Thinking about this a little more, maybe RS could allow one to have separate folders, say named "runtime", and "compiler" alongside the "plugins" folder. While they would normally stay empty, the developer could then choose to have either of these contain an earlier framework or dll version of ones choice.
If a new RB version was installed and one or more of those folders was populated with an older version of the framework, then when launched, RB would use THAT version instead of the one that came with the newer release. This would allow developers to stay with an older version of the runtime as necessary while benefiting from improvements in the IDE. On 27/9/06 12:21, "Daniel Stenning" <[EMAIL PROTECTED]> wrote: > That issue of ubiquity is - sadly so important. > One reason why RB needs to have enough features to tempt C programmers and > others to "switch" and increase the "pool" of programmers (and > "evangelists") out there. > I was annoyed when Java became flavor of the month, because for no good > reason, the freelance contract market switched and PB became less in demand. > Yet Java was a much worse choice for most IT projects. But it is ubiquitous. > > It is sad to see IT projects being comitted to certain languages, simply > because of the need to assure availability of coders instead of suitability > for the task, but those are the realities of life usually. > > Concerning reliability, I think it would be a good idea if RS made it > possible to easily combine separate releases of the IDE, compiler and > runtime library. That would allow one ( within limits ) to use a more > recent IDE with an older and more stable runtime, or let us to use an older > runtime but with a newer compiler or core language additions. > > Naturally this would not always be possible due to inter-dependencies, but > it would help - for example it would allow more developers to start using > newer features ( and thus help the bug testing process ) while still > safeguarding code from failures due to introduced library bugs. > > > On 27/9/06 05:05, "Norman Palardy" <[EMAIL PROTECTED]> > wrote: > >> My point is that C and C++ are widely used even though they may not >> be "the best" choices & people who are competent in them are readily >> available. >> Why ? >> Because they have solid stable runtimes and perform well enough even >> though there may be alternatives that could gain a bit of speed. >> RB needs to have those same characteristics to gain wider acceptance >> as a quality, general purpose >> tool._______________________________________________ > > > > _______________________________________________ Unsubscribe or switch delivery mode: <http://www.realsoftware.com/support/listmanager/> Search the archives of this list here: <http://support.realsoftware.com/listarchives/lists.html>
