The list of things ESRI products can do that Manifold cannot do would be immense. This isn't to knock on Manifold, it's just a fact. A fact, largely due to ESRI's 30 year head start, immense budget, and more than 4,000 employees. Hopefully, the next few years will significantly reduce the gap between the two companies. Well, that could just as accurately be rewritten to: The list of things Manifold products can do that ESRI cannot do would be immense. This isn't to knock on ESRI, it's just a fact. A fact, largely due to Manifold's 10 year head start in technology, immense development budget, and more employees doing actual product development. Hopefully, the next few years will significantly increase the gap between the two companies as Manifold pulls further ahead. Both of the above can be true at the same time because both ESRI and Manifold product families are truly immense in breadth and depth, taking several thousand pages in either case to document. It is expected with any such broad and deep and diverse product families that both will do what most people in general-purpose professional GIS want to accomplish, but that on the way to accomplishing that general end there will be many things that ESRI does which Manifold does not and also that there will be many things that Manifold does that ESRI does not. Both companies have far too many features to expect an exact overlap, and given the large number of features any non-intersecting product regions on both sides will include large numbers. However, for all that, the general opinion of those who know both ESRI and Manifold very well seems fairly strong that at any given cut through the product families the Manifold option will in general provide greater breadth and depth of capabilities at a far lower price. Almost without exception, when ESRI people learn Manifold they will often comment that Manifold does an absolutely astonishing number of things and does many of them better than ESRI. How can that be? The quotation at the top of this letter provides some clues: 30 years, big budget and herds of employees are not positives in this case. Starting 30 years earlier when the competition is all about rapidly changing technology can often be a negative, not a positive. In the audio player industry, for example, tinkering around with reel-to-reel tape recorders or cassette tapes 30 years ago just builds a lot of legacy inertia that does not help you out-iPod Apple in modern technology. If you look at ESRI's technology, that's through and through the case with them where they cling to horrifically obsolete technology. Good example are the paleo-GIS relics of shapefiles within today's SDE, SDE in general, a non-integrated architecture, inability to utilize modern hardware and ESRI's return to time-sharing instead of distributed desktop processing. Time-sharing became obsolete almost 30 years to the day when in 1979 microprocessors started their epic campaign to extinguish time-sharing. I guess ESRI didn't get the memo. They struggle with something as elementary as going 64-bit. They don't do modern things like CUDA at all. I'll take the next two points in reverse order: 4,000 employees is a harsh negative if 3,980 of those employees are devoted to activities other than core development on your core product. ESRI's actual product development team is tiny because they put most of their efforts into other activities, such as very large marketing and sales forces devoted to spending immense amounts of money selling tiny numbers of hyper-expensive products. Much of what they do put into "engineering" goes not into forward development of their core products but rather into applying twists of baling wire and duct tape to patch up living fossil technology. As for very large budgets, if the amount of money one burned up was a metric of quality, we'd all be rushing to use the Post Office instead of Federal Express and the expression "five hundred dollar hammers and two thousand dollar toilet seats" would not have entered the public lexicon. I have run development efforts that had one hundred million dollar plus budgets backed by multi-billion dollar investment into collateral infrastructure. I can assure you that the Great Rule of government financing applies to such efforts as well: "All the money that is budgeted will get spent." But there is no Great Rule that says it will get spent effectively or wisely. In fact, the usual case is that huge amounts of excess cash will be spent foolishly and with poor efficiency, whether you are talking about government or private industry. In both cases, absolute budget power corrupts absolutely and usually leads to nothing but Caligulan bacchanals at the expense of either taxpayers or shareholders. I'm not throwing stones here, by the way, because I've done that too. It's amazing how easy it is when you have a hundred million or so in the budget to convince yourself that dialing back a tradeshow display from costing $700,000 to a mere $600,000 is an example of exemplary fiscal restraint. In point of fact, if you know what you are doing and you are working with an elite team on well-focussed objectives, a few tens of millions is plenty to do whatever you want. Apply more money than that and it just gets wasted on stupid stuff like too many private espresso bars, pinball machines and other stupid toys that people get into when they don't really know what they are doing well enough to become obsessed by the real work of the enterprise. So how does Manifold do much more than ESRI at a fraction of the cost? Easy. The use of more modern technology in Manifold products allows Manifold to deliver much greater reliability, more power, greater sophistication and accuracy of algorithms and greater flexibility all at a much lower price than ESRI products. That's how a $445 Manifold Enterprise x64 license does much more for most real-life users than a basket of ESRI products costing $30,000, such as ArcGIS, ArcGIS Server, ArcSDE, ArcIMS, ArcMap, ArcInfo, ArcObjects and the like, or whatever they are calling those products these days. Does any of that mean that there aren't things ESRI does which Manifold does not, or that it's not highly instructive to learn from ESRI if you are setting out to do a better job? Of course not. The key there, though, is to focus on the general tasks and activities in government and industry and science which although evolving are more or less continuing over the past few decades and into the future. ESRI has a lot of features accreted in their products to serve such activities and the ones that still make sense are the ones to support with newer, more performant, more reliable, more flexible, more cost-effective technology. An addendum: folks who have run successful development teams will rightly point to the tiny size of ESRI's core development teams as a plus. That's true, and if you really know what you are doing you know that given the limits of organization you can't make really good development go ten times faster by throwing 1000 guys at a task instead of 100. As Microsoft has demonstrated with Windows, even if you are really good at development management that doesn't work. Many of the truly outstanding technologies of our era have been created by very small teams. UNIX, for example, was just a couple of guys at first that grew into a core team of about five or ten guys. Add too many heads and you run into the limits of organization, kind of like physical limits brought the end of Moore's law in random-logic chips. But you can get past that a bit with careful modularization, sort of applying parallelism in development, but only if the effort is very well conceived and managed overall. And that's where the 3,980 other guys at ESRI spoil the broth. When the company has headcount overwhelmingly in areas other than development, that's where all the mindshare goes. If you have 2000 guys in marketing and 20 guys in core development when the core development guys want to bite off the pain of doing some new development right but that conflicts with the near-term objectives of the 2000 guys in marketing, well, guess who wins the vote? That's why you see ESRI guided by the inertia and demands of just about everything except better product at a lower cost. This can be changed, of course, in those companies where the non-development people have the longer time scales and greater vision to put a better product at a lower cost first. But given human nature that's very hard to achieve in the scrum of day to day business activities where very short term objectives often rule.
|