Improving logic so incomplete downloads aren't cached

[karagraysen]

🌈

• I've read and understood the *CONTRIBUTING guidelines and have done my best effort to follow.

A Realm user filed an issue on Realm-Cocoa explaining that one of our files (collection_notifications.hpp) could not be found. The specific issue can be found here: realm/realm-swift#4614.

To summarize it:

• The user has CocoaPods version 1.1.1 installed.
• pod update failed with this message:

Installing Realm (2.4.2)
[!] /bin/bash -c
set -e 
sh build.sh cocoapods-setup

core is not a symlink. Deleting...
Downloading dependency: sync 1.0.1
Downloading sync failed:
curl: (56) SSLRead() return error -36

• The user claims their internet connection was unstable at the time.
• The user tried again later when their connection was stable but the files weren't fixed.
• The user mentioned the following script fixed the issue for them:

#!/usr/bin/env bash

rm -rf "${HOME}/Library/Caches/CocoaPods"
rm -rf "`pwd`/Pods/"
pod update

I'm not experienced in this area but is it possible for CocoaPods to improve the logic where downloads occur so incomplete downloads aren't cached in a problematic way?

Please let me know if there's anything else I can provide. ❤️

[benasher44]

@istx25 thanks for the report! This definitely sounds like a bad cache type of issue. I can't guarantee when someone will get to fixing this, but I'll make sure that this issue is labeled appropriately!

[stale]

There hasn't been any activity on this issue recently. Due to the high number of incoming GitHub notifications, we have to clean some of the old issues, as many of them have already been resolved with the latest updates.

[karagraysen]

There's nothing more for me to add. All information should be there and it's just in backlog until someone in the CocoaPods team can get around to fix it.

[dnkoutso]

@istx25 was that part of executing a prepare_command or downloading the actual source (zip perhaps) of the pod?

The reason I am asking is because this issue exists #4587 which might be similar and if solved will solved this issue here as well.

[karagraysen]

I'm not sure. I'd recommend looking at the original issue on the Realm organization that I linked above. I no longer work at Realm so if you need more information, I recommend reaching out on the original issue and trying to talk to others who work there. Cheers

[dnkoutso]

@istx25 thanks, from checking the podspec this seems the same issue where the prepare_command failed and possibly continued and so it never ran again.