- Thu Feb 06, 2020 9:58 pm
#306615
Lots of good points you raised and I have thoughts on plenty of them.
Lots and lots of thoughts....
Like you I have been an ArchiCAD user going all the way back to version 8.1 (and the disastrous version 8.0 that it was immediately released to follow and clean-up after), and even briefly with version 7.2 for a little while, so as you can imagine I have a major built-in backlog of issues and things that I feel nee to be improved in the program as well as general gripes with GS themselves.
- Layers and Layer control:-
Couldn't agree more, especially with the whole "nesting" layers, and even some additional things like using layers to over-ride object properties (like lineweights, pen colour, fill colour and type,etc) or to change opacity on the fly using a slider (think of how useful this would be in fading out funiture layers, for example, when sending out drawings to consultants or third-party partners, or for fading or reducing the opacity of mechanical layer objects and plumbing fixtures, etc.
You would think that the fact that they've been working in conjunction with the folks at McNeel over the Rhino/Grasshopper bridge would have given them some ideas to improve their own layer control set-up and functionality.
As is, it's beyond even just 1990's-level of backwardness. Even AutoCAD in the 1990's offered more control and versatility through layers than ArchiCAD does NOW.
That's embarrassing.
-GDL programming and custom object creation:
Double double agree.
With extra cheese and toppings.
Once again the fact that we speak of Revit's FAmily editor as being a vastly superior way of creating custom parametric objects for Revit (which has ultimately given them the edge in terms of the availability of Revit third-party fully-parametric library objecta available online and from suppliers) in comparison to ArchiCADS literally archaic line-coding to create the same, is agonizingly embarrassing.
Line-coding is wuite literally how we define the 1980's guys. I mean, come on.
Anytime you see one of these nostalgia '80's movies or TV shows where they show someone using a computer or doing something "Advanced" (for back them) it's almost always typing out line codes in C++ or BASIC or some such thing in a clunky console or terminal.
That's were ArchiCAD and the GRaphisfot folks STILL are in relation to how they expect their users to create some of the most complex objects in this, their advanced software and tool.
Shameful.
-BIMCloud
- Can't really chime in much on the whole BIMCloud angle since from the start I always felt that Graphisoft were missing the entire point (and potential) in the whole wave to move software and digital work to the 'Cloud' or to use the cloud to advance their users' workflow, and it would seem I was right.
See also : BIMComponents (remeber that? The GRaphisoft attempt at launching a Google (now Trimble) Sketchup-like Google Warehouse repository of online objects availble to user to share and add to? And how it went nowhere? In light of the limitations of being able to create custom versatile parametric GDL objects, something about putting the cart before the horse comes to mind, right about now).
-On Teamwork
-One (somewhat adjacent) thought about Teamwork is that I recall when they introduced 'Teamwork 2.0' a couple of versions back and a few years not too long ago, and one of the big selling points in its overhaul was the notion that when you work in Teamwork with physically remotely located teams sharing the same teamwork project file online, that the program now would only send changes as opposed to the entire database, which would ultimately increase the speed and efficiency with which teams could collaborate online.
Which seemed very smart if not the logically inevitable practical way it should have worked all along.
But then I thought to myself : "Well, why the hell doesn't the entire parametric change engine of the program itself work this way at a local level in one's own workstation?"
As is, whenever I make even a simple change and switch views (especially to the Sections or Elevations) you can easily fnd yourself waiting several minutes (or more) -especially for bigger and more complex projects - waiting for the program to regenerate the view and drawing just to update JUST THE SINGLE WALL you moved when nothing else changed. It is especially egregious and noticeable with Sections and Elevation windows but also with 3D views as well (the raw 3D windows and really really badly with 3D Documents).
The program seems to use an archaic parametric change engine and tracker which, rather than intelligently just keep track of, and update only just the changes, has to certify the entire model or drawing with every single view switch or change, which, from a workflow perspective is just torture.
I guess what I'm trying to say is that the ArchiCAD change engine and it's own kernel badly need updating - particularly to take advantage of the more robust (and more readily available) hardware (and software technology available today.
Gone are the days when they could sing from the rooftops that they are first (and only) in market with a BIM product that was fully multi-core aware,
To wit: Graphics cards and GPU's nowadays have literally hundreds of cores, and yet we're still stuck on a single monitor use workflow (on the PC side at least).
That picture just doesn't feel right (on top of being constrained........to one monitor)
-Visuals
I'm guessing you mean visualization rather than the look of the program and in this regard, I would be careful what you wish for in terms of Grephisoft getting into bed with third party partners to produce tools to substitute for (rather than supplement) what they themselves have failed to produce natively in the program or at least are doing a poor job of maintaining and developing themselves.
