GDI+ Alternative - C++ Gauge

[Guest fturner]

>>In my opinion, Agg is ready to be utilized to replace GDI+.>>>The pro's far outweigh the con's just in alpha blending,>>gradient, sub-pixel accuracy, anti-aliasing and performance>>alone, and I've only scratched the surface of what Agg can>do.>> All the con's I've come across I've easily worked around>them>>by using other tools offered in Agg. Ed, you have pointed>out>>con's (either right or wrong) and I have been forced to be>the>>opposite.>>I must commend you on the tone of this response.>>I'd also like to mention that I particularly like the way in>which you provide copious comments in your example code, as>well as leaving in the original GDI+ code to illustrate how>the Agg code differs... Nice touch!>>I'd like to extend an invitation to begin a series of>'mini-tutorials' at http://flightsim.com in my Panel and>Gauges forum. I began that forum almost a year ago simply>because the forum here seems heavily weighted towards XML>coding (which is perfectly fine!), and wanted to develop a>forum that is primarly focused on "C gauges"... :)Hi Bill,Thanks for the invite and I will start doing such. But, if any discussions over there turn out to be like this one I'll walk out, and I hope you can understand that. I would like to see discussions that are more positive along the lines of, "ok here's a con, now how can we work around this to achieve what we want" instead of "here's a con, so its no good" type posts as seen here.I did not mean things to turn out the way they did here, as all my intent was to introduce something new and exciting, not into a dueling match as you have described.I look forward to working positivly with the community on this exciting endeavour.Fraser

[WarpD 824]

I believe I asked you to show the functions that directly support poly and path clipping?And using an alpha blend to mimic clipping isn't clipping. It's a workaround... but it is not clipping.Also... how do you encapsulate the behavior of containers in GDI+ using AGG? Without containers, coding some rather complex gauge behavior will require tons more coding.See... I'm asking questions. I believe that's where you offer information regarding exactly how AGG can accomplish it? It's neither a pro nor a con approach at all.I'll also point out your example code is not available, as in... we need it available for people to get... doncha agree?

[Guest fturner]

>I believe I asked you to show the functions that directly>support poly and path clipping?>>And using an alpha blend to mimic clipping isn't clipping. >It's a workaround... but it is not clipping.>>Also... how do you encapsulate the behavior of containers in>GDI+ using AGG? Without containers, coding some rather>complex gauge behavior will require tons more coding.>>See... I'm asking questions. I believe that's where you offer>information regarding exactly how AGG can accomplish it? It's>neither a pro nor a con approach at all.>>I'll also point out your example code is not available, as>in... we need it available for people to get... doncha agree?I have a suggestion... why don't you offer us your answer's to the above questions? Also I invite you to offer some possible work arounds as well if in fact Agg doesn't support what you have suggested above.Frit

[n4gix 4,947]

>I did not mean things to turn out the way they did here, as>all my intent was to introduce something new and exciting, not>into a dueling match as you have described.>>I look forward to working positivly with the community on this>exciting endeavour.What I'd like to see occur is simple, even moreso since I have the ultimate power and authority with which to accomplish said goal in that forum...I'd like to see one specific "feature set" of higher level functionality discussed, developed, bug tested and tweaked......then using my "magic powers" I can then condense the thread into a more polished format and begin building an "Agg API" thread with each developed function listed, along with any relevant ancillary notes of explanation.Does this sound feasible or even desirable?One thing I can promise everyone though is this: I will ruthlessly enforce the standards of decency and good manners by either editing or outright deleting anything that has even a whiff of posturing, chest thumping or ad hominem...

[Guest JeanLuc_]

"you have managed to put Agg down and convinced alot of people that it cannot be used. "Hum,,, Hum,,,I'm sorry, but at least for me, Ed has proven only one thing: AAG is not fulfilling HIS requirements. That's all.Knowing Drawing APIs to some (relative term) extent, I perfectly understand your point of view, and I salute and welcome your innitiative to come in this forum offering alternatives for what I think (personal optinion) is the better for the development of better prodcuts.Now, we all have understood that Ed does not want to work with AAG for many reasons he has exposed. That they are true or false is irrelevant. What is relevant is that more people (silent majority?) are willing to go further, and at this time, I guess this thread should continue as being "hey, look and study with me this alternative solution for this and that reasons" instead of "this is lacking this and that and I won't use it because it does not fit the level of function I require from my development experience".Please, do not go elsewhere than this forum (selfish remark because I don't go elsewhere) and don't be distracted.What is important is not the lack of features in my opinion. What is important is how to use the existing to achieve a goal. In this respect, AAG as as much to offer than GDI+ to achieve the goal of designing/drawing complex rendered gauges. Only the path to achieve it is different than GDI+. Noboby is forced to take the same path, and at the same time, nobody can say only ONE path leads to the goal.Hope this helps!

[Guest JeanLuc_]

I won't speak publicly about it and I even consider I've given enough clues in letting people know that there is no "hack" to do it and it is directly documented in the gauge.h.This reminds me a suggestion in another place, about transparency for gauges, for which the clear issue is not the lack of feature in the SDK, but it is the drawing API that is used that prevents accessing the feature. In this case, the API that limits the access to this potential feature of the SDK, is GDI(+)...

