Djordje wrote:I suppose that whether or not one has a problem with GDL modelling, really depends, in large part, on their familiarity and fluency in GDL scripting in the first place. Which is where the crux of the problem lies for me; in that, traditionally Graphisoft has had the luxury of being able to refer ArchiCAD users to GDL scripting ( or APIs; more on that shortly) to achieve what wasn't doable within and directly from AC's main user interface. At least for a long time ArchiCAD users have had no other option ( outside the GDL route) short of using third party modellers to resolve this issues as other BIM solutions similarly have suffered from the same shortcomings. But the problem is, for how long is this an effective marketing strategy or even operating protocol whereby you make your customers have to resort to acquiring an additional set of skills and expertise ( having to learn GDL and be good at it) to overcome your product's defeciencies. An analogous scenario, would be like Rolls-Royce requiring their clientele to have to learn the intricacies of automative fluid injection adn dynamics details, just to be able to deal with an everyday oil transmission flaw in the engine of their car. ( Yes, I do consider ArchiCAD to be the Rolls-Royce of the Virtual Building field,
Personally, I never saw a problem or had one (and yes, double curves and tents are quite normal stuff around here) using either GDL Toolbox or ArchiForma. Yes, I also knew GDL, but it rusted quite a bit ...
I also don't see a problem in freeform modeling being offered in a form of a cheap API. Except for NURBS, everything else can be easily done with any of the mentioned. It is all about us, in the end.
, despite it's shortcomings, and yes, I do consider the omission of these basic modelling methodologies, such as extrusion along curved paths, to be serious enough flaws if not shortcomings in the program's engine, particularly in this 21st Century, just as an oil transmission defect is in a car engine). The efficacy of this strategy lies in the assumption or rather hope, that other BIM programs ( Revit, Microstation et al) are also not going to develop these same features in their programs any time in the near future, hence leaving AC users no other option. But given Revit's (for example) rather rapid ascension in popularity and usability within its relatively concise period of existence along with its newly acquired marketing and financial (for the time being) muscle in the form of big A(utodesk), how long before they install these features in their next version that even their own users are clamouring for and in essence for how much longer is this 'GDL-or-bust strategy' of GS going to be viable. Smart money says probably until around Revit v7.0!Djordje wrote:As for APIs (yes, 'cheap' is definitely the right description), try to create a curved profile in Archiforma or better yet Profiler, with a radius of 5m or greater (either in elevation or plan it doesn't matter) and let me know how the curve resolution looks like up close, or if (without the help of GDL,) you can adjust said curve to look more smoother than it currently looks. Whereas you say that you don't see the problem, I believe that the main problem lies in the fact that GS is basically lobbing off the responsibilty for redressing these problems onto API programmers who may not be able to do as effective a job or produce as precise a tool as would otherwise be written into the AC engine, such that the troubles I encounter as mentioned above basically become Cigraph's problem and no longer Graphisoft's problem. Once again I believe that this is a seriously and erroneously unsound (if not smug) disposition for Graphisoft to assume regarding the handling of these issues. It's basically the same way they ignored the rendering engine for the longest period of time leaving users (in fact encouraging them, in the case of Abvent ) to seek their rendering solutions elsewhere (Artlantis anyone?) before finally deciding to upgrade it in version 9. Nine versions to upgrade the internal renderer??? How long should we wait for the modelling tools upgrade?
I also don't see a problem in freeform modeling being offered in a form of a cheap API. Except for NURBS, everything else can be easily done with any of the mentioned. It is all about us, in the end.
Of course therere are always work-arounds ( as I gave my own example of the double curving roof in a previous posting) which is one of the great things about ArchiCAD; it's flexibility. But they rarely ever are, 'easily done', as you describe them. ( the Mesh-tool solution in the aforementioned example took up to 10-15 times longer than profiler would have taken had it been working well; and it wasn't even that accurate, when all was said and done)
Djordje wrote:The aversion I was referring to was more so towards stagnation, than it is to change, as you imply. In my personal opinion, there is nothing progressive about GDL scripting, or at least the methodology of it, as a means of creating and modifying 3D models and objects. Sure, it's almost limitless and is quite powerful in the hands of an expert user; as is C++, or Fortran, or COBOL, Turbo Pascal (.... get my drift? ) But in light of the fact that most ArchiCAD users are trained primarily to draw, design and create first and foremostly through graphic visualization as opposed to, say, verbal Coding; wouldn't it make more sense to devise a means with which they can overcome this creatvity barrier in the Virtual Building universe, and which is congenitally attuned to their naturally honed technical and perceptive skills. Is it therefore any surprise that a program such as Sketchup is rapidly gaining in popularity with it's intuitive user interface that is not only easy to learn but more effectively capacitates architects and designers to articulate their ideas in a manner more congruent with the skills they were taught and that they daily use to develop professions and practices? I believe, that for all its power and potential, GDL scripting is nonetheless, simply outdated or is slowly becoming so, as programmers in other fields attempt to abridge the gap between casual user and highly technical and creative tasks. Its the new user-friendly world. In all other aspects ArchiCAD wins in this respect hands down.