I'll give you a couple of examples.
The first one involves a trip down memory lane.
Being a long-term user (or as long as myself), I'm sure you would be familiar with the term,...."Maxonform"?
For those not aware, Maxonform was a "bridge" that Graphisoft developed way back in the day (I believe before Nemetschek bought both them and Maxon out) in conjunction with Maxon and their Cinema4D program as a means of allowing ArchiCAD users to access the power of modern 3D modelling tools found in 3D modeling programs like 3ds Max and Cinema4D - but from within the ArchiCAD user environment (......sort of....).
At least that was the idea...
Except it didn't really work out that way (in terms of being very clunky, both the program/plugin itself as well as the workflow and also the general coordination between the two companies). It was a great idea.....or at least it seemed that way at the time,.....but ultimately poorly executed and in hindsight questionable on many levels and ultimately they gave up on the effort not long afterwards, right before devoting their own resources to developing a internal and (truly) native powerful modelling tool for the program on their own which we now know as the Morph tool. But the path of getting to the morph tool (which itself, while a great improvement to ArchiCAD's toolset, could still use some improvements and a Morph Version 2.0 upgrade of its own integrating SubD polymodeling and NURBS translation) was truly tortured one and littered with lots of bad decisions along the way.
Graphisoft has had a little but more luck working with third-party partners in developing external tools for users in recent years and versions (the Grasshopper/Rhino bridge comes to mind. And also while not STRICTLY a pure third-party tool on its own, the Cinerender engine also falls in the same vein of not developed purely by Graphisoft themselves), but they still face the same issues that this kind of development strategy tend to elicit (like not being in total control of the schedule and timetable during which their partners release their upgrades in keeping with most recent ArchiCAD version releases. For example, the Cinerender Engine inthe current ArchiCAD versionis always an entire version behind the actual current Cinema4D internal render engine, and even in the case of Grasshopper/Rhino - of which, McNeel have a more forgiving release schedule whence they release a new version every 5 or so years - you still have gaps in when you can get the most recent version of the plugin working with the most recent version of Rhino or ArchiCAD respectively.
And you can see from the current delayed release debacle and fiasco of ArchiCAD 23 on the one hand and also the expected release of the new version of Twinmotion (which was itself also delayed from late last year to,....."sometime in early 2020" ...or "first quarter of 2020", depending on which press release you're reading. I understand that they're having issues integrating the Real-time raytracing RTX engine that Graphisoft have already gone out of their way to promote in their own ArchiCAD version 23 promotional videos. Go figure. That's what holding yourself to someone else's promises will get you), this kind of relationship becomes rife with potential problems and risks.
What happens if Twinmotion gets delayed until sometime in early summer, for example? It will mean that ArchiCAD version 23 users who were promised a free version of the lates Twinmotion with their version 23 purchase will only be getting a couple of months worth (at best) use of free Twinmotion before it becomes a paid product again (and this also assumes that GRaphisfot themselves are back on schedule this year with their own releaase cycle.
Graphisoft themselves have no control over what happens over at EpicGames in terms of the development of Twinmotion or the release schedule and this is what you tend to get.
And when you've already promoted YOUR own program using THEIR product (which you have not control or say over), then you have issues.
What also happens if the Twinmotion new version is released (even in the next couple of weeks) and it turns out it's buggy as hell as to be practically unusable?
So I don't really know about using this as a viable strategy going forward - even while I can understand the necessity to have to outsource some of these development functions to outside parties or specialized interest groups with a view to keeping a tight grip on your own limited and strained resources (.....though,....again, they only have themselves to blame here if they feel stretched thin because they're the ones who decided to go on a yearly release cycle that seems to have been nothing but a bane to the existence of both users and developers (on their end) alike. Stop letting the bean-counters and accounting/numbers guys run your company and make strategy decisions for you,....That would be a good start)).
Like I said, ....I have thoughts.
And these aren't even the half of them.
So I'll distill my simple advice and suggestion to Graphisoft to one simple thing:-
You want to increase your marketshare?
Listen to your users and improve your interaction and overall relation with them (beyond those largely useless (IMHO) 'Key Client User' meet-greets and hobnobs that only give you a blinkered perspective and not a well-rounded picture or the necessary tough-love feedback you so often clearly and badly need).
Don't under-estimate how much free-marketing you often can get from happy and satisfied users who feel their needs are being met and their money is well-spent and justified.