[n4gix 4,947]

>Please, do not go elsewhere than this forum (selfish remark>because I don't go elsewhere) and don't be distracted.The only reason I suggested the other forum is that unlike here, I have to ability there to 'pin' threads easily, and maintain a 'condensed version' of the resulting discussion into a single, coherent topic so that it would then become an indexed resource for 'solutions...'This thread already demonstrates the impracticality of using it as a 'searchable resource,' since the little wheat is buried by the chaff... ;)

[Guest fturner]

>>I did not mean things to turn out the way they did here, as>>all my intent was to introduce something new and exciting,>not>>into a dueling match as you have described.>>>>I look forward to working positivly with the community on>this>>exciting endeavour.>>What I'd like to see occur is simple, even moreso since I have>the ultimate power and authority with which to accomplish said>goal in that forum...>>I'd like to see one specific "feature set" of higher>level functionality discussed, developed, bug tested and>tweaked...>>...then using my "magic powers" I can then condense the thread>into a more polished format and begin building an "Agg API">thread with each developed function listed, along with any>relevant ancillary notes of explanation.>>Does this sound feasible or even desirable?>>One thing I can promise everyone though is this: I will>ruthlessly enforce the standards of decency and good manners>by either editing or outright deleting anything that has even>a whiff of posturing, chest thumping or ad hominem...I concur whole heartedly.... but in regards to the other forum, I do not agree with using aggplus. The primary reason for looking for an alternative was to maximize the performance on drawing gauges, especially vector gauges. Using Aggplus, while it having the plus of being "familiar" for people using GDI(+) is another layer of the drawing framework that will slow down performance.When I speak of performance, I'm not just talking about how fast an api can draw something, but the overall impact within FS. The biggest consumption of processor power goes to drawing the 3d scenery around the aircraft. The less time it takes to "draw" something on the screen in the panel, means more time to drawing the 3d environment, so even if we can improve on less time panel processing that is not seeable by the user, will in fact be seeable by the user in the 3d view.So looking at it this way, adding a layer to agg like aggplus will only cause to slow it down. So while its easier for the developer to build a gauge, it will cause a slow down on the user side that could become detectable.Enough said I guess.Frit

[WarpD 824]

>I concur whole heartedly.... but in regards to the other>forum, I do not agree with using aggplus. The primary reason>for looking for an alternative was to maximize the performance>on drawing gauges, especially vector gauges. Using Aggplus,>while it having the plus of being "familiar" for people using>GDI(+) is another layer of the drawing framework that will>slow down performance.>>When I speak of performance, I'm not just talking about how>fast an api can draw something, but the overall impact within>FS. The biggest consumption of processor power goes to>drawing the 3d scenery around the aircraft. The less time it>takes to "draw" something on the screen in the panel, means>more time to drawing the 3d environment, so even if we can>improve on less time panel processing that is not seeable by>the user, will in fact be seeable by the user in the 3d view.>>So looking at it this way, adding a layer to agg like aggplus>will only cause to slow it down. So while its easier for the>developer to build a gauge, it will cause a slow down on the>user side that could become detectable.>>Enough said I guess.>>Frit>>>I have seen no slow down in AggPlus. Simply put, prove your claim.

[n4gix 4,947]

>This thread already demonstrates the impracticality of using>it as a 'searchable resource,' since the little wheat is>buried by the chaff... ;)Never mind!As the new thread had already (within less than 24 hours) already begun to evidence the same trend as this one here has, I've now deleted the entire thread and banned the topic completely there.Perhaps in six months or so I'll revisit this topic there...

[martinboehme 492]

> I won't speak publicly about it and I even consider I've given enough> clues in letting people know that there is no "hack" to do it and it> is directly documented in the gauge.h.Roger.;-)Martin

[Guest DanWalloch]

So now that a few people from eaglesoft have stumbled apon Fraser's kind offering of a dummy's guide to AGG, I sure would like to see some quality work from ESDG and bye bye to that fuzzy #### of FS 5.1 guages they seems to use that kill anyones PC. But what do I know, just that fraser has had this AGG stuff under his belt for the last few months and I know what it will do for gauges if properly utilized.

[n4gix 4,947]

>So now that a few people from eaglesoft have stumbled apon>Fraser's kind offering of a dummy's guide to AGG, I sure would>like to see some quality work from ESDG and bye bye to that>fuzzy #### of FS 5.1 guages they seems to use that kill>anyones PC. But what do I know, just that fraser has had this>AGG stuff under his belt for the last few months and I know>what it will do for gauges if properly utilized.Can't resist a dig, eh? For your information, 'we' have known about AGG for well over a year already, and still find it lacking in the ability to produce the level of quality we're able to develop in GDI+You're right though, we should be ashamed to produce gauges like these, that're obviously inferior... not!http://img177.imageshack.us/img177/5736/av...ec400pfdrb3.pnghttp://img114.imageshack.us/img114/4808/av...ec400mfdwf1.jpg

[Guest Patrick_Waugh]

drool... drool... still drooling!

[WarpD 824]

>So now that a few people from eaglesoft have stumbled apon>Fraser's kind offering of a dummy's guide to AGG, I sure would>like to see some quality work from ESDG and bye bye to that>fuzzy #### of FS 5.1 guages they seems to use that kill>anyones PC. But what do I know, just that fraser has had this>AGG stuff under his belt for the last few months and I know>what it will do for gauges if properly utilized.First, you will notice there is no Eaglesoft in my signature, I am not here for Eaglesoft. Anything I say or do in these forums is as myself. Any effort I have with FS, gauges and code development in the public area of Avsim.com and Flightsim.com is for the public as much as myself.Second, I've had AGG on my system quite a bit longer than 'a few months'. You give Fraser a bit too much credit in that regard, he didn't introduce me to a thing.