Success Story

I've looked at a couple of tools in the past that are designed to take IBM i spool files and transform the data into another desired format. In some cases, it's a PDF file. In other cases, it's a spreadsheet. Or sometimes users want to e-mail a report to a list of folks on a one-time or ongoing basis. The list of vendors that are merging all of these functions and more into either one package or a suite of software continues to grow.

While wandering through the COMMON expo last spring, I ran into a group that touted the ability to go from a spool file to an Excel spreadsheet or graph fairly seamlessly-an interesting concept and one I could see some use for.

Enter Catapult from Business Computer Design International (BCD).

Installation

Actually fairly painless and easy. At least it should be for you with the normal installation. I have to extend thanks to Cameron Ellis for helping me wade through some communications issues on my end. Frankly, there's so much "stuff" on my server that some techs are surprised that it even boots.

Ease of Use

I sat through the demos at COMMON, but there's nothing like doing the 10-minute tech-support walkthrough after you've done the install. After the installation is complete, you should be able to have your feet wet within 10 to 15 minutes, to the point that you're going through the classes included with the product and becoming dangerous. I recommend you have tech support walk you through the demo, then take the classes in the documentation.

There are a few minor methods of doing things that are unique to BCD that you'll need to learn, but this isn't necessarily complex.

It helps if the spool files are in tabular format. What I mean by that is it's best to format your reports so that each column contains only one field-i.e., column one is the name, column two the street address, column three the city, et cetera. You can mix and match with multiple items coming from multiple lines, but it's easier if reports are in tabular format, because the product makes extensive use of pattern recognitional algorithms. Also, blocked tables are easier to recognize and divvy up than busy reports.

I can potentially see where your report might become so busy that it may have some problems doing a proper mix and match on the extractability side. But if you keep things simple, you're home free.

As you might imagine, you have to provide the software with rules for spool file selection and routing. And setting up rules on the Catapult side was fairly painless. (See the Functionality section for more on this.)

Documentation

Now, this is good stuff. Frankly, I referred a few vendors that are in the process of redesigning their documentation to BCD to see how the company does its documentation. It's easy to see what you're supposed to be working with; the instructions are easy to follow; the pages are well laid-out, etc. Basically, this is the way it should be.

Certifiability

I didn't kick the tires on the e-mail functions because they require a mail server, and I'm still using an ISP for that functionality. But everything I did test passed with flying colors.

Functionality/Usefulness

There are several functions to keep in mind here. Catapult monitors output queues on the server based on whatever rules you've established for searching, extracting, processing and routing.

From the operations standpoint, it's a decent way to do report reformatting and routing over a network-local, Internet or e-mail. If you're expecting things to work sooner than they do, keep in mind that the polling is based on a timer, so if you want it to happen faster or slower, tweak the timer.

I also liked having some form of centralized operations with an electronic report distribution over a network directly to where a user might be expecting to find his or her stuff. This allows your operators to do more than worry about report distribution. The monitoring systems watch for whats coming down the spool pipes and then handle the distribution you predetermined. It's kind of a "Look, Ma, no hands!" way of running things.

Warning: Make sure you set up your report-naming nomenclature so that each file name on the PC side has a unique identifier. The default is that a new file would potentially have the same name as the one its replacing, ergo replacing the existing report. The product has enough capabilities built in to easily manage this.

OK, I'm a bit mixed here. For the ad hoc requests for extractability of data from a report or for the ongoing requests, this package offers some really nice functionality. But for data flow needs, such as permanent data transfers, I would recommend writing the RPG code to build a work file (comma separated variable [CSV] format or whatever) and either use a cpytopcd command or an ODBC driver to directly import the data into your PC application.

Support

These guys know their product well. They're also well-briefed on most network communications problems that people like me are apt to find on their boxes. They will hang in there with you to make the product install.

What I'd Like to See in the Next Release

CSV files as an export option-CSVs are universally used in the PC world for moving data from one application to another. (Update note: This feature was added in a subsequent release)

Summary

As we become more net-savvy and -dependent, this kind of utility will help make the transition of data easier for many of the understaffed and over-worked IS shops.

If your network needs to get spool files to the end user(s) in a timely manner and you don't want your operators involved with mundane and potentially error-prone spool distribution, then you should seriously consider Catapult.

Report Card: Honor Roll