CompuServe Thread

#Scientific software ???

11 messages in this thread
#32871From: Herbert L. CrowderAug 24, 1991 4:26 PM
To ALL Amiga Users: In the editorial in this month's Amiga World I was suprised to see their reader survey revealed "desparing wails" about a lack of scientific software. As a physics graduate student I always thought of scientific software as something you wrote when youy needed it. I suppose I've been overlooking an opportunity. I have had my Amiga for over a year now and can program reaonably well in C and use most of the Amiga libraries. I have an MS in physics and did much FORTRAN programming for my thesis. I do not have a job just now so I can devote most of my time to programming. I do have one Question for any Amiga Users who want scientific software: What do YOU want, exactly, in the way of scientific software. I would not want to waste too much time on a program to calculate Clebsch-Gordon coefficients, for example, unless I knew there was some Amiga User other than myself who even knew what a Clebsch-Gordon coefficient was, and wanted a program to calculate them. Also I would welcome any comments on the relative merits of trying to sell software to publishing companies vs. distributing it as shareware vs. trying to setup my own company to sell it. Herbert L. Crowder – via Whap!
#32888From: John ParkerAug 24, 1991 10:29 PM
My Scientific Software wish list: 1) A function plotter with high quality print outs. Every function plotter I have seen just does a screen dump/iff picture of the plotted function. I would like the kind of quality output that is provided by MultiPlot-xlnc (a public domain data plotter). The hard copy print outs from MultiPlot look exactly like those from an HPGL plotter. 2) A Finite Element analysis program (for structural engineering). Not quite your field, but maybe you're interested in such things. 3) A MathCad type of program. I have heard of a program called Maple but have not seen it work. I like the presentation of equations that MathCad uses. 4) Statistical Software. I have seen one Shareware version where the author was asking $150 (If memory serves me). That is abit stiff on price for me. 5) A low cost Fortran compiler. I know of AC Fortran ( $ 195) and BCF Fortran (all of the error messages are in German). 6) A vortex-lattice program for aerodynamic analysis. (If I ever break down and buy AC Fortran I could port one over to the Amiga) I generally go into school when I need to use a program of this type, but it would be nice to see them for the Amiga. I would generally pay around $60 – 70 for programs of good quality.
#32908From: Jeff SchweigerAug 25, 1991 4:20 AM
John – Maple is a symbolic algebra program, similar to Mathematica. Jeff Schweiger
#33104From: Herbert L. CrowderAug 28, 1991 12:50 AM
John: Thanks for your ideas. I have been working on a plotting program, but so far concentrateing on the user interface (still struggeling with Intuition). I had decided I wanted higher resolution output and am thinking of PostScript or HP PCL, perhapes both. If you have other ideas for output let me know, I might find some way to test them, but for now all I have is a laser printer with PS and PCL. A stastical software package is a possiblity for a second project, I'll be thinking on it. I don't know enough compiler theory or machine language to start on a FORTRAN compiler, perhapes some day just don't hold your breath. I don't know what MathCad is, is it anything like Macsyma? If so that is also a bit above my current level of knowledge. I would need to bone up a bit for a finite element analysis program or a aerodynamic analysis program, maybe when I have some income from two or three programs I would have time to study them. What is vortex-lattice, a adaptive grid scheme for CFD? Thanks again, I am keeping these messages so I will have some idea who might be intersted in Beta testing software. Herbert L. Crowder – via Whap!
#33183From: Matthew J. W. RatcliffAug 29, 1991 2:21 AM
Mathcad is like a numerical solution version of macsyma with a nice equation porcessor with wysiwyg appearance, etc. it runs on PCs. Mathematica would be real nice… Clebsch-Gordan coeff.s would be useful for those physical chemistry dudes (or chem. physics!) who actually need the values, which means that they will probably be nasty to figure anyway. for simple systems, it is enough to know that they exist, without actually evaluatoing them…
#32892From: Michael SternbergAug 24, 1991 10:52 PM
Herbert I'll send you e-,ail on my thoughts re: scientific software. In short, ther is a tremendous need on several levels. Be in touch with you soon. Michael
#33116From: Rick CrossAug 28, 1991 12:35 PM
As an author (of MS-DOS programs) distributed both as freeware and shareware, I can speak for the shareware side a bit. Shareware can work very well. The key word is CAN. The program has to be usefull to the end user and easy enough for them to use to make them want to use it. (If it's easier to do it another way, they will). Also, programming for a niche market can also cause "problems". In other words you shouldn't expect 10,000 registrations. The biggest thing I've found in shareware is that crippling doesn't work well. Crippling can be anything from handling a reduced amount of data, to missing features, to time limits it'll work, etc. With missing features and reduced data handling, the user can't truly evaluate the program (which is what shareware is all about). A database limited to 10 records might be blazingly fast… but when loaded with a 1000, might be the slowest of it's kind available. (Sort of like taking a new car for a test drive, but not being allowed to take it over 25 mph). Time bombs, ala a program set to work until a certain date and then quit causes lots of problems for sysops and users alike. (no one likes to download a program and then find out it expired three months ago, especially if it took a long distance call or was obtained from a service like CIS. [More]
#33117From: Rick CrossAug 28, 1991 12:36 PM
[Continued] And even setting it to work for only 30 days after installation is weak. Many times, I've installed a program to check it out, and then not gotten back to it for months at a time. This even applies in my case to some retail software. Plus the easy way around it is to reinstall. The best system found so far is to include reminder screens of the program's shareware status. With maybe a 3 second delay or a random key that has to be hit to get rid of it. This too can be carried too far. After setting up Navigator on the Mac (a retail program) and TapCIS on my '386, I decided to setup Whap! on my Amiga (they won't let you register Whap! until you have it setup and running). While offline and using the Whap! program, it throws up a GUILT screen for something like 8 seconds (even on my 33 mhz '030) BEFORE and AFTER making a selection in the program. This meant that just INSTALLING the program had extra minutes added. This was so frustrating, guess what when into the bit bucket? (Yes I know it doesn't do it while online, supposedly, although it DID happen once when it took awhile for Whap! to realize I was online). So I use TapCIS on the PC exclusively for all my CIS work now. (When I have extra money to toss around and time to waste, I'll prolly grit my teeth, reinstall Whap! so I can order it). [More]
#33118From: Rick CrossAug 28, 1991 12:36 PM
[Continued] The greatest successes in shareware have been fully functioning programs. GO ASPFORUM on here and check their organization out. For those who don't know, ASP stands for Association of Shareware Professionals. The members range from people like me (I've applied) who do this for a hobby, to folks like Chris Magee of AutoMenu (an IBM program) and others who have become millionaires from their shareware programs. The goals of the ASP are to promote SHAREWARE as a means of obtaining quality software, arbitrate between authors and users if there's a problem (if the author is a member), and to help authors out with information on distribution, programming (you can pick the brains of some of the best in the business), and marketing methods. There are only two requirements for membership in the ASP. #1, your program submitted for membership must be considered "non-trivial". Basically, that has been interpreted to mean "if one of us could hack it out in a weekend, it's trivial". #2, it and other programs you release must NOT be crippled. Non-crippled also means that full documentation must be provided on disk. The only differences allowed between trial and registered versions are that the trial versions can have reminder screens (not to exceed 2) which require not more than 3 keystrokes to get rid of, and which if autotimed, last no more than 3 seconds. Also, I belive that as part of the registered package, you are supposed to supply printed documentation. [More]
#33119From: Rick CrossAug 28, 1991 12:36 PM
[Continued] I know these requires sound strict. But following these guidelines have made a LOT of people a lot of money over the years. Plus, they have distribution programs for authors (mass mailings to distributors, et al) and provide approved vendor and approved BBS listings. These are distributors and bbs' which have met ASP's standards for truly explaining what shareware is (ads suggesting that the $5 you spend for a program is all you'll have to pay are forbidden for instance). The ASP also lobbies Congress to protect shareware (the recent Software Rental Act, original contained the phrase, "Public Domain Shareware", an oxymoron. The ASP fought to change it (and succeeded, it was changed to Public Domain Software). You can get free legal advice from copyright attorneys (Lance Rose, Esq. is a fine one and is on CIS in the ASPFORUM) and even help distributing your product in the international market. Currently ASP is all IBM and Mac (they just started the Mac stuff) but supposedly the only reason Amiga authors aren't members is that no one has applied. Why Shareware over Commercial? In shareware, while you pay the costs of advertising and distribution (normally via disk vendors, user groups, bbs' and services like CIS), you also get ALL the money! Going retail, you only get a percentage of the profits, and someone else has full control over distribution, advertising and pricing. And I think it's safe to say, most of the people that would use your software have UseNet, FTP, etc. access and use it to get a lot of their software.
#33180From: John GagerAug 29, 1991 12:51 AM
Rick, I agree with a lot of ASP's philosophy on ShareWare. It is agravating to spend the time downloading a program only to find that SAVE or some other important function has been disabled. When I find out a program has been badly crippled it usually is deleted immediately with no remorse. But I do not consider programs that omit features vs the registered versions as being crippled. For example, the demo version of Whap! is a fully functional program with the most important features enabled. If you like the demo, then you can get a few "Extra" features plus get rid of the guilt screens, which I admit were annoying but tolerable. All of the programs I have written and released have been freeware. But I'm in the process of writing my first ShareWare program. The program will be fully functional as it stands, with the appropriate documentation. But if you send in the ShareWare fee, then you will get extra modules along with the main program which adds more features. I'm not sure if this is against ASP's philosophy, but I believe it is a fair way to handle ShareWare plus get some return on your programming effort. John (in beautiful downtown Walla Walla) – via Whap!