Re: Data organization

Carlson@WHSmith wrote: > OK, I'm at the end of my rope here, people. [...] I then attempted > to create a filing structure (Netscape messages) at my PC with a lot > of DejaNews research. [...] I'll admit that I now file the messages > locally, so I'm creating a new data repository with recent > information. Just would like to eliminate this large stack of > printouts of older messages from off my desk. > > How do you people keep all this information under control? [...] Well, AFAIK I've never persuaded anyone else use the system I use, so it probably isn't as useful as all that, but what works for me is: * Everything is treated as a mail message. * If I read something in a news reader that I want to save, I mail it to myself. (OK, so I don't trust a news reader to do what I want, and in any case, my mail/news archive is at the office and I tend to read c.d.i at home). * I have a directory ~jleffler/mail which contains about 60 sub-directories. There are big ones, like OnLine, and little ones like Informix3.30, and lots of intermediate ones. * When I want to save something, I choose the most appropriate top-level directory (eg OnLine) and then save it under some suitable name (eg lvm for logical volume managers, a vast file). So the information about 2 * 18.2 SCSI disks causing aggravation because of the 2 GB limit, let alone the 70 GB DiskSmart system which I read a few minutes ago were forwarded to my work account, and will be saved in the file mail/OnLine/2gb. They will probably also be forwarded on to relevant parties in the ongoing war to get the 2GB limit removed. 8.3x will have that limitation removed by all accounts, at least on machines where it can be removed (meaning, in all probability, all machines XPS is ported to). The changes are due to be merged into a forthcoming Foundation.2000 release, but I don't know which one (it isn't in the first one). * When I need to look something up, I have to guess which file I saved it in. This is the least satisfactory part of the system. When I checked for datetime and interval, I found about 6 files for each in different directories -- I need to rationalize that. Usually, I can work out which file it is fairly readily. Otherwise, I have to get inventive with grep (or Perl). * I use a mailer which automatically appends a new message to an old file without asking me silly questions. It also prepends the mail directory to all relative pathnames, and it has a filename completion option when saving. * I have around 30 MB of email archive handled this way. I can usually find an old answer to a question if I need to. -- Jonathan Leffler (jleffler@informix.com, jleffler@earthlink.net) Guardian of DBD::Informix v0.60 -- see http://www.perl.com/CPAN #include <disclaimer.h>