A Clarification on RAD Studio Versions Numbering

Since version 11, Embarcadero has modified the version numbering scheme of its releases: 11.2 is an update of 11 Alexandria, not a new major version.

As announced when Embarcadero released RAD Studio, Delphi and C++Builder 11 Alexandria (see this blog post), the product now has a different version numbering scheme compared to the 10.x series. Let me clarify.

First, RAD Studio has two different types of release, major releases and minor ones. The core difference is not based on marketing, the amount of new features or anything like that. The main difference is in the binary compatibility with DCU files (compiled Delphi units) and BPL files (compiled component packages). Minor releases maintain binary compatibility, so the same third party component or code you have build should keep working as is, without recompiling. Major releases have compiler Delphi or RTL changes that break the unit compatibility, requiring recompiling all binaries.

With this difference in mind, over the past few years, we released major versions with a double number like 10.1, 10.2, 10.3 and 10.4. They were followed by minor releases like 10.2.1 for 10.2 or 10.4.2 for 10.4. Each major version generally has 2 or 3 minor releases following it, maintaining compatibility. There were reasons for this choice, including both Microsoft and Apple staying on macOS 10 and Windows 10 for many years.

As announced in the 11 Alexandria release time frame, we have now changed the numbering scheme, to use only the first number as a major release number. So 11 was a new major release followed by 11.1 and now 11.2, both minor releases. In fact, the full name of the last release is " RAD Studio 11 Alexandria Release 2 ". When we'll have a new major release we expect it to be called 12 or 12.0. I hope this clarifies the fact that RAD Studio 11.2 Alexandria is a minor release, while 10.2 Tokyo was indeed a major one. We believe that going forward the new numbering scheme will be easier to understand.

Hi! Embarcadero quite often refers to However, "this is what our customer want". I seldom feel I'm included in these statements: To all of you "Embarcadero customers" out there, do you really not want new functionallity, language improvements, more competitive frameworks (helper classes for TDateTime are neat, but not imp, receive), Linux ARM64 support, etc., etc. Delphi is lagging behind big time now. And no improvements are to be seen in the sky (Marco talked about C++ being a priority area for the upcoming time at the release webinar, does this mean we'll see 11.,3, 1.1.4, 11.5 etc, before any 12.0?). The campaigns targeted Delphi users not renewing their licenses is great. But should you not also focus on us loyal actually paying customers? It seems you take us for granted, which (sorry to be blunt), seems to be a part of your business plan.

Year based numbering worked best in my opinion. I can easily remember which features appeared in Delphi 2005 vs Delphi 2007 vs Delphi 2010 etc. When it comes to the XE series, they're all a blur. Cities names are pretty good but remembering the chronological order is starting to get a bit difficult.

Regarding "what customers want", here is my two cents: I don't really crave new language features or frameworks. After all, I spend 100% of my time writing native Win32 desktop apps using the Win32 API and the VCL. And using these very old frameworks, you can create extremely-high-quality applications with GUIs essentially free from bugs or quirks, and that can be used either with a mouse or a keyboard. Superusers can work really efficiently in these. Firemonkey, on the other hand, ... Well, let's not go there. Similarly, it didn't take me long to find several bugs in the new "modernized" Microsoft Paint app. It's much buggier than any of my apps, which is surprising since Microsoft is a fairly large company with a lot of resources. No, I -- and many other Delphi developers that have been doing this for 20+ years -- don't want new frameworks and are quite happy with the language as it is today (but I do love some recent additions like operator overloading and inline variable declarations). Instead, my main issue is the quality of the IDE and Code Insight. And fortunately, I do feel that Embarcadero is prioritizing this. So I'm quite happy with the direction, but I do find it a bit slow. There are still some rather nasty IDE and Code Insight issues present.

Thanks Andreas for your response. I did have a short look at your homepages and the products you develop (cool stuff!) and I think I understand your position (a single-team hobby developer, as it seems): it makes sense. Just to clarify my comment a bit more, it originates in from a company where we have many software engineers working, some of them using Delphi. For us, the business perspective on the software we develop is of the highest importance and includes many things. One of them is selecting the right tool for the right task. And not just for the current being but also looking into the future. E.g., we don’t use Delphi because we like it; we use it because it makes business sense in a longer perspective. For many companies/developers, Delphi makes sense just because of legacy reasons. For us, this is not sufficient – we must provide cutting edge products also in five-ten years. And other tools (programming languages, IDE:s, frameworks, etc.) are evolving. They not only try to be on par with the latest technologies but make it more cost- efficient for teams to maintain software over time (including stuff like code readability, exception handling, efficient writing code, adaptation to new technologies, team cooperation features etc. etc.). IDE improvements etc. are very welcome but no way near sufficient for us to continue to make investments in Delphi-developed applications when it comes to a business perspective. When it comes to what has actually been delivered, Embarcadero seems not to be interested in targeting what I describe above. I wish they could be more honest with their strategy, but I suppose that honesty would drop sales, so instead, they are quiet. Earlier roadmaps have been more focused on providing false hopes than deliveries, and when this behavior was disclosed by the community, the roadmaps were made blurry (is there even one today??). It seems “obvious” they try to hide the decreased investments. The GM, a few weeks ago, posted a message to the community, seeking for dropped subscribers to renew their subscriptions. I personally would have liked to see some efforts from the GM to reach out to existing subscribers, actually paying for subscriptions, and telling them what we can except from Embarcadero going forward. Not much, it now seems…

Just to clarify: Yes, Algosim is my personal project, and I do use Delphi every day at home as a "single-team hobby developer" in that context. But I am also a software developer by profession working for an IT company, and we use Delphi to create enterprise-level Win32 software for our customers (who are companies).

Marco Tech Blog