Reader Feedback on One-Man System i Shops

Published: April 11, 2007

Hey, Bruce:

I read your recent article about using OpsNav to monitor jobs. What caught my eye was your lead line about being a one man shop. I didn't think there were too many of us left! Anyway I just wanted to bounce some ideas off someone.

The first being, how do I decide what kind of RPG modernization product I should be looking at. LANSA, Profound Logic, BCD, CGIDEV2--there are so many out there, what do I choose? (Well, there is also WebSphere, but I have not heard great things about it.) Or should I be looking at Zend Technology's PHP (which I just installed) or some other methods? I should point out I am very much a novice at RPG IV.

Second question, we are a small steel distributor and my boss is concerned that if anything happens to me, the company will be in trouble. So we are kind of looking at packaged software. I have a gut feeling we will spend a lot of time and money and energy getting packaged software to work the way we want it to. And we may lose flexibility to adapt to changing business conditions. I am trying to line up an IBM business partner to have a programmer/consultant available to back me up, but I just don't get a warm fuzzy feeling about how well someone will be able to pick up my pieces. I guess that also relates to my out of date documentation. As the saying goes, "When you are up to your neck in alligators, it is hard to remember to document draining the swamp." (It was something like that.)

On the plus side, I have a college intern that comes in about four hours a week. In the fall, he will take an RPG course at his college. (Wow, a college that has an RPG class: Youngstown State University). I am kind of grooming him to work for us, but I don't know how exciting we are to a young guy. Then again, a job is a job.

Thank for listening.

--Dale

Let me assure you that we are not the only one-man shops. There are plenty out there, along with lots of shops with five or fewer people--even some with zero!. As you indicated, one of the biggest challenges we face is having someone to bounce ideas off of. A newbie would flounder all by themselves.

I won't claim to have all the answers, but let's see if we can tackle your questions now.

1.    Modernization. IBM keeps telling us to modernize. Then IBM either leaves us adrift or gives us 80 options to choose from with no suggestions. I've been struggling with the same issue myself. Like your situation, I work for a small company. I have a saying: I buy all the free software I can afford! If there's a free or low-cost solution, I'll investigate that first. I've been able to get reasonably priced change management software and have investigated some of the modernization tools you mentioned.

Of course, CGIDEV2 falls into the free category. I've looked into it and have tried it out. The only hitch I've come up with is the fact that IBM locked these tools down about 18 months ago. They later unlocked them, but I still wonder if IBM knows what it has here. One vendor--ProData Computer Services had a promotion at the time where it gave away its tool (RSP) if you paid for a year's maintenance. That isn't quite free, but it was pretty close, so we bought it and I'm using it--sparingly for now, but I think it will do what we need. Whether you choose CGIDEV2, RSP, or any other toolset (they all look pretty good), do the research. Cost isn't the only factor, but it's an important one. Tim Prickett Morgan recently had a blurb on an open source toolset called Renaissance, which marries CGIDEV2 into a framework including Ajax enablement--free on top of free, it's hard to beat that price. I'll be looking into that in the next couple of months myself--for curiosity if nothing else.

If you have a local user group in your area it can be a great resource. We have a strong one in Milwaukee and there is another in Madison and I know people in both. Just being able to throw an idea at someone for a sanity check can help.

2.    Programming Resources. I had an intern here last summer and will most likely have another this summer. We have a technical college about 50 miles from here that has an IBM midrange program for programmer/analysts. (I'm also on the Advisory Committee). We need more of them.

Getting connected with a consultant sounds like a great idea. You're ahead of me on that. Like you, I try to document everything I can so that the world won't come to an end if I'm suddenly not here any more.

Canned software has its pros and cons. I strongly believe that if you buy canned software, you should run it as vanilla as you can. If there are special processes that are unique to your company, try to find a way to do it within the confines of the software. If you can't, keep those processes separate from the canned software, but find a way to link it back (as opposed to modifying the vendor's source). Only as a last resort would I suggest modifying vendor software.

Having said that, I remember making several changes to MAPICS DB code about 15 to 20 years ago. I wonder if any of those changes have survived the MAPICS updates (I don't work there any more). Usually vendors have exit points where you can pass data into or out of their databases to make life easier. I'd look there first for handling special circumstances.

What I said earlier about newbies floundering only applies if they are all alone. If they have you for a mentor, that is completely different. It is a great opportunity for them to learn. Programming isn't just heads-down RPG or COBOL coding. It is a juggling act. I currently have two major projects in the fire and can be pulled from them at any time for even higher priority requests. A newbie needs to see you doing that juggling and have some of that themselves to get a feel for what is required. We don't have tech support staff to lean on in a small shop. We need to do whatever needs to be done.

I don't know that I've actually answered any of your questions, but I do like to ramble. I think you are approaching it from the right perspective: Change is needed and you're smart enough to know that you don't have all the answers. Here are the suggestions I have:

• If you don't already belong to a LUG and there is one in your area, join.
• Make use of the forums. You never know who might be lurking with good information. You can create an account on the IT Jungle Forums by clicking this link.
• Return in investment is key. Cheap doesn't always get the job done, but it's as good a place to start as any.
• Having access to a consultant is like the company's insurance policy, and it is always a good idea.
• Grooming an intern is like the rider on that insurance policy.
• Canned software is a possibility if you can find a good match and don't modify it back to being a home-grown version.
• Keep focused on your goals. Tools are only valuable if they help you achieve those goals.

If you have specific questions about tools or techniques, I'll be happy to offer my opinions and experiences. I can't guarantee that it will solve your issues, but I'm happy to offer my two cents worth. It's good to hear from someone in a similar situation--misery loves company, right?. Perhaps we can work together and help each other.

--Bruce

                     Post this story to del.icio.us
               Post this story to Digg
    Post this story to Slashdot