Parchive

Parchive is a correct system of errors which can be applied to a whole of files to allow their rebuilding when one or more these files are missing.

History

The groups Usenet were originally conceived for abstract conversations and were thus not conceived to be a reliable transmission resource. One of the limitations for the conversations was that the messages generally concise and were limited to the text in ASCII 7-bit.

To pass from the binary data of 8 bits on a channel of 7 bits and to use Usenet to transfer them, various techniques were conceived like uuencoding and Base64. Later the Usenet software allowed the use of the wide ASCII 8 bit, thus of new methods as yEnc appeared.

Whereas the problem of transmission of the data was solved, the reliability of Usenet left something to be desired. In 2001, Tobis Rieper and Stefan Wehlus Parity Volume Set specification 1.0 proposed. By transmitting additional data and by using the correction of error Reed-Solomon, a user can rebuild the missing data starting from an incomplete remote loading.

Overall picture

The files By and Par2 are used to repair the files damaged while being based on their binary model. The name By comes from By ité. These files can be used to rebuild damaged files or to recover missing files in a whole of files. For example, if you downloaded a file cut out in 47 files but that only one of them is missing, you can simply download a file BY for this file. Once you have this file BY, you can use a program BY rebuilding the missing file.

Versions

There are two incompatible versions (1 and 2) of the specifications of the format.

For version 1, by considering the files f1 , f2 ,…, fn , Parchive will consist of a file of index ( f.par ) and of a certain number of volume of parity ( f.p01 , f.p02 , etc). If there are all the original files except only one (for example, f2 ), it is possible to create the file missing ( f2 ) starting from all the original files plus any of volumes of parity. Moreover, it is possible to recreate two files missing starting from two volumes of parities, and so on.

The files index (*.par in version 1 and *.par2 in version 2) are not necessary to recover data. The indices are only hashages which make it possible to identify the target files quickly, and their contents are duplicated in each volume of parities. The files index are used to check quickly if there is an error in order to know so of the additional files of parities are necessary. They were more useful in the verison 1 where volumes of parities were larger than the indices. All the data files By contain the index in entirety, but the easiest checking of the complétude is to use a simple file index.

Most important of the limitations found in the real use of Parchives was when an error of only one bit in a file forced the algorithm to put at the variation the file in entirety. To improve this situation, a second version of Parchive was created, which cut out all the source files in smaller blocks. These small blocks can be seen like file Parchives themselves; so sufficient blocks are present, all the others bloces can be recreated.

Files PAR2 use the convention of naming in general: fichier.vol000+01.PAR2 , fichier.vol001+02.PAR2 , fichier.vol003+04.PAR2 , fichier.vol007+06.PAR2 , etc the elements +01 , +02 , etc in the file names indicate the number of blocks which they contain. If a file index of one téléchargment indicates that 4 blocks are missing, the way simplest to repair the files will be to download the file fichier.vol003+04.PAR2 . However, thanks to the redundancy, the file fichier.vol007+06.PAR2 is also acceptable.

Other uses

The Parchives files can be used within other frameworks that the Usenet transmission.

When one uses CD-R cheap, it is possible to insert data of redundancy, while engraving of Parchives with the data.

Certains software using Parchive can cut out a source file in several smaller files, to abstract itself from the limitations imposed by systems like FAT-32. Even with 0% of redundancy, the software can rebuild the file moose starting from the small portions.

Software

Microsoft Windows: QuickPar, Parbuddy
Mac OS X: MacPAR deLuxe
GNU/Linux : Parchive, PyPar2

Other uses of the extension of file

A file BY can also be a file spreadable Portlet SAS for the Gate of supply of information SAS. The nomenclature follows the rules of the files GRAVEL BANK, WAR and EAR. These file are simply files ZIP.

To see

RAID

External bonds

Parchive project - complete specifications with the mathematical approach

Random links: