2/8/2012 6:31:49 PM #

>> As previously noted, I haven’t played with FireMonkey  much

I'd wait for XE3.

The design time experience is horrible and the performance isn't spectacular either at the moment.

Jeremy North |

2/9/2012 8:10:52 AM #

The point with firemonkey isn't that it's perfect out of the gate, it's that it's on an interesting trajectory, and the eventual performance and feature set that is possible puts it in a league of its own. Right now it's "beta" level, but really really interesting.

Warren

Warren Postma |

2/12/2012 5:46:23 AM #

I remember back Anil Korkurt 'saying' something like, 'If you want to achieve something in engineering business, you will have to get your hands dirty'. This is the Engineering Business.
EMB does concerning more sophisticated application of computer graphics with the FireMonkey, I hope people/users will not wait for the turn-key solution and also get their hands dirty (engagement at every level) before. Progress does come when the whole network is focusing on one goal. Wait and see is a convenient perspective but a stopper argument. Tennis at the baseline is nice for pensioners and hobby tennis players at the amateur level. FireMonkey imo is more about Serve Volley. Just my opinion. I personally think, don't wait until something so called 'better' is shipped.
The FireMonkey from the overall perspective is one of the most convenient ways to get in touch with the world of 'computer graphics' on one hand and on the other the decision to put the focus on the HID part is imo a first realistic step to on-board a broader audience into a challenging but interesting topic.
If we think back, from vector screens, parallelism, vector computers, vector graphic cards there have been many approaches aiming at compensating the absence/shortage of computing power required. Finding the compromise for a specific situation on one hand and on the other hand consolidating different requirements but a growing number into one technology is challenging and does require input from those who intend to use.
The MS approach until 'today' has been more the thread parallelism but fast calculation in sequence in pure C, this is a multi purpose approach that does work well but requires in depth knowledge of the underlying hard-ware and the philosophy implemented in hardware. I am not sure if this approach found it's limitations on a long term. The moment special libraries and C as well a parallelism do come into focus, is the one to get in touch from the point in time perspective. CUDA and family have been the action, I think now the consolidation 'reaction' is on the way.
Convenience and integration do not come for free. I hope people will at least take the opportunity and get to know the 'new' programming model.

Michael Thuma |