Free file hosting software for linux

AlternC is a set of user-friendly hosting software. AlternC Hosting Software is easy to install and based on open-source software only.
AlternC includes an automatic installation and configuration system, and a web-based control panel to manage users accounts and web services (e.g. domains, emails, ftp accounts, statistics...)
AlternC is based on the Debian GNU/Linux system (Woody version), and it requires other softwares such as Apache, Postfix, Mailman ...
It also contains a documented API, so you can customize your web panel quickly and easily.
AlternC was initially written in French. However, the debian package includes an English version. The translation of the project into other languages such as German and Spanish is in process. Volunteers are welcome ! (See the internationalization page). The documentation is only available in French (so far).
AlternC has been created by the system administrators of Lautre Net, members of Lautre Net and Globenet.
Enhancements:
- This release continues the Debian compliance challenge.
- Quota management was corrected, bind now forbids recursion and domain poisoning, and the Debian package uses po-debconf.
- New functions were also added, such as postfix saslauthd with rimap and MySQL user rights managment in the Web desktop.
- The .eu TLD is now properly handled.

Hosting Backup provides a set of scripts for backing up your hosted Web sites.
Hosting Backup is a set if PHP command line scripts that provides an easy way to backup hosted Web sites to your local computer. It works with sites hosted on Linux servers with MySQL databases using FTP.
HostingBackup uses lftp (http://lftp.yar.ru/) to download all the files from your remote server.
HostingBackup uses rc4 encryption, so you do not need to have PHP compiled with mcrypt.
HostingBackup is tested on Linux local computer with PHP 4.3.x and lftp 3.0.x.
The remote component will manage some operations like optimization and dump of MySQL databases.
The local component will manage the execution of backups and other utilities like the generation of crypted passwords and the automatic installation of the remote component.
To allow an easy way to backup all hosted MySQL databases we store the main password for the account directly on the remote server, crypted with rc4.
When we call the dump script we pass the decrypt key.
To catch the login data an hacker should read the crypted password written on the server and intercept the key on the net.
Enhancements:
- Add: "--comments=false" to mysqldump
- Add: "set FOREIGN_KEY_CHECKS=0;" to mysqldump
- Add: "--add-drop-table" to mysqldump

file*HANDLER project is primarily a Perl script which coordinates some free media conversion packages (cited later) with a PostgreSQL back end accessing server to index and serve binary and text files direct from the database. A simple CGI file with embedded AJAX makes calls to this server which is also constantly looking for other alive f*H servers on the Internet.
Since the server caches media into the SQL database on demand, as the network grows, the network improves.
Its gridded directory sharing/browsing/searching with streaming audio/video as well as flat text/doc/pdf/image display for everyone. Its written with a few hooks for tags that would be included in your actual front page so that the UI is discard-able -- anyone can quickly rewrite a whole new [GT]UI without having to worry about the syntax of the newest version of dojo.licio.r or whatever.
If you wanted to ignore the JS/HTML/CSS hooks then you can easily use the system to make direct requests that just return lists formatted as HTML table-bodies. In other words, the markup IS the markup.
As such, Ive whipped up a Dojo 0.2 Widget that coordinates the serving backend with a UI so anyone can embed f*H functionality anywhere, or easily customize a provided default page.
A file*HANDLER server is really a few constituent parts Ive tied up for you (top down):
- A local web page providing the UI(served by an HTTP server of your choice) that is generated by a cgi script with embedded AJAX.
- A secondary portion of the same CGI script, acting as middle-ware, which communicates, via AJAX, with the local front page to reconcile asynchronous JavaScript requests with the file*HANDLER sub-network back-end.
- An always-on network server written in PERL that serves to the front and end communicates laterally with everyone elses file*HANDLER back-end PERL server additionally, it manages indexing of content directories you choose to serve.
- A PostgreSQL database that is accessed only via internal PERL routines called from your front page.
So for example, a remote user comes to your site. First, not only can they browse and search your files, but they can also browse and search the files of anyone else hosting a file*HANDLER server that your local server knows about. (file*HANDLER identifies other servers on the network automatically). The user can now read/view/listen/watch by stream any content they find from whomevers server. Theres no download, so theres so actual sharing, just direct streaming to the users browser.

File::LockDir is a Perl basic filename-level lock utility.

SYNOPSIS

use File::LockDir;

INTERFACE

new

Initializes the class. Returns the singleton object.

nflock($file, $nap_till, $locker, $lockhost)

Locks the supplied filename. Only $file is required.

