Oh my, the world is abuzz with stories stemming from the recent announcements. A lot of negative press has wafted through the mailing lists and online magazines, and to be honest, I was pretty incensed myself. But you don't need to hear more ranting and raving; you need a bit of real information. So first I'm going to put on my journalist hat and try to present as unbiased a view of the situation as I can.
First Things First
Let's get some of the larger pieces of information out of the way. In the past, I've complained that the Rational group and the WebSphere Development Studio Client for System i (WDSC) team don't talk enough. Well, be careful what you ask for, because IBM has definitely fixed that: The WDSC product is now entirely under the Rational tools umbrella. Analysis of this particular event is difficult: Does it mean that WDSC will get diluted into a Rational tool with System i add-ons as an afterthought, or will the Rational products start to get a better understanding of the System i world? That's going to depend on the management of the group.
Which brings me to the second point. Bob Cancilla has been appointed to head the Rational tools for System i development. While I'm not sure exactly what this entails, it does mean that there is a System i person in place in the Rational management infrastructure, and that can only be good for the System i. I'd like to get some real, official statements as to what all this means, but for now I can give you my sense of the situation from conversations with Bob.
The most important issue is that Enterprise Generation Language (EGL) is really high on IBM's internal radar as the development language of the future; I hear that not just from Bob Cancilla, but from everybody I talk to at IBM. At the same time, though, Bob assures me that RPG will not get short shrift and that, if anything, we will see even more development in the mainstay of the System i product line.
And Now for Something Completely Different (Not)
IBM is once again in the process of revamping its financial strategies, and of course the System i is in the middle of the bean-counting crosshairs. I hear tell that the System i revenues are down, so they have to "shake up the pricing." What this means is a bunch of terms that I'm uncomfortable with, such as "unbundled" and "deprecated" and "entitled." Let's get these straight:
- Unbundled—You used to get this for free. Now you don't.
- Deprecated—You used to have this. Soon you won't.
- Entitled—You will be able to get this. We're not sure how much it will cost you.
CODE400 is deprecated. The only GUI screen tool (CODE Designer) we currently have will no longer be maintained. New keywords will not even be added. I'm not sure it will even be shipped with the next version of WDSC.
EGL is unbundled from WDSC. If you use EGL today, you may be grandfathered in; the word IBM uses is "entitled," but there is no official word whether that means free or not. New customers and people not currently using EGL will have to pay for it. How IBM can determine whether you are currently using the tool is beyond me, but that's the official position, and they're sticking to it.
The part that is rankling everyone, though, is the fact that the supposed replacement for CODE Designer—an Eclipse-based screen design tool—is only currently available in the advanced edition of WDSC, a superset of the free WDSC tool that costs about $3,500 a seat and thus is out of the price range of just about everyone. And while it hasn't yet been cast in stone that this is the final delivery strategy for the new tool, it leads to the larger issue.
What IBM Just Doesn't Seem to Get
OK, journalist hat off, System i developer hat on.
I have heard this pricing policy rationalized thus: IBM can't possibly give System i developers software for less than they charge non-System i developers. This is the underlying premise that I take issue with. It is completely backward! By definition, you should be charging people who buy System x and System p boxes more than you charge your loyal System i clientele. System i users spend a premium for their hardware and OS licenses expecting to get an integrated package.
But this truism falls on deaf ears; the people trying to get more money from the System i developers just don't seem to understand the difference between a Corolla and a Corvette. Someone buying a Windows box (or a System p/x machine) is your Corolla buyer. They are looking for economy above all else. They pay considerably less for their base model vehicle, and they expect to pay extra for features like a leather interior or a five-disc CD changer. They'll sit and haggle and agonize over each option and in general make the salesperson's life miserable (well, at least I know that's what I did).
A Corvette driver doesn't want to be bothered with all that. The idea is that he is buying a machine with all the necessary pieces in place: an integrated vehicle, if you will. When you purchase a Corvette, you pay a premium for the car, but at the same time, you don't expect to get billed extra for the floor mats. And making System i developers pay extra for development tools is like charging for the gas pedal. It's hard justifying the cost of the box in the first place; imagine having to go back and tell upper management you need another $3,500 a seat...to create green-screens.
If you want to get some money out of the System i community, tack a few bucks onto 5722WDS, maybe $500 or so. Provide a limited number of seats of the tooling for free (number based on processing tier—for example, maybe two seats for a P05). Trust me, jacking the base price of the Corvette a few percentage points is going to be a lot easier to sell than having to go back and ask the boss for more money to get windshield wipers. Add all of those moneys to the WDSC development pool and you're talking about millions of dollars. Just how many $3,500 seats of WDSC are you expecting to sell? I can tell you that the sum total of AE license fees among my clients is going to be very close to zero.
And more importantly, by including the tool with the base price of the system, you guarantee two things: Shops will try to use it, and developers won't shy away from it. Remember, if ISVs can't be sure a shop will have a given tool, they won't try to build a product that requires that tool. And if the tool is that good, then the larger shops will indeed pay for it instead of other products. Make sure it's easy for them to get additional licenses for a reasonable price (no offense to the Rational team, but no tools are worth $3,000 a seat).
In short, quit trying to nickel-and-dime the folks that already pay you a hefty premium. Sell the System i as the integrated machine that it is, give developers access to all the tools as part of the overhead, and if your product really is worth the money, people will buy more of it.