t is important to keep in mind why developers need to know about Borland's plans.
1) Some developers have to make their own decisions about what tools to use and do
not want to start new multi-year projects on tools that are not going to be improved
much, if at all.
2) Some of us have to sell management. If Borland doesn't provide us with enough
information to sell our own management then we will be forced to use other tools and
we view the alternatives as less appealing. There are guys posting here who know
that they are going to be pushed off BCB at some point in the coming months. Others
already have been and are gone.
3) If Borland is really going to do something we want them to do and they would
reveal this information then we could save a huge amount of effort involved in
ranting or sending them messages arguing for what we want them to do.
If a problem is going to be solved we can move on to other problems, of which
there is no shortage given the line of work we all are in.
In other product lines Borland recognizes this need for advance information and
supplies it. Borland doesn't supply this information for BCB in spite of the fact
that the need on the part of the average BCB user for this information is even
greater than that for the average Delphi or JBuilder user. The reasons for that
greater need are:
4) BCB has suffered from serious regression bugs in its last couple of releases and
some of those bugs were never fixed.
5) BCB is behind in versions as compared to Delphi. Borland's commitment to BCB is
obviously much much weaker. Our ability to assure our own managements of Borland's
commitment is weak at best.
6) BCB is lagging in C++ standards compliance.
7) BCB Has long-standing and as yet unfixed problems that predate the regression bugs.
At this point BCB users are being given hints that some positive news about BCB will
be forthcoming. Yet BCB users continue to suffer from an information deficit as
compared to other Borland product users. One issue that Borland apparently does not
sufficiently appreciate about its BCB users is the issue of trust. Consider the
8) We know BCB is less important to Borland than Delphi or JBuilder so we think
Borland could walk away from its unofficial hints for a new version unless it makes a
very public commitment to a new version.
9) BCB has had previous releases that did not address serious problems and instead
introduced new problems. Well, even if BCB is going to be revved to v9 we have very
credible reasons to fear that even then existing serious bugs may not be fixed and
new ones may even show up and then not be fixed.
10) Borland promised its C++ users in an open letter about its C++ plans that it
heard loud and clear the need for better support of a C++ product. BCBX was released.
It had some serious IDE bugs that caused many to try it and give up on it. Many
months have passed and those IDE bugs (eg in BOE have gone unfixed. So when Borland
says that it has heard the message about the need for better quality and support that
hasn't translated into bug fixes.
I only hope that the management turn-over in Borland has not been so high that what I
said above would be news to some Borland managers who are in the decision-making
chain of command for BCB and BCBX.
My message to Borland is that your C++ developer users and their managers have plenty
of reasons not