especially not the people that…
@golbiga commented on GitHub:
The fix is in master, please test. Closing issue.
@golbiga commented on GitHub:
This has been fixed in the 0.9_prerelease branch, it will be included in the release.
Thanks for reporting it.
@golbiga commented on GitHub:
@peterlewis, one thing to note. The output will still be in the mobileconfigs folder. There will just be an additional folder within, based on the selected…
@monkz commented on GitHub:
@Anachron: interesting approach - making the API enhancement optional! I was total fix on default API behaviour - my bad.
We've already parameters for max downloads…
@peterlewis commented on GitHub:
That's great. Thanks, @golbiga, for sorting that out so quickly!
@nl5887 commented on GitHub:
Actually I found out that the difference between 1.4.2 and 1.6 in SSL handling was causing the issue. I was reading and writing to S3 with https, causing the cpu to…
@robertgendler commented on GitHub:
The script has been removed in our current release. So this has been resolved.