standtrue wrote:On the one hand we have the existing user/consulting community with ~20 odd years of developing & interfacing in Architect and Perspectives, and on the other we have IBM introducing the very Cognos/EP-ish Performance Modeler & TM1 Applications portal.
As someone mentioned before the new TM1 10 Design & Develop course is all in Performance Modeler and the class (probably new clients) are not seeing Architect at all.
Consultants will want to develop in the "legacy" tools they feel comfortable in,
Hmmm... careful with the implications of that. Someone once suggested that I stayed with an "older" technology only because it was what I was used to. I took serious exception to that because I'd already stated that I stayed with the older technology because it was, in my view, a better and more effective technology for anyone who already knew the tool and didn't need their hand held. (Not that I think that the new tools are all that intuitive anyway, but that's a whole other conversation.) Similarly I don't believe that there are many consultants, certainly not some of the better ones and certainly not some who inhabit these halls, who wouldn't be all over a newer technology regardless of the learning curve
if it provided advantages to both themselves and their clients. Not everybody in this field can't be bothered moving on to something new, they just need to be convinced that there's an up side to doing so.
But therein lies the key problem. All the slick, glossy, virtualised-to-the-cloud, new-paradigm marketing presentations notwithstanding... the new tools just don't offer such an advantage. Steve Vincent nailed the key issue:
Steve Vincent wrote:but the more this meanders to web (which is still too slow and restrictive when compared to an installed client & Excel)
It's become almost an article of faith that "web/cloud is good, desktop is bad" but the fact of the matter is that a well designed, dedicated and regularly updated desktop client, one which is optimised to communicate with its corresponding server, is
always going to outperform (and not just in speed) a corresponding Web application simply because browsers are so generic. Whatever you want to do on your web "canvas" needs to come from the server side. The price of the zero footprint is that you can't build in things on the client side which help handle the load. That's why there's a difference even with something as simple as cracking open subset editor, at least with larger dimensions. In Perspectives I just click the thing and get on with it. In Web I suck in my breath through clenched teeth, click and wait as the elements buffer, and scroll, and wait, and scroll...
However pretty the interface might be (and not very in this case, as I've written scathingly about Sloth Modeller previously), that's the kind of thing that turns end users off.
But that's not the worst aspect of it, not by a long margin.
The worst aspect is the attempt to move away from Excel.
However good the core product is (and it is) it can never be a complete solution. It's basically a numbers machine and the intention is to seek out patterns within those numbers. You can never anticipate in advance where poking and prodding at the numbers will take you and I'll betcha that in almost every site at least
some of the analysis on TM1 data is done in Excel simply because it's the fastest and easiest way of getting from A to B. Do a slice, whack in a formula, copy it down, recalculate, slap on an auto filter... Compare that to trying to drag and drop an endless series of objects around a web canvas. The IBM "wobsession" smacks to me of them (yet again) trying to dictate the direction of the market and keep it all to itself, shutting out any and all competitors. (Even ones which they can benefit from a symbiotic relationship with.) In which case they would mark themselves out as a company which cannot learn from history... even their own.
Ultimately it's the users, not the developers, who will decide whether a product succeeds or fails. And given that the object of the exercise for most people is to do their work in the minimum amount of time so that they can get on with their
real lives, slapping slow-moving, click and drag, click and drag, click and drag web interfaces in their way is more likely to make them ask "Hmm, so what other options do we have for this?" The appeal of eye-candy lasts only so long, even in the i-Everything, reduced attention span age that we now find ourselves in.
standtrue wrote:and newly-trained clients will have to try to maintain or modify models in the new interfaces with predictably interesting results.
So will the existing consulting base grit their teeth and jump into the new tools once they are quasi-stable and stop doing silly things like generating feeders from rates, or will they simply make the rounds fixing the horrible auto-generated models until IBM decides to adjust their product roadmap back to a more manual & technical nature?
Or will they perhaps see an opportunity for bolstering training courses in how to use the product properly, bypassing the slower components of it? It may be a matter of perception but I seem to be seeing an upswing in the courses offered by some consultancies and certainly I'd probably be more inclined to recommend some of those over any "official" courses. But I'd agree that the above two will play a part as well.
standtrue wrote:We KNOW there are people at IBM that want TM1 to be awesome.
The problem is that many of those are ex-Applix and I'm not sure about the extent to which they're thriving (or not) in the IBM biosphere. I also think that there are too many from the IBM and Cognos realms who still don't really comprehend what drew the user base to TM1 in the first place, and we're seeing that in where the development priority is these days.
standtrue wrote:The engine HAS improved under the covers of the past few years. Front end not so much.
The engine hasn't been completely
ignored, notably with the change in locking model post-9.0. And there has been an attempt to implement more parallel work, though I shall leave Mr. Usherwood to comment on the effectiveness of that. I'm not sure that I'd describe these as leaps and bounds improvements, but they're heading in the right direction.
standtrue wrote:Going to be interesting times...
True enough.
