Archives are single files that contain entire COGENT research programmes. In COGENT's normal mode of operation, the models within a research programme are each contained in a separate directory or folder within the user's project directory. This makes transfering models and/or complete research programmes between machines difficult, because you need to transfer the contents of an entire directory. It also makes uploading and downloading research programmes from the web more difficult than necessary. By stuffing all of the relevant files into a single file, archives provide a convenient portable format for transfer and exchange of complete research programmes.
The archive browser is an interface tool that allows research programmes to be transformed into archive files and vice versa. The browser window (shown below) consists of a list of research programmes (on the left) and a list of archives (on the right), with two buttons between the list for transforming one format into the other. To archive a research programme, select the desired research programme and click on the >>Archive>> button. To extract an archive, select the desired archive and click on the <<Extract<< button. Preferences relating to archives (e.g., the format in which archives will be created, the locations of archives and research programmes, and whether research programmes/archives should be retained after they are transformed) may also be set from the archive browser. (See Archive Preferences.)
COGENT supports several archive formats:
Each archive format has pros and cons. The CAR format is most portable, and generally recommended. CAR and COG files may be renamed, and the archive will then extract into a research programme of the new name. TAR and TGZ files contain the names of the research programme within them, and so renaming these files will not alter the name of the extracted research programme. Furthermore, TAR and TGZ formats are not supported under Windows. However, the TGZ format is most compact.
If archive file size is an issue, be sure to reinitialise all models in the research programme before archiving it (this generally ensures that message logs and current contents files are small). For reductions in archive size can be obtained by compressing CAR files (using a zip utility), though they must of course be uncompressed before COGENT can extract the contents.
The format of an archive is indicated by the last three characters of the file's name. COGENT uses these characters to work out which extraction routine should be applied to the file. Don't change them!