Home > Uncategorized > An Open Letter To Embarcadero

An Open Letter To Embarcadero

Hello all!

After a long silence, I finally had some free time to create another post. No matter what I do, Delphi is my language of choice, and I end up running it for new projects. I felt extremely bad about the future of Delphi, and decided to write a letter to Embarcadero. If you’re interested, their public contact mail is info@embarcadero.com.

Here’s the letter I sent to Embarcadero:

“Dear Embarcadero,

First of all, I want to congratulate you on your purchase with CodeGear. The join is a good fit for both companies. I would like to tell you about the fatal mistakes of previous management, hoping that you’ll take them into account.

First of all, let me introduce myself. I’m a passionate Pascal/Delphi developer, and using this language since Turbo Pascal 3, where Borland introduced Turtle Graphics (which had later named as BGI). I have used every Pascal related IDE which Borland released, until Delphi 7. Although I have tried to use latter versions, the problems and slowness prevented their usage. With over 15 years of loyalty to Borland, I suppose I have the right to make some recommendations about Delphi development platform, hoping it will get noticed in its new home, unlike CodeGear.

CodeGear, as from community standpoint, has never been a successful company. No matter how their sales good, the company failed to achieve its users’ expectations. The IDE line they had was best IDE tools ever created, until 2001, Delphi 7. The usability and quality of the products significantly lowered with the release of Delphi for .NET (Delphi 8).

No IDE tool achieved same quality as previous versions, after Delphi 7. Almost all of Delphi developers are pissed about Delphi now, and using Visual Studio to develop solutions. There are a lot of reasons for this, and I’m listing some here.

– Lack of dedication, innovation and quality
– Avoiding community requests and recommendations
– Trying to look like Visual Studio
– Building its IDE technology on .NET, which is a competetor’s technology.
– Depending IDE on slow .NET framework, instead of native code.
– Failing to keep up-to-date with major open source projects (Like MySQL, PHP, PostgreSQL etc)

The future of Delphi language is not clear now. As you have bought all IDE line, you should had a look to the previous versions of the IDE tools. A product lives with its pros and cons, and there must be lots of differences from other tools in the market, for it to survive. Delphi versions from 1 to 7 are extremely powerful, different, has their own approach to problems, and flexible. Please analyse the new IDE’s, andย compare them to old ones. You’ll see the vision difference. You’ll see why people like me became passionate about Delphi. My reason for sending this letter is that, I want to see Delphi back to its shiny days. I want to see my precious quality environment again.

As I see, you bought the tools, and started to sell them as is on your web site. This is a mistake, because the tools are not usable in their current conditions. If you want these tools to thrive, you should work on them constantly, before making them public again. Also, you should *really* listen comments and recommendations from community. CodeGear got 40 pages of feedback, when they asked people the reasons for not upgrading. Guess what? They took the comments down, and the next release was same on every area with previous ones.

This was a long letter. But when Delphi IDE dies, I don’t want to say that “I could do something for my precious IDE, but I didn’t”.

Hope your vision and dedication will be far more higher than CodeGear.

Best Regards,”

So far, I didn’t even receive a response from Embarcadero. I don’t even hope that they will respond back.

Anyway, I hope they will listen us, unlike CodeGear.

Categories: Uncategorized
  1. July 21, 2008 at 6:16 pm

    Thank you for the blog post. We are definitely listening and working on new versions of Delphi. It is true, that the Delphi IDE is a mix of native code and .NET managed code (mostly for the UML modeling parts and for .NET development itself).

    The community has many ways to give us feedback, including blog posts like this, putting requestions and problem reports in Quality Central (http://qc.codegear.com/), taking part in Delphi surveys, participating in newsgroup conversations, and tuning in on webinars and community chats. And you can also contact us directly at our offices and by email (davidi@codegear.com, and Nick Hodges Delphi product manager nick.hodges@codegear.com).

    You are right that we have more work to do, especially now that we are part of Embarcadero, a company that cares about developer and database tools. You definitely bring up valid past/historic points that we need to improve upon.

    I am a little confused about one point that you mention, “Failing to keep up-to-date with major open source projects (Like MySQL, PHP, PostgreSQL etc)”. We do support MySQL 4 and 5 in our drivers. We do have Delphi for PHP, innovating by adding the same Delphi component model and visual/RAD design for PHP applications.