Aversion is more like it; as is aversion in general to any change. While the imagination might be soaring, the MEANS of doing your (mine?) work are so primitive; drawings? Please ...
I understand that for those of us well versed in GDL this is really a non-issue ( as would indeed be the case for anyone who is an expert in one field whereas everyone else isn't)and was similarly discussed adn likewise dismissed in one of the Wishlist discussions. But as Virtual Building becomes not only more popular but also more conventional, and with more converts streaming in daily from the Flatlands of CAD - converts, who like us, are architects and correspondingly time-constrained, with lives filled with perpetual project deadlines - expecting these new users to have to learn to program to do what they would otherwise easily accomplish in other programs or more conveniently represent by hand on top of having to learn ArchiCAD itself, is to me a most imprudent marketing strategy on Graphisoft's part which should be ( should have been with v9.0) addressed effectively.
(sorry for going on too long. but I reallly feel quite strongly about this!)
Bricklyne wrote:Bricklyne, your not even close to being alone in your thought despite what the handful of GDL evangelists will try to persuade you to believe.
I suppose that whether or not one has a problem with GDL modeling, really depends, in large part, on their familiarity and fluency in GDL scripting in the first place. Which is where the crux of the problem lies for me;
Bricklyne wrote:Excellent analogy!!!!!!!, but it will probably be falling on deaf ears.
An analogous scenario, would be like Rolls-Royce requiring their clientele to have to learn the intricacies of automative fluid injection adn dynamics details, just to be able to deal with an everyday oil transmission flaw in the engine of their car. ( Yes, I do consider ArchiCAD to be the Rolls-Royce of the Virtual Building field,
Bricklyne wrote:Until Revit or some new kid on the block, like Sketchup hands GS there lunch.
How long should we wait for the modeling tools upgrade?
Bricklyne wrote:You actually would prefer to use your time more productively, what are you thinking?
Of course therere are always work-arounds ( as I gave my own example of the double curving roof in a previous posting) which is one of the great things about ArchiCAD; it's flexibility. But they rarely ever are, 'easily done', as you describe them. ( the Mesh-tool solution in the aforementioned example took up to 10-15 times longer than profiler would have taken had it been working well; and it wasn't even that accurate, when all was said and done)
Bricklyne wrote:Bricklyne, the GDL evangelists will never "get your drift" on this issue. What nerve you have expecting to spend your time creatively designing, presenting, and documenting your architectural projects. How dare you expect to be a "good architect" without being a GDL technician. Don't you know that I.M. Pei, Frank Gehry, Richard Meyer, Norman Foster are all GDL experts? Buck up and learn the montra, I MUST LEARN GDL!!!!!
The aversion I was referring to was more so towards stagnation, than it is to change, as you imply. In my personal opinion, there is nothing progressive about GDL scripting, or at least the methodology of it, as a means of creating and modifying 3D models and objects. Sure, it's almost limitless and is quite powerful in the hands of an expert user; as is C++, or Fortran, or COBOL, Turbo Pascal (.... get my drift? )
Bricklyne wrote:Again with the logic, this one is SMART!
But in light of the fact that most ArchiCAD users are trained primarily to draw, design and create first and foremostly through graphic visualization as opposed to, say, verbal Coding; wouldn't it make more sense to devise a means with which they can overcome this creatvity barrier in the Virtual Building universe, and which is congenitally attuned to their naturally honed technical and perceptive skills.
Bricklyne wrote:
(sorry for going on too long. but I reallly feel quite strongly about this!)
Brycklyne, very well stated. But if you read enough of these posts, will find that some ( though not all ) who will respond as apologists for GS on this forum will try to coerce into believing that expecting the most advanced and forward thinking software, especially given the cost of the software, from GS is YOUR problem. Your points about deficiencies in the modeling capabilities in AC are well reasoned.
It is amazing to me that we still cannot extrude along a path, or lathe a shape, or rotate in 3d without learning GDL or buying many additional tools just to name a couple of modeling shortcomings.
In my opinion, 9.0 should be called 8.5. Some perfume and lipstick was added. Yes you should enjoy some efficiencies with this version. Not sure how some have quantified this as 20%, not yet anyway.
But we still don't see any terrain modeling improvements ( oh, I'm sorry, I keep forgetting the earth is flat, just ask Christopher Columbus, he works at GS ), stairmaker......well we all know what a piece of garbage that is, and adding Lightworks while a good step forward was only a half step. Buy making it part of AC, instead of a free standing package like Artlantis is, ( oh, and what a good idea it was not to add in radiosity since, the apologists have determined for us that we don't REALLY need it, I allways love looking at their renderings ) it forces you to put AC away while a rendering task is being performed.
One of the benefits of Artlantis being a stand alone package is the rendering can process in the background while you continue to work on an AC project in the forground.....uhm, hello GS!!!
OK, sit back and enjoy the comebacks, should be fun. Keep up the heat, its the only way to burn away the chaff!
Regards.
Still looking?