Ignore them at your own overall peril.
gavinNZz wrote: ↑Thu Feb 06, 2020 3:18 am I see the main problem being that Archicad is a piece of 1990's software trying to wear some 2010's clothing but still dances like it's in the 90's. I have been with Archicad since V8 (15+years) and there are some functions that have not been touched in that time. How is that even possible??.......
Lots of good points you raised and I have thoughts on plenty of them.
Lots and lots of thoughts....
Like you I have been an ArchiCAD user going all the way back to version 8.1 (and the disastrous version 8.0 that it was immediately released to follow and clean-up after), and even briefly with version 7.2 for a little while, so as you can imagine I have a major built-in backlog of issues and things that I feel nee to be improved in the program as well as general gripes with GS themselves.
- Layers and Layer control:-
Couldn't agree more, especially with the whole "nesting" layers, and even some additional things like using layers to over-ride object properties (like lineweights, pen colour, fill colour and type,etc) or to change opacity on the fly using a slider (think of how useful this would be in fading out funiture layers, for example, when sending out drawings to consultants or third-party partners, or for fading or reducing the opacity of mechanical layer objects and plumbing fixtures, etc.
You would think that the fact that they've been working in conjunction with the folks at McNeel over the Rhino/Grasshopper bridge would have given them some ideas to improve their own layer control set-up and functionality.
As is, it's beyond even just 1990's-level of backwardness. Even AutoCAD in the 1990's offered more control and versatility through layers than ArchiCAD does NOW.
That's embarrassing.
-GDL programming and custom object creation:
Double double agree.
With extra cheese and toppings.
Once again the fact that we speak of Revit's FAmily editor as being a vastly superior way of creating custom parametric objects for Revit (which has ultimately given them the edge in terms of the availability of Revit third-party fully-parametric library objecta available online and from suppliers) in comparison to ArchiCADS literally archaic line-coding to create the same, is agonizingly embarrassing.
Line-coding is wuite literally how we define the 1980's guys. I mean, come on.
Anytime you see one of these nostalgia '80's movies or TV shows where they show someone using a computer or doing something "Advanced" (for back them) it's almost always typing out line codes in C++ or BASIC or some such thing in a clunky console or terminal.
That's were ArchiCAD and the GRaphisfot folks STILL are in relation to how they expect their users to create some of the most complex objects in this, their advanced software and tool.
Shameful.
-BIMCloud
- Can't really chime in much on the whole BIMCloud angle since from the start I always felt that Graphisoft were missing the entire point (and potential) in the whole wave to move software and digital work to the 'Cloud' or to use the cloud to advance their users' workflow, and it would seem I was right.
See also : BIMComponents (remeber that? The GRaphisoft attempt at launching a Google (now Trimble) Sketchup-like Google Warehouse repository of online objects availble to user to share and add to? And how it went nowhere? In light of the limitations of being able to create custom versatile parametric GDL objects, something about putting the cart before the horse comes to mind, right about now).
-On Teamwork
-One (somewhat adjacent) thought about Teamwork is that I recall when they introduced 'Teamwork 2.0' a couple of versions back and a few years not too long ago, and one of the big selling points in its overhaul was the notion that when you work in Teamwork with physically remotely located teams sharing the same teamwork project file online, that the program now would only send changes as opposed to the entire database, which would ultimately increase the speed and efficiency with which teams could collaborate online.
Which seemed very smart if not the logically inevitable practical way it should have worked all along.
But then I thought to myself : "Well, why the hell doesn't the entire parametric change engine of the program itself work this way at a local level in one's own workstation?"
As is, whenever I make even a simple change and switch views (especially to the Sections or Elevations) you can easily fnd yourself waiting several minutes (or more) -especially for bigger and more complex projects - waiting for the program to regenerate the view and drawing just to update JUST THE SINGLE WALL you moved when nothing else changed. It is especially egregious and noticeable with Sections and Elevation windows but also with 3D views as well (the raw 3D windows and really really badly with 3D Documents).
The program seems to use an archaic parametric change engine and tracker which, rather than intelligently just keep track of, and update only just the changes, has to certify the entire model or drawing with every single view switch or change, which, from a workflow perspective is just torture.
I guess what I'm trying to say is that the ArchiCAD change engine and it's own kernel badly need updating - particularly to take advantage of the more robust (and more readily available) hardware (and software technology available today.
Gone are the days when they could sing from the rooftops that they are first (and only) in market with a BIM product that was fully multi-core aware,
To wit: Graphics cards and GPU's nowadays have literally hundreds of cores, and yet we're still stuck on a single monitor use workflow (on the PC side at least).