    Thanks for the comments and encouragement. We are working on Project Tiburon, the next editions of Delphi and C++Builder and future innovations for 64-bit, multi-core, and additional platforms. Keep pushing us and keep giving us your thoughts.


    David Intersimone “David I”
    CodeGear Chief Evangelist
    Embarcadero Technologies

  2. Excessive
    July 21, 2008 at 6:56 pm

    Dear David,

    First of all, thank you very much for taking time to comment here. I can’t describe how happy I am after reading your post.

    I’m extremely excited by Tiburon project, and following it closely. I really liked new features, especially anonymous methods, VCL enhancements and Unicode support.

    But these are not enough. I love Delphi by heart, and if one can fall in love with a programming language, it is Delphi for me. You and the rest of the Delphi team are the most valuable people in the world. You were the ones who created it, managed it, developed it in the past. You can do it again.

    No matter what happens, I’ll always use Delphi, as no matter what I do, I’m returning to it back. It’s home.

    As for open source projects like MySQL and PostgreSQL, I didn’t mean now, I wanted to talk about past. While these projects are shining, Delphi failed to support them. People still uses components like Zeus for MySQL connection, because there wasn’t anything to connect MySQL databases at that time.

    Thank you very much again for your time to read my post and for your comment here. I’m expecting good news from Embarcadero, and your comment is also an indication of your dedication to community feedback.

    I am sure everything will be good again..

    Best regards,

  3. July 21, 2008 at 7:42 pm

    Project Tiburon is all about Native code. And we will do more, but we have to prove it to everyone by doing and not just saying.

    “But these are not enough.” Okay. I agree. What specifically can we do to innovate beyond:

    Anonymous Methods,
    Better compatiblity between native Delphi and C++ languages,
    DataSnap enhancements (see Steve’s blog post at http://blogs.codegear.com/steveshaughnessy/2008/07/20/38912/)
    VCL enhancements,
    Unicode support,
    Improved documentation (yes, we have to do more – see Dee’s blog post at http://blogs.codegear.com/deeelling/2008/07/18/38298/)

    How about sharing your favorite list of innovations/enhancements beyond what is in the roadmap (see http://dn.codegear.com/article/36620)?

    We’re communicating and listening ๐Ÿ™‚
    David I.

  4. Excessive
    July 21, 2008 at 8:18 pm

    Dear David,

    Thank you very much for your second comment, and giving me a chance to express my thoughts.

    For me, the most important thing for success of Delphi are:

    – Difference: The IDE must be redesigned that, once users or developers sees it, they must say “What is this? Wow!”, instead of “Oh, it is a flavor of Visual Studio.”. This is the most important item in the Delphi success list. Delphi team once did this, and this is why Delphi had its success.

    – Unique approach for solving problems: Please have a look at previous Delphi versions. They were the tools of the future, in terms of expandibility, usage, feature set, and speed. I also remember that, to keep up with Delphi, Microsoft had to provide lots of extra components for Visual Basic.

    – Unique and beautiful components: Delphi has always provided superior components compared to its competetors in the past. This needs to be inside Delphi again; futuristic, unique and beautiful components.

    – Database connectivity: Even Delphi 1 had SQL Links for Oracle, Sybase and Informix, while other tools lack support on those database systems. People need to know that, “For Microsoft technologies, there is Visual Studio. For everything else, there is Embarcadero, and its beautiful, fast, stable, and superior IDE’s”. Once this take place in people’s mind, everything will be crystal clear, and people will not think again while choosing Embarcadero’s IDE line.

    – CodeGear Framework for web & managed development: If you had to support a framework, it should be yours. Because you have exact control on your own framework, and you can increase pros on your own framework all the time.

    – Documentation: What about writing your own help system? Adobe does it, other companies do it, so what keeps a company which are full of world class development experts from writing a lightning fast, stable, extensible and expandable (for component developers) help system?

    These are key development factors to any product, but it fits exactly to my beloved Delphi development platform.

    Please tell me what you think of these ideas. Thank you again!


  5. July 24, 2008 at 7:29 am

    Dear Excessive and David,

    Thanks for “pushing” requests and “trying” to follow ๐Ÿ™‚
    This is my first post in blog in my life, so please forgive me if I write something wrong…

    First of all, I’m developer with 22 years experience, started to work in Russia with 8-bit DOS and now in South Korea with Linux/MacOS/Windows. During my life I was involved in many-many projects from individual to big internet multimedia (for Hitachi).
    And, sure, I used Borland product from beginning (Turbo series, even TurboProlog) and at that time I was ABSOLUTELY satisfied by Borland products. By my job I had to work with several OS – so Kylix was a best shot for me (and my team)!
    When I stopped to use Borland products ? After Delphi 7… Excessive expressed exactly why…
    What I use now:
    gnu (gcc, objc++, gnat) – multiplatform, standard de facto, static,
    Java – multiplatform, standard de facto, semy dynamic :), very fast

    What I will use in addition:
    Lisp – multiplatform, has standard, full dynamic, very fast.

