Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!cs.utexas.edu!sun-barr!decwrl!ucbvax!pro-exchange.cts.com!rich From: rich@pro-exchange.cts.com (Rich Sims) Newsgroups: comp.sys.apple Subject: Re: Market Research Message-ID: <16039.apple.info-apple@pro-exchange> Date: 16 Dec 89 06:23:55 GMT Sender: daemon@ucbvax.BERKELEY.EDU Organization: The Internet Lines: 81 In-Reply-To: message from done@pro-sol.cts.com (Don Elton) I have no feel for what "most" users would think was more important, but for myself, I'd go for the right database. "Right" meaning that it has some features I would find useful. Some of the things I'd like are as follows: (Gee, I never got to "design" a data-base before!) 1. Ability to merge data from other sources in a variety of formats, such as fields delimited by tabs, returns, commas, etc. 2. A moderate "macro" ability during data entry, with the capability of either using a 'standard' value or duplicating the previous entry, as desired. 3. Some way of linking long text entries to specific fields, either by direct inclusion in the data file or as a separate file that's read when that field is displayed -- sort of "semi-relational" I guess. By "long" text entries, I mean I'd like to be able to include a text document that's a few k long. 4. I wouldn't find it particularly useful, but since this is a GS application, I suppose some way of storing graphics would be mandatory for most GS users. 5. Ability to define specific fields to hold particular data types and disallow invalid entries. This should include an optional maximum entry length. 6. A field (or two) that are *automatically* updated with the current date and time whenever any data in that particular record is changed or the record is created (or data added). This could be an optional field, not one that's always created, although either way would be acceptable. 7. Mouse usage should be optional, not required! 8. Ability to create data entry "templates" that are accessible quickly and easily by someone who's not necessarily familiar with all the "bells and whistles" that the database provides. This is to allow creation of the file format and data fields, and then to allow anyone to easily enter the data from a prompted or menu-type environment. 9. It should recognize text and numeric data, and provide for treating either type as the other. 10. Once the data is in the records, I'd like to be able to manipulate it almost without limitation, with the options of modifying the data and either replacing the existing data with the result or creating a new field to hold the result. This kind of thing should be programmable as a script or series of macros that can be called up at any point, so the sequence can be defined once and used over and over. Right along with the above, search/manipulation capability should include "pattern" or "expression" matching, rather than just being able to match literal data chunks. 11. How about two variants? One to be RAM based for small data files and one to be disk based for larger amounts of data. If only one were available, I'd choose disk-based. After all, we can already keep small stuff in AppleWorks. 12. At the risk of drawing a lot of heat, how about designing the thing for performance, not designing it to be useable on a one-floppy system. (If it still is useable that way, fine -- if not, well, that's life!) 13. Report generation capabilities - along with all the normal stuff, it should be capable of tabulating sub-groups of data as well as the entire group, and preferably without any physical re-organization of the data file. Let it do something simlar to this.... Sort the output by last name, then first name, group it by ZIP code, and give me a count of the number of entries using each different phone number prefix. (Not that I'd use that particular one, but similar stuff would be very useful to me.) Oh yeah -- multiple data screens available, so they can be organized into logical groups without cluttering up the display by trying to see everything at once. I can think of a dozen other features I'd like, but this is already too long, so I'll give it up for now.