Hallo René, > First of all, its 'only' a compiler and IDE.
Very true, but it's a very mature and stable one that, on numerous occasions, we use far more deeply than most programmers, have access to tons of code, examples, libraries, articles, books and other supporting materials for, have _very_ long experience with, and a lot of the created code is tailored to work with those compilers. We've been using Borland compilers since TP 1. Even a lot of IDE elements, like keystroke memory, have been deeply ingrained by now <g>; probably right down to the DNA level! > but there are alternatives That do not as yet measure up to these compilers... nor will there be support for the extensive library "extensions" that we use from the enterprise versions any time in the visible future... I have used a lot of compilers over the last 25 years; a bunch of them were Pascal compilers (notably CP/M-86 Pascal MT+86, IBM Pascal, Mac Pascal, and THINK Pascal...). I REALLY don't want to even contemplate the need for moving this *large* codebase, and _many_ libraries, any time in the near future, even though the modern compiler alternatives target Borland's Delphi as the natural standard, and thus should be ever-more compatible with it. I will definitely look at them again at some point (at least for some projects), but they're completely out of the question now. > So, maybe with some additional > work, but (y)our codebase is not lost forever, it is still compilable. One way or the other, the codebase would not be lost; but it could cost an enormous amount of work, on top of what's already been done. That's not acceptable, when the licenses are still valid. > It doesnt care about registration, it just works. Well, that's great news, but they may have handled the registration issues differently for the academic versions. ? > As last cent, with newest releases of delphi going further away from the > traditional pascal, with this .net stuff taking a more and more central > place - whereas i target platform-independant code as much as possible, > i really don't care anymore what latest and possibly future version(s) > of delphi brings, i'm probably going freepascal sooner or later anycase.. I am (almost <g>) TOTALLY with you there, and have been since the monstrosity that was Delphi 8. I was really amazed when Borland went for the .NyET bait hook, line, and sinker (translation: when Borland followed Microsoft completely). It was obviously a fool's game unless Borland could use the MS strategy of Extend, Embrace, Extinguish against MS (by making cross-platform compilers compatible with .NET on other platforms, which was likely impossible for several reasons). It took me quite a while to completely arrive at this conclusion, but I haven't cared about ANY Delphi releases since D7 and K3! And that sentiment is even reinforced by Borland's choices, since those were the last 2 cross-platform compatible compilers! These compilers support the platforms that we're interested in: Win32 and up, and several flavors of Linux... Where you and I might differ in opinion is in the need to move to another compiler. After years and years of being amazed at the new developments Borland brought out with each new version, it's a little strange to feel this way, but I'm pretty convinced that several generations of the old compilers could really be considered "evergreen." (At least as far as we can see now.) The huge base of existing operating systems and computers mean that future environments would have to support today's code just as Vista still even supports legacy code, such that MS-DOS programs from 1981 can still run on it! We're now doing VERY different things with databases, communications, security, and replacing certain built-in features (like the random functions for certain programs), but the core product and much else, is so powerful and extensible (and "clean!") that these older compilers along with all the "history" gathered about them are still the most powerful development environment available to us. At one time, I thought a 64-bit compiler would be a good incentive to "upgrade" or change. Not any more. It's almost silly. For the vast majority of programs, there's no need for it. And in situations where large memory spaces (...) were needed, one could use several communicating 32-bit processes, instead of one big one. Modern hardware performance is just fantastic, and coupled with these good compilers (and decent libraries and a good thread pool ;-), can be made to do anything (very quickly!). Native Unicode support and controls would be nice too, but we'll probably buy a product to supplement what facilities we don't already have... Lots of companies (and academics) are still VERY happily using much older compiler versions, that accomplish everything they need. > Thanks for bringing this issue to my attention though :)) Well, we pretty much knew something like this was coming for the last couple years... As long as the old compilers keep working, even this is a "non-event" for me! And thank you for mentioning the non-problem (!) of running the compiler without phone-home capabilities. With any luck, that's how the other classes of products will behave also! ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ synalist-public mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/synalist-public