    What I will never use – NET, Ruby, Smalltalk, Scheme, …
    But I wouldn’t concentrate on mistakes, I would express MY NECESSARY items.

    0. Tool MUST BE INTELLEGENT – It should not make my mind NARROW, like “if you need TCP Server – take this component and connect to that” – programming is NOT construction of SAME BUILDINGS from SAME BRICKS. If you do like that after 20 years of development you will be ROBOT. Development is Architecture. Best buildings are UNIQUE. Best paintings are UNIQUE.
    1. Tool MUST BE multiplatform and crossplatform – without any “but” – even now I develop for Linux/MacOSx86&PPC/Windows XP&Vista this is ABSOLUTE MINIMUM. I didn’t mention about mobile & Solaris which I use from time to time. Will I switch from modest multiplatform gcc to BEST GREAT PERFECT TOOL FOR WINDOWS ? Sure never. It just doesn’t allow me to do my everyday job.
    2. Tool MUST BE native – should use ONLY proprietary code (or if necessary ADOPTED) , but in any case solution should be self contained and not rely on 3d party RTs like NET.
    3. Tool MUST PRODUCE compact & fast code (and it basically should be better than free opensource tools – gcc or SBCL ๐Ÿ™‚
    4. Tool MUST PROVIDE perfect local/remote debug/profiling and bug finding tools (and they should be better than free Valgrind based and Gdb based tools ๐Ÿ™‚
    5. Tool MUST PROVIDE project/file overview, variable/function navigation & refactoring – (and it should be better than free Eclipse/Netbeans ๐Ÿ™‚
    6. Tool SHOULD PROVIDE frameworks like QT, DB Access, Distributed processing and so on, but:
    – This frameworks shouldn’t be huge, multi-multilayered and complex
    – It’s not necessary to implement in Designers ALL components – I very rare use QT Designer ๐Ÿ™‚ Hands type faster…
    7. Tool COULD PROVIDE REPL/Change code without full recomile – it speed up development in times especially with huge source code…

    And just several thoughts:

    Until Delphi 7 it has higher level than average development level, so it pull developers up and attract them. That’s why I like Queen group – it was always higher, but not too much !!! to confuse. Now I feel you try to catch the train, but from my experience it’s wrong way. You have good team & background to make your own plain to outperform those train.

    Just have a look around what’s going on ? Dynamic languages rise. Seaside attracts attention to Smalltalk, Semantic web attract attention to Allegro Lisp, even slow Ruby has his own fans. And Adobe is not sleeping with Flash 10/Flex platform. I don’t mention about CLR/Silverlight just because I don’t like Microsoft. Java platform 7 is almost ready for dynamic & GlassFish is developing fast. Eclipse Ganymede & Nebeans 6.5 could support C++/Java/PHP/Ruby/DB and even Lisp development by plugins.

    What are you doing meanwhile ? You made Delphi for PHP…. “strong” reply and you spent time and money. You are finishing Tiburon. I hope it will compile QT 4.4 anyway it’s not useful for crossplatform or internet/distributed development.
    And what is important you are making one hummer for Static languages, another hummer for Web.

    What should you do (just my guess):
    1. You have to make NEW DYNAMIC LANGUAGE (with static compile options) and compact, fast, multiplatform VM/RT – not like Java ๐Ÿ™‚ What could be compiled could be compiled what couldn’t be – could be compiled at runtime (JIT or like embedded TinyC, say TinyDelphi) – Flash 9 & 10 has perfect speedup, isn’t it ๐Ÿ™‚
    2. Language could be any you want, but it MUST include first class functions, multiple and context dispatch (double/triple/??), GC of course.
    3. Based on this you could make new framework, and, then new IDE.
    4. It should allow users to create standalone and/or distributed multiplatform applications, services, agents. Fast, compact, stable and fully self contained, remote debugged & profiled and even updated at runtime.
    5. It should have bindings or private implementations to main modern technologies, like 3D (OpenGL/DirectX), Multimedia (QuickTime/Core…/Phonon), you choose…

    You have to make NEW TOOL whose abilities is not worse than best abilities of what other already have:
    Gcc, Gnat, Java, Allegro Lisp, Flash/Flex, Eclipse, Netbeans.

