My first order of business in joining this (or any) organization is to evaluate and review the "raw material" that LRI has to work with -- owned intellectual property as well as existing and/or previous personnel resources. As an executive who has many times been tasked to "figure out and fix what's broke," I've found that all existing organizations come with a mixed bag of useful or obsolete institutional knowledge, products (in this case software) that may or may not be well-designed, properly coded, and able to be evolved instead of tossed and totally rewritten, and personnel who may be anxious to work with the new guy or who resent an outsider's meddling. As usual, in eDoorways and LRI I've discovered challenges and opportunities abounding.
On the software front, most of what I had expected to need rewriting from scratch is more-or-less fairly usable for our near-term development. Note that eDoorways has a relatively long history, having actually begun in the late 90s during that decade's tech boom. Back when the company was worth tens or even hundreds of millions of dollars (quite a contrast to the pittance it's worth today), it heavily invested in the SmartONE learning module building and interpreting code that has languished on its shelves to this day. Similarly, just last year eDoorways launched a fairly complex content management system (CMS) that they called PowerChannels to support user-generated content uploading, managing, and marketing via the eDoorways.com website. My initial reaction to SmartONE was rather negative (to even load it, I had to roll back my Windows 7 system to Windows 98 compatibility mode!). But the more I've talked with the A.I. expert who wrote it, the more I realize that its C++ core engine can easily be "reskinned" to meet today's design standards, and its core A.I. software remains world-class to this day, just as Apple's Siri actually contains many responses that derive from the A.L.I.C.E. open source code whose history itself harkens back to Joseph Weizenbaum's ELIZA written at MIT over 50 years ago. Now, given that core validity, software user interface design is something that I both enjoy and am pretty good at. Moreover, I have accumulated a variety stellar software design personnel resources over my years in this industry. So dragging SmartONE kicking and screaming into the 21st century should be reasonably straight-forward.
eDoorsways.com will be more difficult challenge, but it's actually something I'm planning on attacking first, since it's where the early revenues might lie. The site, presently hosted on a Rackspace server (but needs to be moved to Amazon's EC2 cloud to handle our anticipated traffic), went down this past weekend. Its original outsourced developer was unavailable on short notice, and eDoorways' previous project manager, a Windows server guru, and I were all unable to bring it back up. The original developer finally checked in and waved his magic wand (as software types do), and it was back in 5 minutes. During the fire drill of trying to figure out what had gone wrong, I learned far more about its internal database structure than I had expected in a very short time, and I now understand that there is more there than its (again, very old 90s-ish) look and feel implies. For instance, it's got, but has currently disabled, an interesting twist on social networking already built-in to it, with the ability of mutiple PowerChannel users to interconnect based on keyword searches (e.g., "Are there any other users online now who are interested in today's solar flare activity?"). It has a chat client, again currently disabled, that I expect to replace with something more powerful and current. It has a Paypal-based shopping cart module, again non-functional, that I also expect to replace with something more recent and superior. But overall, its underlying Java-and-MySQL database implementation of PowerChannels for individual users, when suitably renamed and augmented with a designed-and-built-from-scratch graphical user interface (GUI), can be used intact to support our iTunes-like learning app store concept.
Bottom line so far via analogy: Whenever you inherit a broken-down car, it's hard to tell how much work it needs to get it running again. While this one has broken headlights, a ripped headliner, and sagging seats, and has rusty and dented fenders, out-of-favor fins, and an AM radio from way back when, the engine and chassis actually look and run just fine. So, at least at this early stage, I believe that it shouldn't be all that expensive -- in time or money -- to let my designers and developers have their way with it, building from that strong foundation. As promised before, I will keep you appraised of their progress via this blog.
On the software front, most of what I had expected to need rewriting from scratch is more-or-less fairly usable for our near-term development. Note that eDoorways has a relatively long history, having actually begun in the late 90s during that decade's tech boom. Back when the company was worth tens or even hundreds of millions of dollars (quite a contrast to the pittance it's worth today), it heavily invested in the SmartONE learning module building and interpreting code that has languished on its shelves to this day. Similarly, just last year eDoorways launched a fairly complex content management system (CMS) that they called PowerChannels to support user-generated content uploading, managing, and marketing via the eDoorways.com website. My initial reaction to SmartONE was rather negative (to even load it, I had to roll back my Windows 7 system to Windows 98 compatibility mode!). But the more I've talked with the A.I. expert who wrote it, the more I realize that its C++ core engine can easily be "reskinned" to meet today's design standards, and its core A.I. software remains world-class to this day, just as Apple's Siri actually contains many responses that derive from the A.L.I.C.E. open source code whose history itself harkens back to Joseph Weizenbaum's ELIZA written at MIT over 50 years ago. Now, given that core validity, software user interface design is something that I both enjoy and am pretty good at. Moreover, I have accumulated a variety stellar software design personnel resources over my years in this industry. So dragging SmartONE kicking and screaming into the 21st century should be reasonably straight-forward.
eDoorsways.com will be more difficult challenge, but it's actually something I'm planning on attacking first, since it's where the early revenues might lie. The site, presently hosted on a Rackspace server (but needs to be moved to Amazon's EC2 cloud to handle our anticipated traffic), went down this past weekend. Its original outsourced developer was unavailable on short notice, and eDoorways' previous project manager, a Windows server guru, and I were all unable to bring it back up. The original developer finally checked in and waved his magic wand (as software types do), and it was back in 5 minutes. During the fire drill of trying to figure out what had gone wrong, I learned far more about its internal database structure than I had expected in a very short time, and I now understand that there is more there than its (again, very old 90s-ish) look and feel implies. For instance, it's got, but has currently disabled, an interesting twist on social networking already built-in to it, with the ability of mutiple PowerChannel users to interconnect based on keyword searches (e.g., "Are there any other users online now who are interested in today's solar flare activity?"). It has a chat client, again currently disabled, that I expect to replace with something more powerful and current. It has a Paypal-based shopping cart module, again non-functional, that I also expect to replace with something more recent and superior. But overall, its underlying Java-and-MySQL database implementation of PowerChannels for individual users, when suitably renamed and augmented with a designed-and-built-from-scratch graphical user interface (GUI), can be used intact to support our iTunes-like learning app store concept.
Bottom line so far via analogy: Whenever you inherit a broken-down car, it's hard to tell how much work it needs to get it running again. While this one has broken headlights, a ripped headliner, and sagging seats, and has rusty and dented fenders, out-of-favor fins, and an AM radio from way back when, the engine and chassis actually look and run just fine. So, at least at this early stage, I believe that it shouldn't be all that expensive -- in time or money -- to let my designers and developers have their way with it, building from that strong foundation. As promised before, I will keep you appraised of their progress via this blog.
RSS Feed