Borland C++Builder Revolt

florescent_beige writes "Developers using Borland's C++Builder RAD tool are in revolt. Borland apparently obsoleted this product one year ago. However, the promised migration path (to be described in a now infamous open letter) never materialized. In a last-ditch effort to convince Borland to support them, users have put together a letter justifying (and begging) for continued support."

Same old Borland

[Anonymous Coward]

I am shocked that anyone thought Borland wouldn't pull a stunt like this. They have a consistent history of making products that are a great leap ahead and then just sitting idly by as everyone else catches up and surpasses them.

For developer environments, the software does in fact "rot". For example new APIs and components are introduced in the underlying operating systems, and the compilers and libraries etc usually have to be updated to use them correctly. I remember sitting idly by during the Windows 95 beta programs waiting for a version of BC++ that could actually produce Win95 programs. (It wasn't a big leap as they already produced Win32 programs for NT but some updates were needed for win95). As another example, you need some tweaks done if you want a Win32 program to use the proper widgets under WinXP.

Unless there is steady stream of effort to at least keep up to date, and fix bugs then the development environment will become useless.

Re:I agree Builder is not Open Source...

[Anonymous Coward]

The Professional & Enterprise editions of BCB included the VCL and the RTL (although maybe not all of that; I can't remember). In BCB5 Pro & Ent, TASM & DCC32 were provided for compiling assembly & Object Pascal when needed - don't know about other versions, but I would expect the same is true of BCB6 Pro & Ent.

The majority of the Delphi & BCB IDEs are implemented in Delphi with VCL, while the JBuilder & CBX IDEs are built in JBuilder. I don't know about Kylix.

There used to be a patch project for OWL much like what you describe - don't know what the status of that is these days.

T

Re:not verb

[mjc_w]

From dictionary.com:

obsolete

\Ob"so*lete\, v. i. To become obsolete; to go out of use. [R.] --Fitzed. Hall.

Source: Webster's Revised Unabridged Dictionary, © 1996, 1998 MICRA, Inc.

Re:The top gun...

[seanyboy]

It's Anders Hejlsberg. [wikipedia.org]

Re:Borland dead? I don't think so.

[Anonymous Coward]

This really look interesting. I would gladly jump in this bandwagon BUT, I'm not in the OSS stuff. How the heck am I supposed to distribute applications on the internet without also distributing the sources? If there's a way, I'd like to know :)

This is certainly possible. Remember how you distribute Java executables without source code? You compile the source to bytecode, then distribute the bytecode (after some formatting and massaging) to be run by the JRE.

Python will work similarly, but since the Python runtime is NOT a standard piece of installed software on a lot of machines, most "compilers" will throw the Python interpreter and a large chunk of the standard library into the package as well. Depending on the size of the program you're writing, this can have an impact on your total executable size-- be warned.

Note that some of the compilers will need a copy of the Python source tree. Some won't. py2exe is a good example of a program that doesn't need the Python source-- but it's Windows-specific. A cross-platform compiler that doesn't require the Python source tree is cx_Freeze (Google for it).

If you're positive that your customers have a proper version of Python installed on their computers (plus any other dependencies!), you could also try distributing .pyc files, but that's probably NOT what you had in mind.

Hope that helps!

Re:Borland cannot be trusted

[NavySpy]

Let's clear up some of the FUD in your post, Mike.

Borland has not killed off Kylix. That is FUD. Kylix is "on hold" this year, but they did announce some open source initiatives at this years Borland Conference. Kylix isn't thriving at Borland, but it's FUD to say that it's been killed.

Borland /may/ kill off the VCL side of C++Builder. They have set Dec 15 as the deadline to announce whether C++Builder will become part of the BDS/Galileo IDE or not. If they miss that deadline, then we know the answer. But it is FUD to say that it has been killed.

CodeWright has been killed, yes, and incorporated into Galileo and Primetime IDE's.

Borland never made any promises about CF support in Delphi for .Net, and to say that they have is FUD. Borland very much wants to do this, but MS is dragging their feet. It's a limitation in Delphi, yes, but to portray it as broken promises is vicious FUD.

Delphi works fine on XP systems with SP2. What are you on about here?

Man, clearly all that baloney up took some work.