Tomek wrote:
Just chime in a little here. It appears that there are two separate issues here:
Firstly, complex software is complex. It's hard to evolve, maintain, upgrade. Especially so for a system that has endured for as long as ArchiCAD. There is a tonne of software architecture baggage under the hood that can't just be ripped out and replaced. Software development just doesn't work like that. Otherwise we'd have GS rewrite AC every year. Clearly that would not work at all.
Secondly, and more importantly, GS have screwed up managing their customers expectations. They have failed to promise less and deliver more. That is all on management and they need to figure out a way out of this.
I think most people are aware of the difficulties, complexities and hurdles involved in software development in a general sense, and in particular, with a software as complex as ArchiCAD is.
Which is why many of us also questioned the wisdom of their decision way back when to shift to a yearly upgrade cycle that tends to inevitably lends itself to these kinds of issues - i.e.....progressively more diluted versions released each year, with fewer or less impactful features or new tools (if any); more instability with the need for more hotfixes to get it to a stable working and production-ready version; which itself being a function of a short beta-testing and debugging period to iron out the bugs and problems; long-standing unaddressed issues and requests from users which are subject to their almighty roadmap that no one outside the development team or inner circle knows anything about
,...and now delayed releases.
Look, I get it.
A yearly cycle combined with a subscription payment model provides a more stable revenue stream for Graphisoft - hence a large reason for the switch and the decision.
But at what cost? Especially to us, the users, and more so smaller users who frequently tend to get overlooked.
At the end of the day, the die is cast and this is the world we live in, and I don't think they'll be going back to a longer release cycle any time soon (even though circumstances like we're going through now seem to suggest that they would benefit from it almost as much as customers if not more - at least on the developers' side)
As for expectations, if they are finding a mismatch between what they are producing and what customers are expecting, then they only have themselves to blame for that, thanks in large part, as has been pointed out, to their reticence in communicating more frequently and with more clarity to their customers almost to the point, seemingly, of contempt.
If anything I would argue that they actually now benefit from a considerably lowered expectations bar from users thanks to years upon years of not really delivering on a lot of things people hope to get addressed in the program (just visit the Wishlist Section anytime to get an idea). So much so that when they do actually address a long-standing issue (like the Stair tool, for example, or this new version's, "Beam cover fill" long-standing issue) it's treated like they've practically invented the cure for cancer. While all along people forget that these are things that have been sitting in the Wishlist request section for the better part of two decades.
At this stage in the game, them promising less would be just about what most customers expect anyway since we barely hear from them outside their Key Client User conferences, and as for delivering more? Given their track record, I wouldn't hold my breath on that one.
