Sybase is currently in the middle of the most aggressive marketing program that I've seen for PowerBuilder in recent history. If you've been following my blog (and if you haven't you should) you'll know that Sybase recently ran a four-page ad in Visual Studio magazine promoting the release of PowerBuilder 12 (http://xrl.us/bhmxfe). They've also been conducting a worldwide series of "Developer Days" promoting PowerBuilder 12 and a number of other Sybase products. One of the things they show during that road show is a rather tentative "road map" of which features are planned for future versions of PowerBuilder (http://xrl.us/bhmxfr).

The release of PowerBuilder 12 triggered some complaints about some of the features that were removed with that release (primarily support for creating COM and COM+ components). Associated with that were complaints about features currently supported in "classic" PowerBuilder targets that are not supported in the .NET targets (e.g., data pipelines). There were also some complaints that the requests in the ISUG enhancement system (http://my.isug.com/enhancements) weren't being adequately considered.

In response to such complaints, I set up a number of polls on my blog (http://xrl.us/bhmxh9) to gain feedback on:

1. New PowerBuilder.NET features listed in the product road map
2. Existing Classic PowerBuilder features that should be included in PowerBuilder.NET
3. Top-rated ISUG enhancement requests
4. Additional features that could be dropped from Classic PowerBuilder

That last item is related to an article I wrote almost five years ago suggesting 10 features that Sybase should remove from PowerBuilder ().

If you haven't voted in these polls yet, I would encourage you to do so. I'll leave them open indefinitely. However, it might be useful to look at the preliminary results.

With regard to PowerBuilder.NET future features, the one that seems to be attracting the most interest is support for Silverlight. My initial reaction to the road map was that Sybase may be focusing on cloud computing support too soon (next release) and Silverlight support too late (two releases out). The initial responses from these polls would confirm that impression.

Support for .NET 4.0 was the second highest future feature, which Sybase has tentatively in for the next major release. I wonder if the folks voting for this are that familiar with what .NET 4.0 provides. Perhaps I'm missing it, but I'm having trouble finding the killer new feature in 4.0 (http://msdn.microsoft.com/en-us/library/ms171868.aspx).

Other future PowerBuilder.NET features that seem to be drawing the most interest include DataWindow support for Entity Framework, something I discussed back in April (), and support for HTML5. The latter I'm not particularly concerned about for the immediate future. HTML5 is largely still in its infancy, and there appears to be too much opportunity for it to fragment as previous HTML standards have done (http://xrl.us/bhmxkf). Its currently limited capacity appears to be the reason that companies like Hulu have decided not to move to it (http://xrl.us/bhmxj7).

With regard to features that are supported in Classic PowerBuilder and not (at least yet) in PowerBuilder.NET, it's surprising to me that the feature that gets the highest interest is support for email. Of course, the email support in Classic PowerBuilder is support for MAPI, something I wouldn't recommend when moving forward to .NET. However, SMTP email support is built into .NET and should be rather easily added to any PowerBuilder.NET application. Perhaps what people are looking for are some internal PowerBuilder classes that map to Classic PowerBuilder mail object calls when the application is migrated to .NET, but which implement SMTP under the covers.

The poll regarding which potential features to remove from Classic PowerBuilder hasn't yielded any significant information just yet, except perhaps that PBNI appears to be the feature that people would be the least willing to give up.

With regard to the ISUG enhancement requests, top honors goes to having the ability to save a DataWindow to Excel so that format is saved as well (not just data). Some of the other top vote getters are (I believe) addressed by new features in PowerBuilder 12 or future features already planned for post 12.0 releases. For example, support for automatic autoresizing should be addressed by the layout capabilities in WPF targets. Similarly, support for browser-independent web applications will be addressed by support for Silverlight in a post 12.0 release.

A couple of things strike me about the ISUG enhancement requests. One is the level of participation. I selected the enhancement requests with the highest average vote in the system to include in the poll. That Save to Excel request has been in the ISUG system since 1999 (11 years!) and yet so far has garnered just over 40 votes. I've gotten more votes for it in the poll I opened just a week ago. Of course, voting in my poll doesn't require signing up for an annual membership. Perhaps it's also because the voting process is a bit simpler. The bottom line is that I believe the results of the enhancement process would carry more weight if there were more participation. I know that the Novalys annual surveys get thousands for responses (http://brucearmstrong.ulitzer.com/node/1299679). Something in that range would seem more authoritative of the user community.

The other is that (with the exception of browser-independent web applications) none of the enhancement requests I included were what I would consider "release defining" enhancements. That is, they were all good candidates to include in a major release at some point as an "and in addition" type new feature, but they wouldn't be the feature that would be the major selling point for the new release. If people want to see the ISUG enhancement process more involved in the overall direction of the product, we need to see the kind of enhancement requests that would affect that overall direction.

In any event, that's just some of my conclusions based on some week-old polling. Overall, it does look like we're headed in the right direction; the only question raised was the speed at which we're traveling - primarily with regard for support for Silverlight. Polling isn't closed yet though, so feel free to participate if you haven't already.