i wish to note that i have no direct association with Paul Graham. i only know of his work through discussion and observation of his code with Yahoo! Shopping (and the after effects of it), so i harbor little opinion about him. i read his recent post about how Yahoo! failed with interest, and for what it's worth, it does contain a number of things that are true. Mostly, these relate to the issues around the "easy money" that was had and the poor decisions that resulted.
i'll note that he also touched on the "media centric" issue. It's very to say that was a mistake because the past decade has shown very clearly the error, but at the time, it would be impossible to do so. Google was interested in only being a search engine and filling a niche that was poorly supported. i'll note that they didn't turn profitable until some time later when they discovered they could sell ads out of it, but i digress.
i do draw some issue with the latter half of his post, where he basically becomes highly dismissive of the engineering resources that Yahoo! had. To that end, i wish to remind the world of a few things. One of which is that he wrote the Y!Shopping cart in eLisp, which he proudly notes:
We wrote our software in a weird AI language, with a bizarre syntax full of parentheses. For years it had annoyed me to hear Lisp described that way. But now it worked to our advantage. In business, there is nothing more valuable than a technical advantage your competitors don't understand. In business, as in war, surprise is worth as much as force.
No, that's not a business advantage, that's job security. A proper business has these nifty protocols and procedures to generally keep folks from viewing fiscally sensitive code like shopping carts. Yahoo! had those. He wrote in Lisp because it was his pet language. Unfortunately, not that many developers at the time knew lisp, nor do many know it now. This means that while he was quite proficient in it, it was very difficult to fill his role once he left, a little over eleven years ago.
i won't say that the replacement code was more elegant or efficient than what Mr. Graham designed, but it was far easier to understand, work with and expand to meet the ever growing demand. i know that very little time was wasted in completely striping out his code and replacing it so that new features and scaling support could be added.
That took some fairly smart folks. People Mr. Graham thinks very little of.
While i do not claim to have the level of intellect as Mr. Graham, i do know that any organization of sufficient size will have a wide range of developers. Indeed, you want a wide range of developers because not everyone wants to do things like support log parsing mechanisms. i also recognize the fact that when Mr. Graham left Yahoo at the close of the last millennium, there wasn't quite as strong a "hacker" culture as later developed.
So yes, please do read his treatise on why he believes Yahoo! failed, and understand that he does note with great clarity a number of business decisions that time has served to highlight. Also, please note, that at the time of his meeting with Jerry Yang, Mr. Yang was not the CEO, or COO of the company, but merely one of the founders who had quickly determined that his own inexperience would be disastrous in the role of lead business decision maker. Of course Jerry wasn't interested in the technology being offered, other people had already made the decision to acquire it. Jerry's interest was at a more personal level.