$file is the file to be locked; $nap_till is the total amount of time to wait before giving up; $locker is a name identifying the locker; $lockhost is the host requesting the lock.

nunflock($file)

Unlocks the supplied file.

nlock_state($file)

Checks the state of the lock for the supplied file. Returns a list: the first item is true if the file is unlocked, and false if not; the second item is undef if the file is unlocked, and the identity (name and host) is it is locked.

DIAGNOSTICS

%s already locked

Seen when youve already locked the requested pathname. Informational only.

No pathname to be locked

You didnt supply a pathname to be locked to nflock. Fatal.

cant write to directory of %s

The directory where the file resides cant be written, so the lockfile cant be created.

cant get %s: %s

The named lock cant be gotten: the reason is supplied. Failure occurs after ten tries to get the lock.

%s %s[%s]: lock on %s held by %s

The lock on the specified file is help by the noted locker. Informatory message, printed only when debugging is on.

close failed for %s: %s

The file containing the lock information couldnt be closed for the reason shown.

releasing lock on %s

Debug message; notes that the lock on the specified file was successfully released.

LSF::Hosts is a Perl module to retrieve information about LSF hosts.

SYNOPSIS

use LSF::Hosts;
use LSF::Hosts RaiseError => 0, PrintError => 1, PrintOutput => 0;
($hinfo) = LSF::Hosts->new( [HOST_NAME] );
@hosts = LSF::Hosts->new();

LSF::Hosts is a wrapper arround the LSF bhosts command used to obtain information about lsf hosts. The hash keys of the object are LSF bhosts header values. See the bhosts man page for more information.

CONSTRUCTOR

new( [ [HOST_NAME] ] );

With a valid hostname, creates a new LSF::Hosts object. Without a hostname returns a list of LSF::Hosts objects for all the hosts in the system. Takes no arguments (jet).

Host on Imageshack is a service menu tor upload images on Imageshack.

The tarball contains the .desktop file, the original script and an icon (modified version of Azureus icon).

Based on imageshack_upload.sh -> http://www.bchesneau.info/articles/2006/10/23/bash-script-to-upload-image-to-imageshack

The tarball content:

#NAME #CONTENT #DESTINATION

1.imageshack.desktop the servicemenu file ~/.kde/share/apps/{konqueror|dolphin}/servicemenu
2.imageshack.png an icon for the menu ~/.kde/share/icons
3.imageshack_upload the original bash script ~/bin
4.README this file

Fast File Search is a crawler for FTP servers and SMB shares that can be found on Windows or UNIX systems running Samba.
It provides a web interface for searching files. It is optimized for searching files by a wildcard when there are some normal (not * or ?) chars specified in the beginning or in the end of the mask (for example *.iso).
Fast File Search crawler runs on UNIX (currently only Linux has been tested but I do not now any reasons why it should not work on other UNIXes). Fast File Search uses MySQL database, web interface needs a web server with PHP >= 4.0.3 and the crawler needs some perl modules.
The crawler (ffsearch.pl) crawls the network (FTP servers from the list and all reachable SMB hosts on the local network) and stores the information about files into database. It is invoked at certain times each day via crontab entries.
There are two modes of operation of the crawler: complete crawl and incremental crawl. The crawler expects a command line argument that tells crawler which mode to run (-c or --complete for complete crawl, -i or --incremental for incremental crawl). Both modes retrieve a list of the active SMB hosts in all workgroups.
The complete crawl tries to scan all active hosts and all hosts that are listed in database. The complete crawl should be run once a day.
The incremental crawl tries to scan active hosts and hosts listed in database that have not been scanned since the last complete crawl because they were unreachable. The incremental crawl should be run several times a day, for example each 3 hours.
How does the crawler get know whether the host has been crawled since the last complete crawl?
Each time the complete crawl is executed, the expire count is incremented first. When the host is crawled, expire count is set to zero. So all hosts whose expire count > 0 were not reachable since the last complete crawl. Moreover, when expire count reaches value specified in configuration (i.e. it was unreachable during the time period of complete crawls) the information about files on the "expired" host is deleted from database.
Web interface is used to search the files in database, details how to search are described in the Help section of the search page.
You can also add a FTP server to a FTP server list, edit FTP server in the list or delete FTP server from the list through the web interface. So that anybody could not do anything with the server list only the record about abcdef is editable from host abcdef. There are also admins who can edit all records in the server list. The admins login through the web interface.
Enhancements:
- fixed few bugs in the crawler
- added a possibility to exclude some SMB shares
- www: improved Russian and Ukrainian translation