[The OS/2 Supersite: The Starting Point for Warped Web Surfing.]

Celebrity Corner: Robert Mauro

To celebrate the inaugural edition of "Celebrity Corner" this month we asked Robert Damian Mauro, President of Cybercom, to share some very interesting philosophies his company has regarding software development (and a few other things). Cybercom is the developer of the highly anticipated AccuCount/2 accounting and business management package being exclusively developed for OS/2.

During the course of the past half a year, Mauro tells us that some Windows users have become interested in AccuCount/2 and have inquired about a possible port to their chosen platform. This month, Mauro tells us what Cybercom's answer is to that request, and why.

Why We Don't Write Windows Software

First, a word of thanks to the editor of OS/2 e-Zine! for inviting us to this month's "Celebrity Corner".

We (Cybercom) have gotten numerous messages asking us our position on developing Windows software, about IBM and the OS/2 market. We hope to clarify our opinions on these matters and give you an idea of where we as a company are going as well.

Cybercom is a PC Software and hardware consulting company. We specialize in custom solutions as well as hardware upgrading, repair and system building. Our primary focus is a commercial software package called "AccuCount/2" -- an OS/2 only accounting and business management package.

During our development of AccuCount/2, we've learned a lot about the computer market. First, Windows users aren't quite the mindless lemmings many portray them to be. Many of them honestly do not know about OS/2, or the advantages it has over Windows. Second, many Windows users seem to be looking for solutions other than Windows, or better software for it.

Also, we do not believe the OS/2 market is dead, no matter how others portray it. We think most people just aren't aware of OS/2. I don't know how many times we have consulted with a client or potential client and suggested OS/2 and have been asked, "what's that?" or, "does that run under Windows?"

Yet still, with all the doom-and-gloom out there, we have gotten requests for AccuCount for Windows. Our response... "Sorry, no."

To this reply and to our explanation why, we have actually found some Windows users ordering our software, and planning on dedicating a machine to OS/2 so they can run it.

Why won't we develop for Windows? Many reasons. One is that we don't believe Win95 is a viable platform for "Mission Critical" software, and we don't believe WinNT is up to par when it comes to performance/hardware ratios. The abysmal attempt to use NT on the MSNBC servers or PC Magazine's tests of Warp Server Single Processor against NT on 4 are perfect indications why.

What do we define as "Mission Critical"? Well, to us, "Mission Critical" is anything you run, anywhere you run it. In other words, when I get home from work, I play games or surf the web -- does that make it less "Mission Critical" for me? If you bought a new car, and it ran terribly, but you only used it for joyrides around the block, would that make the drive less "Mission Critical" (and excuse the breakdowns the car suffered from)?

Whether you are a home user, SOHO user or a Fortune 500 company, we think everything should be treated as "Mission Critical". To us, that alone discounts Win95 as a viable development platform. Even Microsoft claims if you need a "Mission Critical" platform, you should use NT. To us, that is an admission of failure for Win95. It's akin to saying, "If you don't mind something that is buggy or poorly written, use 95, otherwise use NT".

As for NT, MS still doesn't seem to know for sure what direction they are going today. Cairo is getting later by the year, and many promised innovations we keep reading about just never materialize except as another year appended to the product name that we never seem to see. We can't justify writing products for an inefficient OS whose features are vaporous.

Well, what about a port? NO. One thing we feel about OS/2 and the OS/2 market is if we (the OS/2 community) are right (and I believe we are), and OS/2 is the best commercially available OS for the PC, then shouldn't we have the best software as well? There are some great companies out there -- like SPG (ColorWorks), SouthSide Software (PMMail), and The Keller Group (FaxWorks for OS/2) to name a few -- who seem to realize this and deliver top notch products that exemplify such a belief.

In developing AccuCount/2, we decided from the beginning that our software should also exemplify this belief. In (hopefully) following through on this belief, we are nearing our release date for a product that incorporates features that just aren't possible to implement on a platform like Win95. By "implement", we mean, "incorporate such features in so the features are usable and not a detriment to the speed and performance of the program or other programs simultaneously running".

Some of the areas we cite are multithreading and interprocess communication. "You can implement those under Win95!" some people tell us. Sure, with a lot of work, we can, producing a package that is very slow, hogs the system, and proves once again that Win95 doesn't multithread or multitask well. No thanks. We refuse to put our name on such a port and we refuse to let anyone else port our product and release such "crippleware".

Why don't we de-thread the program and/or remove the OS/2 specific features? Our plan and our company paradigm is to release state of the art software, with advanced features. Such a decision would not fit with that goal. That would just amount to releasing another version of everything else out there on the WinXX platforms.

A word to developers...

To their credit, IBM has been a phenomenal help to us. Yes, we said, "IBM." When we ran into problems obtaining information on the new PMPrint Engine in OS/2 (which does most of the device independent printing under OS/2), Sheila Ryan and Marta Gomez were instrumental in helping us. Instead of just providing us with a little documentation or a WWW URL, they passed us along to Mark Vanderweile from Warp's PMPrint Team. He not only answered specific questions both by voice and e-mail, but also sent us documentation, sample code, actual PMPrint code, much more than we hoped for -- about 11MB in all! We were stunned. With all the complaints about IBM's developer support, we didn't expect it.

We also are very lucky to have two local IBM reps who have been fantastic in their assistance who should also be mentioned. They are Michelle McKnaught and Craig Stone. These two work primarily with large (Fortune 1000) companies, but were always willing to spend the time needed to help us out, finding us software, development tools, tracking down leads or departments in IBM to help, providing us with promotional literature -- you name it. They realized we had a need and no one else locally to fill it, so they picked up the ball and got things rolling for us.

As for the horror stories you may have heard, yes, they are true. We could tell you many ourselves. For instance, it took months to get the information on PMPrint that we needed. Not because of inefficiency on the part of someone at IBM who promised it, but because we weren't talking to the right people. Many people claim, "Well, I have talked to 50 people at IBM and haven't gotten any help." IBM has roughly a quarter million people working for it so talking to 50 really doesn't say much. Perhaps, like us, you will get lucky and hit upon the right person. Perhaps, like in many cases, you will not.

Our suggestion: first, join the IBM Solution Developer Team. They are a great bunch who care about their job. They've gone through some reorganizing lately and some developers have temporarily suffered because of this, but once you find your support person, they can be a great help. Second, don't give up. If IBM can't help you, try someone else. Perhaps someone outside IBM knows who inside you can contact or can pass you to another developer who has a tool or trick to suit your needs. For us, such companies were OnLine Data, who has a very nice database development tool set, The Keller Group with FaxWorks for OS/2 for AccuCount/2's fax integration and SouthSide Software with PMMail for AC/2's e-mail integration.

The bottom line is, OS/2 is the only platform we will develop for. We support others for our clients that use them, but will NOT develop for those other platforms. Finding support and technical information for OS/2 may be very difficult at times, but the rewards of sticking it out are phenomenal. I, for one, couldn't live with an imitation of other WinXX packages when I know what our software can do under OS/2. All of us at Cybercom are committed to releasing top-notch software -- not software that is "feature competitive" with its WinXX counterparts, but instead, software that sets new standards that Win XX counterparts cannot match.

Robert Mauro, President of Cybercom, is one of their lead programmers as well as an A+ Certified Hardware Technician. He went straight from DOS to OS/2 2.0 beta and never looked back.

[Index]  [ Previous] - [Feedback] - [Next ]

[Our Sponsor: J3 Computer Technologies - Serving the Global OS/2 Community, large & small!]

This page is maintained by Falcon Networking. We welcome your suggestions.

Copyright © 1997 - Falcon Networking