###   Projekte und Informationen rund um den KC85   ### 

 

CFOG's PIP, May 1986, Volume 4 No. 7, Whole No. 43, page 12

A Report on PdbM -- A Data Base Program

by Benjamin H. Cohen

I saw an ad in the October 1985 Foghorn offering a new data base program, the Personalized Data Base Manager or PdbM. The author of the program is the same person who wrote the Desolation game for the Vixen and who uses the company name B.C. Software. The ad promises that PdbM is "simple to understand", offers full screen editing, and is 100% compatible with other formats from dBase II to SuperCalc. With a record limit in excess of 65,000 and over 200 fields per record, PdbM sounded like it might be what I needed. One nice feature promised that you could add a field to a data base with a simple command. In the April 1986 Foghorn a very short report appeared, praising PdbM to the skies. I decided to take B.C. Software up on its 30 day money back guarantee.

PdbM comes on a single double desity disk with a small (5.75 by 7.375 inches), thin (0.125 inch), neatly perfect bound manual nicely printed with a daisy wheel printer. The manual sure looks nice. A little editing an proof-reading would have halped readers, though, even more than the nice appearance. After I figured out that a "filed" was a "field" -- a typographic error repeated throughout the manual -- there wasn't too much difficulty understanding the manual.

It was irritating, however, that the author follows no known rule of capitalization, that typos abound, that there are references to definitions of terms in Section 7-1 that aren't found there, to an ASCII table in Section 14-1 (for the sort order) even though the manual ends on page 13-1 (that it's not just a missing page in my copy is attested to by the table of contents -- it's not listed there, either.) On page 5-3 the author tells how to 'string' the Find Command with other commands: "Type (F) for Find, (D) for Output and (C) for Continue..." He means, I'm sure, "(D) for Delete", but it sure stopped me there for a minute.

When I got the pacakage I was at my office where I have an Osborne Executive on my credenza. I copied the disk for backup, put the master away, stuck the copy in Drive A:, and entered PdbM<cr>. Unfortunately, the Executive froze. Not to worry, the application of BIOS22.COM made a version that would run ond the Executive. (I don't believe the program was tried on an Executive. As additional evidence I note that the manual says that the DELETE key is supported for the Executive -- now you support a key that doesn't exist is beyond me. Perhaps that reference should be to the Vixen.)

At home with my Osborne 1 (tan case, Osborne 80 column and double density upgrades), with the small (192K) Drive C: RAM disk, I loaded PdbM again. Since I use the Dvorak keyboard I utilize SmartKey or XtraKey to reconfigure the keyboard. This is hard on program that are memory hogs, since the Drive C: uses about 1110 bytes for its software and the keyboard reconfiguration programs and their definition files subtract another 2K+ from the TPA (the Transient Program Area, the portion of the nominal 64K RAM that's left to run programs after subtracting the CP/M system overhead and in my case the Drive C: software and SmartKey or XtraKey.) unfortunately, while PdbM will run with the Drive C: software, it won't run in the Drive C: with either SmartKey or XtraKey present.

I've used PC File and DIMS and other programs of that sort. If you run these programs with your data on a floppy disk it takes forever to sort the data. Well, maybe not literally, but a sort that takes about 15 minutes when PC File is run on a floppy system taktes about 2 minutes in my Drive C:. I'm not about sacrifice that speed, so the sheer size of PdbM forces me to return it to B.C. Software. Another factor against PdbM is the need for macros while using any file management program. When I have to update a list to show who has paid current dues, a macro lets me enter the name [or enough of it to distinguish it from other names] and then hit one key to update the entry and initiate the search for the next name. In my RAM disk all of this is very fast.

The size of PdbM isn't all bad. With the program all in one piece you get the ability to move from function to function without delays. Programs that don't use as much TPA often have to have some of their functions in separate "overlays" -- that means each time you use those functions you have to wait for them to load from the disk. WordStar users are quite familiar with this phenomenon. PdbM on the other hand, has no overlays and thus moves from funtion to function quickly.

I tried out PdbM only in a small way, since I knew I wasn't keeping it. I can't really evaluate it fully since I didn't spend a lot of time with it.

Some of the problems with the manual show up in the program, too. One of the options in the main menu allows you to "Preform" (sic) arithmetic functions. Other spelling and grammar mistakes abound in the abundant and otherwise generally quite useful help screens.

You can't give fields long names, with spaces in them -- up to 65 characters.

[What neither the manual nor the help messages nor the plug on the back of the box nor the advertisements tell you is that field name plus field length is limited to 69 characters. If you try to enter a field name plus length that are, combined, greater than that, you'll get a message that the limit is too big -- but it doesn't tell you what the limit is. It's not listed anywhere that I could find it: I determined the number 69 by trial and error.] When you define your data base the field name is followed by a colon and then either a comma and the number of characters allowed or by a series of fixed characters terminated with a carat (¢). If you do the latter, the fixed characters will appear in each record in non-editable form. For example, you could create a field name

Telephone Number:(   )   -    ¢.

When you got to enter a telephone number the paranthesis and hyphen separators will always be present and PdbM will skip over them as you enter the data.

On line help is a nice feature, too. Error handling seems excellent -- PdbM won't let you enter incorrect commands and generally tells you what your choices are at a particular pint in the program. but there weren't always help screens available, and with the skimpy unindexed documentation I was several times left high and dry.

You can add or delete a field in an existing data base by a direct command. Editing a field name, however, seems impossible.

When you as PdbM to find a record you can enter as many characteristics as you want in a 'mask' -- if you want all records that have Zip code starting with 606, last name starting with G, and telephone number starting with 686, you enter those characteristics as the 'mask' in a dummy record and PdbM will find the matching records. Or you can find records that don't match your posted characteristics. I couldn't figure out, however, to find records with a Zip code greater than 606 -- a standard type of search for data base programs.

PdbM offers arithmetic functions. If your doing something that needs to be numbered or incremented, PdbM will do it.

PdbM offers a feature called overlay. This lets you overlay the data from the previous record in the same field in the current record. It's a great help. Many mailing lists have a lot of people from the same city. Most programs let you enter an apostrophe and hit the <cr> or use some other convention to repeat the required entry in each record without retyping it. The lack of an index or help screen prevented me from figuring out how to get this to work in PdbM. Indeed, in a number of places, hitting the "?" for help resulted only in a grind of a disk drive and a rewrite of the screen, without any help.

PdbM commands are performed only once unless you tell it to continue on the command line. If you tell it you want to insert more records in the file, you have also to tell it you want to continue or it will automatically stop after one addition. Somehow, it seems to me that the default ought to be in the opposite direction, since a single keystroke after the insertion of a record into the file will terminate record insertion.

As I said, this isn't designed to be a complete review or evaluation of PdbM. PdbM may work fine for you if you don't have the TPA problem that I run into. In any event, there's the 30 day money-back guarantee.