That picture just doesn't feel right (on top of being constrained........to one monitor)
-Visuals
I'm guessing you mean visualization rather than the look of the program and in this regard, I would be careful what you wish for in terms of Grephisoft getting into bed with third party partners to produce tools to substitute for (rather than supplement) what they themselves have failed to produce natively in the program or at least are doing a poor job of maintaining and developing themselves.
I'll give you a couple of examples.
The first one involves a trip down memory lane.
Being a long-term user (or as long as myself), I'm sure you would be familiar with the term,...."Maxonform"?
For those not aware, Maxonform was a "bridge" that Graphisoft developed way back in the day (I believe before Nemetschek bought both them and Maxon out) in conjunction with Maxon and their Cinema4D program as a means of allowing ArchiCAD users to access the power of modern 3D modelling tools found in 3D modeling programs like 3ds Max and Cinema4D - but from within the ArchiCAD user environment (......sort of....).
At least that was the idea...
Except it didn't really work out that way (in terms of being very clunky, both the program/plugin itself as well as the workflow and also the general coordination between the two companies). It was a great idea.....or at least it seemed that way at the time,.....but ultimately poorly executed and in hindsight questionable on many levels and ultimately they gave up on the effort not long afterwards, right before devoting their own resources to developing a internal and (truly) native powerful modelling tool for the program on their own which we now know as the Morph tool. But the path of getting to the morph tool (which itself, while a great improvement to ArchiCAD's toolset, could still use some improvements and a Morph Version 2.0 upgrade of its own integrating SubD polymodeling and NURBS translation) was truly tortured one and littered with lots of bad decisions along the way.
Graphisoft has had a little but more luck working with third-party partners in developing external tools for users in recent years and versions (the Grasshopper/Rhino bridge comes to mind. And also while not STRICTLY a pure third-party tool on its own, the Cinerender engine also falls in the same vein of not developed purely by Graphisoft themselves), but they still face the same issues that this kind of development strategy tend to elicit (like not being in total control of the schedule and timetable during which their partners release their upgrades in keeping with most recent ArchiCAD version releases. For example, the Cinerender Engine inthe current ArchiCAD versionis always an entire version behind the actual current Cinema4D internal render engine, and even in the case of Grasshopper/Rhino - of which, McNeel have a more forgiving release schedule whence they release a new version every 5 or so years - you still have gaps in when you can get the most recent version of the plugin working with the most recent version of Rhino or ArchiCAD respectively.
And you can see from the current delayed release debacle and fiasco of ArchiCAD 23 on the one hand and also the expected release of the new version of Twinmotion (which was itself also delayed from late last year to,....."sometime in early 2020" ...or "first quarter of 2020", depending on which press release you're reading. I understand that they're having issues integrating the Real-time raytracing RTX engine that Graphisoft have already gone out of their way to promote in their own ArchiCAD version 23 promotional videos. Go figure. That's what holding yourself to someone else's promises will get you), this kind of relationship becomes rife with potential problems and risks.
What happens if Twinmotion gets delayed until sometime in early summer, for example? It will mean that ArchiCAD version 23 users who were promised a free version of the lates Twinmotion with their version 23 purchase will only be getting a couple of months worth (at best) use of free Twinmotion before it becomes a paid product again (and this also assumes that GRaphisfot themselves are back on schedule this year with their own releaase cycle.
Graphisoft themselves have no control over what happens over at EpicGames in terms of the development of Twinmotion or the release schedule and this is what you tend to get.
And when you've already promoted YOUR own program using THEIR product (which you have not control or say over), then you have issues.
What also happens if the Twinmotion new version is released (even in the next couple of weeks) and it turns out it's buggy as hell as to be practically unusable?
So I don't really know about using this as a viable strategy going forward - even while I can understand the necessity to have to outsource some of these development functions to outside parties or specialized interest groups with a view to keeping a tight grip on your own limited and strained resources (.....though,....again, they only have themselves to blame here if they feel stretched thin because they're the ones who decided to go on a yearly release cycle that seems to have been nothing but a bane to the existence of both users and developers (on their end) alike. Stop letting the bean-counters and accounting/numbers guys run your company and make strategy decisions for you,....That would be a good start)).
Like I said, ....I have thoughts.
And these aren't even the half of them.
So I'll distill my simple advice and suggestion to Graphisoft to one simple thing:-
You want to increase your marketshare?
Listen to your users and improve your interaction and overall relation with them (beyond those largely useless (IMHO) 'Key Client User' meet-greets and hobnobs that only give you a blinkered perspective and not a well-rounded picture or the necessary tough-love feedback you so often clearly and badly need).
Don't under-estimate how much free-marketing you often can get from happy and satisfied users who feel their needs are being met and their money is well-spent and justified.
Ignore them at your own overall peril.