Medusa is software that allows you to quickly search your system for particular types of files, using an index. CAVEATS ------- Medusa is still in its early stages. It is not yet feature complete, and there are known bugs, some of which can take up lots of unnecessary resources on your machine. Please do not use this code on an afs system or over nfs volumes; the file permissions are not compatible with medusa. Running medusa on these systems is completely insecure. HOW TO BUILD MEDUSA -------------------- To make and install medusa, do the following from within the medusa source directory: ./configure make make install (do the last step as root if you are installing to a place where you do not have write permission) RUNNING MEDUSA ON YOUR SYSTEM ----------------------------- Medusa comes with a cron job that will be installed on your machine. By default a regular indexing process will not be run on your machine. You can turn the regular indexing process on by running "medusa-index-configure --on". You can turn regular indexing off by running "medusa-index-configure --off". Nautilus also allows you to turn indexing off and on. You can also create an index and set up the service by hand. More information about the indexer and how to set up medusa by hand appears below. You should have at least 64, and preferably at least 96 megs of RAM to run the indexer. Removable media will also be indexed by medusa, so remove CDs, zip drives, and floppies before indexing. Medusa's indexes will be created in proportion to the amount of data on your hard drive. In general the size of the indexes will be between 0.25-5% of the size of all of the data on your computer, so be sure you have enough free space to make the indexes before running medusa. After the indexer is finished, it will print a message to stdout, telling you that this is the case. To start the indexing process, as root run medusa-indexd Let it keep running, and wait for it to finish indexing. When the indexer is completed you should start the medusa search service. As root, run medusa-searchd After both the indexer is completed and the search service has begun, you can perform searches. This can be done by using the "search" feature in Nautilus, or by using the command line testing tool, called "msearch" For documentation about the command line tool, see the file docs/how_to_use_msearch. It is possible that this file is slightly out of date. CONFIGURING MEDUSA ON YOUR SYSTEM --------------------------------- Because indexing the contents of your machine each night can be expensive, Medusa by default does not run a nightly indexing process. You can turn on a nightly indexing process by running the following command: medusa-enable-indexing --on It can be turned off by running medusa-enable-indexing --off. You can also set a preference to "use fast search" in Nautilus, which will turn Medusa on. Any user on a system by default can turn the Medusa service on. A system administrator may also disable the service for everyone by using the medusa.conf file. This file is installed by default into the /etc directory. Setting blocked=yes will keep Medusa off and prevent any user from turning the service on. The medusa.conf man page also has information about the configuration file included with Medusa. REPORTING BUGS -------------- Medusa bugs can be reported using the eazel bug report tool, which is located at bugzilla.eazel.com. Report bugs for the "medusa" component. Bug reports that occur during the use of Nautilus should be reported using the Nautilus bug report tool if possible, but can also be reported using bugzilla. You can send email regarding medusa to medusa-list@lists.eazel.com, or to rebecka@eazel.com. MEDUSA FREQUENTLY ASKED QUESTIONS --------------------------------- 1. How is medusa different than slocate? Medusa is able to find files by their names quickly. However, medusa is also able to find files by their path, size, content, and owner, or in any combination. 2. Indexing takes a long time! When will this get better? Indexing your whole drive does take a long time, and we hope to improve this as much as possible. To some degree, however, this is a byproduct of the slowness of hard drives. However, we hope to introduce "incremental" indexing soon, so that files will only get reindexed when they are altered after the last indexing.