Well, according to my experience, java *is* slower than C++, but in most cases it is totaly insignificant when used in an interactive context. Besides that, I think that the efficiency of a program depends much more on the programmer itself than the language. I have seen enough C/C++ programs that behaved far worse than similar java ones.
Going back to the initial problem, if such high mathematical computations are involved, i'll suggest to use on the server side a computation oriented language like Fortran. C++ is fairly comparable to java in that regard. Cheers, Florin -----Ursprüngliche Nachricht----- Von: John Wright [mailto:[EMAIL PROTECTED] Gesendet: Freitag, 28. März 2003 13:24 An: [EMAIL PROTECTED] Betreff: Re: [JAVA3D] AW: [JAVA3D] speed question Florin, as usual you provide some very good responses (and are a credit to our community). My quip (which I now regret making) was more related to the implication that C/C++ was needed (vs Java) on the server and my thought was that in a remote situation the performance difference between Java and C/C++ would be trivial (compared to the transmission time). - John Wright Starfire Research Florin Herinean wrote: > > Actually, I can imagine several situations. > > For example, simulating some nuclear reactions in a particle accelerator. > You can get the first positions and display them, while waiting for the next > results. The results may be calculated at specific discrete times, while the > client can display nice animated morphing as to give the illusion of smooth > transitions. The computations involved in such a scenario may need several > seconds or more per result set, depending on the accuracy degree. That makes > the transmission speed insignificant, especially in the case of a local > network. More than that, that solution gives the posibility of simultaneous > brodcasting of an experiment to several client stations, for example in a > particle physics laboratory, where the data is taken from real-time > experiments. > > Cheers, > > Florin > > -----Ursprüngliche Nachricht----- > Von: John Wright [mailto:[EMAIL PROTECTED] > Gesendet: Donnerstag, 27. März 2003 22:13 > An: [EMAIL PROTECTED] > Betreff: Re: [JAVA3D] speed question > > Galen Wilkerson wrote: > > > {snip} > > > > - make calls to fast C/C++ routines on the server that the applet > > downloaded from > > > > I find this rather humorous. "fast C/C++ routines on the server" If > you are calling a remote routine (even on a LAN) the transmission speed > slows this down to the point of dwarfing any difference between Java and > C/C++. Why do people still seem to think they need to use C/C++ for > speed? > > - John Wright > Starfire Research > > =========================================================================== > To unsubscribe, send email to [EMAIL PROTECTED] and include in the body > of the message "signoff JAVA3D-INTEREST". For general help, send email to > [EMAIL PROTECTED] and include in the body of the message "help". > > =========================================================================== > To unsubscribe, send email to [EMAIL PROTECTED] and include in the body > of the message "signoff JAVA3D-INTEREST". For general help, send email to > [EMAIL PROTECTED] and include in the body of the message "help". =========================================================================== To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help". ==========================================================================To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help".
