A few thoughts about the near future of KDevelop.

Roland Krause rokrau at yahoo.com
Thu Aug 2 01:03:46 UTC 2001

First, let me say a huge "Thank you" to all contributors to this
release,  especially to (Giga)Falk who contributed numerous hours
to get the MDI interface stabilized.

I am now able to use KDevelop at work on a fairly large code base.

My work project consists of ~ 2100 files with 615000 lines of code
written in Fortran, C and C++. It uses a custom build system based
on GNU make and native compilers on currently about 10 different
platforms. The main development platform for Unix is currently SUN
Solaris 6 and 7.

When being introduced to such a large code base, the most important
thing a user expects from the IDE is support for searching, cross
referencing and code browsing. I usually work having 10-12 files
open at the same time, opening and closing files as needed during 
the process of developing code. 

Having the information, which files are currently edited, at hand
is very valuable to me. Such is the possibility to switch between
definitions and declarations of variables, classes, subroutines
etc. in a fast way. I.e. without having to go through many steps.

All these features have been either enabled or greatly enhanced by
the MDI interface.

That said, there are a bunch of features that I, and my
"Studio/Emacs/NEdit spoiled"  colleagues are missing.

First of all, a modern, scriptable and extendible editor. I would
like to phase out the old kwrite based code, that we currently
more or less maintain. At its place I would put a kate part.

There is a proof of concept patch for this already. The patch is
based off the  KDEVELOP_1_4 branch from a few weeks ago and can be
rebased in a matter of  hours.  There is some substantial clean up
work necessary for this to go  in. A project to implement a kate
part could go in three phases:

1 change all menu and toolbar stuff to KAction, resp. KStdAction 
2 apply/update the kate patch 
3 fix docviewman to work with kate and reimplement CEditWidget to
  be a flexible class derived from a popup menu.

I believe that kate is available and usable now. It is not the
greatest editor of all times, agreed, but it is also maintained.
Additionally, afaik, noone uses kate's editor interface currently,
few people seem to use  kate at all. If we want to improve kate
for the KDE project in general,  someone has to start actually
using it.

I would like to propose to create a branch KDEVELOP_2 off the
current  KDEVELOP_1_4 branch. KDEVELOP_2 would lead to
KDevelop-2.1, to  be released seperately or with KDE-2.89 if HEAD
is not ready.

Bugfixes can be merged down to KDEVELOP_1_4 if applicable.
KDEVELOP_1_4 can  lead to KDevelop-2.0.1 and be released with


Roland Krause
In the garage of life there are mechanics and 
there are drivers. Mechanics wanted!

Do You Yahoo!?
Make international calls for as low as $.04/minute with Yahoo! Messenger

to unsubscribe from this list send an email to kdevelop-devel-request at kdevelop.org with the following body:
unsubscribe »your-email-address«

More information about the KDevelop-devel mailing list