On 3/28/2007 6:19:33 PM, Ted Green wrote:
>
>We are finding it very difficult to get back to basic
>VEDIT development - Tom Burt works full time on VEDIT, but
>has been spending 100% of his time the past 6 months on
>EBCDIC conversion jobs. I even hired another programmer (Tony
>Taylor) for the EBCDIC jobs, but both have been maxed out
>lately.
If it is difficult to find Assembly programmers, the why not hire another person for EBCDIC conversion jobs and let Tom to concentrate in developing Vedit?
>
>As you may know, the core of VEDIT is written in assembly
>language, which is difficult on which to train new
>programmers and not portable.
>My goal is to hire two C programmers later this year to
>rewrite VEDIT in C.
One of the main advantages of Vedit have been the speed and compact size that result from Assembly coding. I am afraid those will be lost if Vedit is rewritten in C. And further, rewriting would be much the same as developing a new program, thus much losing the 30 years of development and optimizing.
On the other hand, most enhancements needed are related to the user interface (e.g. fully configurable menu system, more toolbars, side panels, window splitter etc.). Those are already written in C. Enhanced printing functions could be in C, too.
There are some things that require changing the Assembly code, too. Perhaps the most important would be the Unicode support. More flexible color system would probably require changing the core code, too. (But perhaps something could be done without Assembly coding, such as option to change palette colors.)
If Vedit is rewritten in C, then I hope at least the most time critical functions, such as searching, will be still Assembly routines. (You can create C routines for other hardware environments, if needed.) This way, you could retain much of the speed, even if not the compactness.
--
Pauli
|
|