    How long will it take to develop ? With your team probably 1.5 – 2 years.
    What will be the profit ? Nobody will never use something else (may be except funs), because it’s COMPLETE, SELF CONTAINED, MODERN (even more than current) system which allows to do everything what other particular tools allow, but with less efforts, and more results.

    I hope you will.

  6. Excessive
    July 24, 2008 at 2:51 pm

    Dear friend,

    Remember Turbo Basic? I have also used it back in time ๐Ÿ™‚

    Thank you very much for your comment, and I appreciate the points you made. As you know, we were saying these over and over again, since Delphi 8. The points and the suggestions you made are vital to the success of new generation Delphi.

    Also, you should post more on blogs. Your comment is extremely accurate and full of information. Hope to see your blog in the future. When you have one, please indicate it here.

    I’ll be honored to read it. Keep coding ๐Ÿ™‚

  7. Zexx
    September 26, 2008 at 1:08 pm

    I’ve read all the comments above, but you’re all aiming too high. Lower the expectations. Let me ask a primitive question:

    IN WHICH VERSION will the damn form editor have the UNDO function?

    It’s not that I’m waiting for that feature for a long time. Only about 13 years. As it is now, if I accidentally (or deliberately) move or stretch some controls, I can return them in previous state only if I saved the form. To make a 5-step undo just save the form in 5 copies. That’s how Borland, CodeGear and Embarcadero think about productivity.

    No, I don’t expect they will implement Undo before December 2019. Since Borland sold their best programmers, their know-how, their stocks and their soul to Microsoft, the development of Delphi has stopped. It was blocked so that Visual Studio could thrive. Nobody wants to say it openly, so here, I’m saying it.

  8. November 28, 2008 at 8:34 pm

    I changed from D7 to D2007 and found lots of improvements, BUT:

    The help system sucks! The D7-help system was better and faster and with lots of examples. Please go back to a native help system that focus on Delphi. Expand the system with good examples.

    Try to focus on RAD tools for generation of Web-forms. Intraweb is pointing in the right direction, but I am sure that you can do even better if you focus on it. Many of us do not give a damn about the finer workings of Web2.0 – we just want to write and deploy the applications as fast and eficiently as possible.

    Get the TeeChart components – or something equally efficient – on the pallette of components that ships with Delphi.

  9. mark santos
    December 22, 2008 at 6:36 pm

    Please revive delphi… for me delphi is still the best. I have use VS.NET but there’s something that is missing. I am still return back to my home which is delphi.

    Support Delphi!!! Delphi Rules!!!

  10. mark santos
    December 22, 2008 at 6:36 pm

    Hi Iam from Philippines. Please revive delphi… for me delphi is still the best. I have use VS.NET but there’s something that is missing. I am still return back to my home which is delphi.

    Support Delphi!!! Delphi Rules!!!

  11. mark santos
    December 22, 2008 at 6:40 pm

    I think it will be very helpfull to Delphi if you guys create a delphi group that is approved and advertise by embarcadero so that it will be popular. i have a lot friends here in the philippines that is very loyal to delphi we can join to support delphi

  12. Descendant
    February 10, 2009 at 3:54 am

    Delphi 7 was definately one of the more significant releases and I perhaps logged more time in D7 then any other version. I prefer D2007 overall and am looking forward to D2009 soon.

    I’ve used Delphi 2,3,4,5,7 & 2007.

  13. Aqge Priwibowo
    June 18, 2009 at 7:19 am

    Wow, until an hour ago, i thought it was my right decision to start learning JAVA. I ve been feeding my family by mastering delphi sold it for almost 5 years. watching all my fellow programmer in indonesia moving from delphi to other languages, (mostly .net) was convince me to tight the belt and push the eject button get away from delphi.

    event sounds like so much drama but is’t true people does fallen in love to delphi. i cant give some expert or useful comment here because i m not that master at all, but i m hoping (with all deep and strong feeling i cant write down) that delphi will return.

  14. September 10, 2014 at 4:30 pm

    Con seguridad uno de los mejores blogs sobre el
    asunto que se puede ver en la red. Felicidades otra vez por
    el utilรญsimo contenido. Un abrazo!

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: