Collaboration / Beta Testing

What is language unindependent is the so called “monkey test” . :laugh:

aaraaayen wrote: Pls let me know What is Thread Testing Thread is rated for shear and tensile strength. Testing involves subjecting thread to stress, and measuring the force needed to tear or shear. Or something. ---- It appears that everybody is under the impression that I approve of the documentation. You probably also blame Ken Burns for supporting slavery. --Raymond Chen on MSDN

In re-reading your original question, one other issue may exist -- input. If you cannot type your language with a conventional keyboard, then your interface of your application will have to map keystrokes to the characters of your language. To simply the whole system (as for a project), you might build a simple text editor or memo field editor. The input algorithm could translate characters. If on the other hand, your language involves words of different lengths (when translated back and forth to and from the human intelligible language), then you will have to build a dictionary which looks up the word of one language and converts it to the other. This content can be pasted or programmatically transferred to targeted media, and re-converted on the other end. If you have a language which the human understands however, you hav no computer project at all, except perhaps in the case of searching, where your dictionary is going to have to translate from the new language to forms which are friendly to the search engine.

I tell you frankly that imho the most promising direction to pursue is probably C++ CGI. I too write in C#, but .Net programmers tend to think everything new under the sun is somehow better. CGI is probably much more secure and efficient -- particularly if you only transfer data across bandwidth. Your MySQL driven content will fly in C++, and the operating overhead will be minimal. What you are doing in .NET may seem to be powerful, but you can deliver any web content in very straightforward, efficient ways from C++. All you have to know to do anything under the sun is basic DBMS implementations (record maintenance and navigation) and raw XHTML. A tool which was a good alternate solution until a few years ago was Delphi. A prebuilt web server component allowed you to read your tables and build your content into a simple CGI run by the server (of different kinds). You can't really get more straightforward than that -- and the peformance is so much better than .Net it's amazing it isn't the mainstream approach. For good material on this alternate approach, you can pick up a used copy of Mastering Delphi [7+...etc.] by Marco Cantu. A single chapter covers this approach -- and there isn't anything you can't do with the *basic* process, except embed all the ActiveX your .Net implementations require... which is a huge security risk to content consumers. If *I* were dictating security policy, I would disqualify .Net on this count alone.

Hourly charges allow for abuse. Jobs should be by contract unless there is something about the job which is so hypothetical its cost cannot reasonably be projected (accurately enough to be bound by contract). A person who maintains a real estate website at $50 per hour is very expensive if they produce 1 or 2 pages an hour. A person who produces higher quality pages at a rate of 20 or 30 pages per hour however (by pre-designed format of course) can make 2-300$ per hour at $10 per page. In the 90's purported web boom (false), the Industry Standard (a principal e-magazine about the false boom) published figures showing per-page costs to be in the $1,000 range in the early 90s, to $1,300 per page or so in the later 90s. If I could have found those clients, I'd have made many millions. In truth, much web development competes with relatively amateur technical skills. If you make good earnings, it almost certainly will have to be by contract and high skill. There should be few clients these days willing to hire anyone by the hour.

You'll find many answers if you research exception handling for various OOP languages. One language should be enough for diverse material, except perhaps that type safe languages may somewhat de-emphasize type mismatch errors. Research error codes and the built-in exception types (yes, exceptions are actually types in OOP) of mainstream languages such as C++ and C#.

I did not know it before today! There is even a special tool for thread analysis! http://www3.intel.com/cd/software/products/asmo-na/eng/threading/219785.htm[^]

Following definitions are according to ISTQB syllabus (International Software Testing Qualification Board, something like a standardisation gremium) "Load testing" tests the behaviour of a system (or component or whatever) with increased load, e.g. the number of users on the system, number of transactions etc... Goal is to determine what load can be handled by the system. "Stress testing" evaluates a system at or beyond the limit of its specified requirements. Since that is exactly what you need to do for load testing (to see how much your system can handle), both, load and stress testing can be regarded equivalent. "Performance testing" determines the performance of a system, i.e. the degree to which the system accomplishes its tasks within given constraints in terms of time and throughput rate. Example: You have designed your system for 10 users. In load / stress testing, you test its behaviour with 1 user, 2 users, ... 10 users, 11 users. Focus of your test here is not the speed (it may get terribly slow with 10 users) of the system but if it does not break, if it really allows 10 users (and not only 8) etc. In performance test, you verify that your system meets given timing constraints. For example, if your requirement is that the system should respond within 10 seconds in any case, you will focus on stopping the time. Of course, you will have to test the speed with a different number of users (1,2 ... 10) logged in, since that may affect the system. So somehow, both test strategies are intertwined. Did this help a little bit???

Hmmm. When you have written it as a learning exercise, why do you want to have it tested? And how intensively?

Hi Alleyes. I'm interested. Mail me at @: s20237736@nmmu.ac.za Artman