Managing director of
Heart Informatisering B.V.
Heart - Profit Systems B.V.
Has some experience on FoxBase, FoxPro, Visual FoxPro and ?ERP;
Being an oparator some 28 years ago, used to watch flickering lights of CPU vs. I/O on Philips Mainframe P1400 with main memory of 256 KB and mass memory of 1 MB (yeah) during night-shift in order to be home earlier.
Started programming (Fortran) in order to win a primenumber speed contest, which he did. Having the results out of 1,000,000 in a few secs (how many Khz was this P1400 ?), started to work on the derival of a formula for primenumbers in which he almost succeeded, but stopped because it was boaring.
Developing a chess-program was not, until the creator couldn't win of his creature anymore.
Worked at several larger companies such as Shell during his stay at BSO/Origin for 7 years, and made a hobby out of normalization in not using the 3, 5 (Codd) or even 9 known forms, but invented some more (time-based). Created a several-hundred user system on a PDP11in Mumps with super-response.
Wanted to start a company as part of BSO/Origin in 1986, developing mainframe-apps based on PC-networks after some experience in DBaseIII. No one believed in this stuff, so he started his own company in 1986, Heart Informatisering B.V., and selected FoxBase+ (multi-user, yeah !) for the development-tool.
Too much was asked from FoxBase and regular contact with Fox Software fixed the most crucial bugs (thanks for the Dutch Distributor of Fox too !).
When the first version of Heart - Profit ( ERP) was running at several customers, Microsoft bought FoxPro, and from version 2.00 Heart - Profit started to run under that. Okay, it did not, because the memory-management of 2.00 was really nothing. Helpdesk numbers of Fox Software became numbers of Microsoft, and 2.00a was doing it's job a lot better. 2.00a became 2.5 and now everything was officially running, though the app contained several hundreds (or more) work-arounds.
2.6 (Dos) was released, but problems started again and it was layd aside.
Some 6-7 years ago VFP 3.0 was tried for a few simple things, but again, causing too much memory-problems; in order to start attempting to run Heart - Profit graphically, one year was worked on 2.6a (Windows), and just when this was finished with nice result, of course this was not it, though all big ERP-apps were creating a graphical shell, having no real objects etc.
Now VFP 5.0 was tried, en yeah, this seems to work allright. A real OS was created in order to map the procedural Dos-code to the Object-environment, and in the 4 years of working on this, no more than 10 real bugs where discovered in VFP5. This, for being the first VFP5-buyer in Europe (ie very premature official version) was nice stuff.
Over a year ago again a very first version of VFP6 was bought, but messed up all classes due to having used custom method-names etc. now existing as VFP-names in 6. Stupid, and own fault (hmm ?).
Right at this moment 2.5 is still used for the Dos-version of Heart - Profit, and VFP5 (no SP) is used for the Win-version, not giving any problems.
The app is developed in simple (!) procedural code in Dos (over 7,000 prg), but runs Object-Event for Win, allowing for over 100 users on one server, not noticing they are all there on a database consisting of 800 server-tables and 1,400 including local tables. One logical transaction may use 80 tables and in such an environment over 1,000,000 transactions may be performed in one day. All together Heart - Profit has several thousands of users, and a helpdesk of one person is sufficient.
Peter's explicit wish is to make the impossible possible, doing fine right now. Peter -of course- is not Peter-only, and the employees for sure must be included.
Where Heart - Profit uses native tables, one of the main tasks for the next period is the DB-independancy, and Peter "promises" to succeed without re-writing the app which just is impossible. He 'll let you know when he (no, we) succeeded.
Besides form this, all functionalities will become COM-objects (or whatever), making all the thousands of functionalities available to all (other apps). May take another few years, but just wait ...
All together Peter "lives" from tuning things, not wanting to perform even one I/O (etc.) extra where it is not needed. For that matter, Peter believes in normalization of not only data, but of all processes. His general idea : eliminate all redundancy you can think of, and responses are optimal by itself. Not only system-responses, but organization-responses too. Peter's own organization is the first example, being able not only to create the big app with a handful of employees, but being able to maintain it too with all of it's users.
Peter therefore is proud on his employees, willing to think the same as he does.
ps at heart-profit.com
( Topic last updated: 2003.06.19 10:12:14 AM )