Direct news from RapidShare on 22 dec.2006
Our linux department was on holidays for two days with the order to program two OpenSource upload-scripts. The target audience is the professional linux user, who is searching for comfortable ways to upload files to RapidShare. For the first time we have used the new RS API, which contains numerous valuable functions making it possible to write comfortable and feature-rich upload-managers for RapidShare. For example its now possible to resume aborted uploads including a MD5-check after the upload is completed to ensure the upload worked as it should. Two scripts have been created. The first script, which you can download here already contains everything you need to upload files as a free/collector's/premium-user. Just the upload-resume is missing here. However, this script is rather well documented and easier to understand than the second script. The second script is based on the first script and extended by the new upload-resume. This feature makes the second script more complex than the first one. You can download the second script here. Please notice that these scripts are created for the professional computer user. If you need help with this scripts, please talk to computer experts you know. Unfortunately we cannot provide support for these scripts. We hope that our windows department will include the new RS API in their programs within the next weeks so that upload resume is possible there as well. Since the RS API functions are available for everyone, we hope that we will see some nice upload managers with upload resume very soon. We will try to contact the programmers of the best upload managers and implement the programs in our website. If you have suggestions regarding our scripts, feel free to contact our support. Drop us a mail as well if you want some additional RS API functions for your programs. Maybe we will implement it.
1 comment:
I've been looking at the rsapiresume.pl script and I've been using it successfully. The API seems to have a feature missing.
I modified the script to test the checksums after each block is uploaded. This is to not waste time uploading the rest of the file if a block is corrupted.
However, the first block uploaded does not return a checksum. This means you have to upload an entire file before you know if the upload worked if the *first* block was corrupted. Subsequent blocks can be checked ok.
Perhaps the API could be extended to provide the MD5 checksum of the first block uploaded in some backwards compatible way.
Another useful way to save bandwidth would be to extend the protocol so that if the upload client detects that a sent block was corrupted, there is a way to resend the block rather than start sending the file all over again from the beginning.
Post a Comment