| IBM has been accumulating a persona among my cadre (developers who consult for companies that are bootstrapping a new tech division on top of a profitable business). The problem is that having a boondoggle division reflects on the whole company. When I'm making decisions for a client, I never even think about what IBM solutions I could apply to a problem. I've only had a few projects using IBM products, admittedly most very old legacy systems -- but those introductions failed at just basic dev hygiene things. Documentation behind paywalls, couldn't get things to run on my dev machine without very experimental projects made by individuals - last updated in 2003. Just frustrating. I don't like advocating complete rewrites, I've happily updated a company's internal tool that was written in as3. Could have squeezed more hours out of them if they wanted to go with html5, but doing so didn't open up any new possibilities they wanted so I didn't. If your company is running on RPG, and it's just a CRUD app -- I'm going to save you money by recommending a rewrite under the two circumstances I've encountered in the wild. Honestly, this isn't their fault. It's awesome that their products have lasted so long for companies. I don't expect a company to keep up to date documentation / interoperable interpreters for 30 year deprecated tech. IBM's just so old that they've accumulated many layers of fossilized products, and those things have caused an unfair bias in my mind (and probably the minds of others) I'm sure that there are problems IBM research is uniquely qualified to solve, my perception is that they're the wrong choice for most problems. I'm pretty thankful for companies like them. (IBM, Siemens, SAS etc) Their expensive ERP systems have enabled me to raise my rate over the years off the information asynchrony that forms the foundation of their success. |