S3 has a feature where you can generate URL-tokens (hashes) to grant download access to a file not on the CDN. This allows the developer to keep more private and sensitive files off the CDN but grant download access to users without Cloud accounts.

Without this feature, developers must host these files themselves or create a download proxy script which defeats the bandwidth and performance benefits of the cloud.

Index/default pages support would allow one to access:

http://cloudfileserver/folder/

instead of:

http://cloudfileserver/folder/index.html

Amazon just announced the ability to have default documents and 404 documents on AWS S3, essentially allowing you to host a website from S3. You can read more about it here:

http://www.allthingsdistributed.com/2011/02/website_amazon_s3.html
http://docs.amazonwebservices.com/AmazonS3/latest/dev/WebsiteHosting.html

This is something I have been long wanting for, and it would make Rackspace Cloud Files great.

Please add the option to view cloud files bandwidth usage per bucket. This way we can have seperate clients in seperate buckets and work who is using what bandwidth (and charge accordingly).

Currently each bucket is only 1 folder deep, that isn't the best solution when you have hundreds of thousands of files. So a richer folder experience would be awesome.

My site has some 19000 small images which I'm offloading to the CDN. Total size about 80 Mbytes. It takes several hours to upload them using a script through the API, since a single transaction takes about 0.5 to 1 second and the files must be uploaded one by one. How about uploading a .tar or a .zip file in a single transaction through the API or the control panel, and have the Cloud Files server unpack the archive in the container? Would also reduce the load in the API service when uploading small files.

It would save my sanity if this seemingly simple function could be implemented in the Cloud Files container view. I can only assume the underlying infrastructure makes it technically very difficult to implement this, otherwise it surely would have been done when the Cloud Files container management UI was built.

The idea would be able to FTP/SFTP into our CDN here at Rackspace. With us having over 100 files and adding more each day this would greatly streamline our production line.

Display stats on the number of times a file has been downloaded, providing useful information should you wish to analyse popular files and requests. At the moment downloading log files and analysing through logs doesn't seem the right solution

Maybe something as simple insisting on an Referer header present matching a supplied regex (for expected host domains). That would stop casual hotlinking and bandwidth theft IMHO.

Allow CDN URL requests to specify max image width and height and send a resized copy (cached on Cloud Files for future requests).
That way we'd only need to submit one version of any image to the cloud, any size could be requested.
When the master image is deleted, all the resized versions get deleted too.

Amazon S3 supports browser-based direct uploads.

http://docs.amazonwebservices.com/AmazonS3/2006-03-01/dev/index.html?UsingHTTPPOST.html

This feature is extremely helpful for sites with tons of user generated media files. Currently our users need to upload their files first to our Slicehost servers (which are slow at handling such uploads) and then we push that content to our Cloud Files containers. However the user could have just uploaded the same content directly to Cloud Files in a much shorter time.

It would be great if you can add a similar functionality.

When using CNAMEs and SSL in Cloud Files, like "https://cdn.example.com", I would like to override the SSL certificate Akamai issues with a "verified" certificate specifically for the CNAME.

Please request upstream to CDN Akamai to provide the referrer data in the CDN access logs. This is important for a number of reasons.

Needed to determine traffic sources.

Used to determine if bandwidth stealing is taking place by direct linking to images hosted at rackspace.

Ability to limit outbound bandwidth on an account.

I'd like to use Cloud Files to provide some video streaming but I do not want to use bandwidth I can't afford.

The ability to apply a self imposed usage quota would be very useful to prevent overspending.

It would be great if we can set a custom origin to Cloud Files that was just a cloud server (or any host) rather than use Cloud files itself.

Perhaps as a fork of CloudFiles, or just as an additional feature, I'd like to have persistent storage available to any server via a 'mounting' interface.

... like Google doing it at Google Books, DOCs and PDF reader, Picasa (everybody used it, if pressed "Zoom in/out". Allow to build similar services.

It would be GREAT to have the ability to map network drives to a cloud files solution (or something similar). This would allow us to move most of our clients file server data into the RS Cloud.