3 ms·
"heads down, iterating and releasing the most crucial features first" That's the key. The "number" of technologies involved is moot. Every piece of software i
by hammerdr 16y ago
"heads down, iterating and releasing the most crucial features first"
That's the key.
The "number" of technologies involved is moot. Every piece of software is build upon thousands of layers of technology. Adding one or more ingredients to the soup doesn't necessarily ruin it.
Often the simplest way to solve a problem is to duct-tape two different items from different technology stacks just well enough that it works for you. In that simplest case, you may be duct-taping 2 technologies together with a 3rd.
- kls 16y agoI disagree somewhat, one of the reason the web is a battle ground of corpses, that where once projects is due to the fact that it became a convoluted mess. The bar to be proficient required HTML, CSS, XML, (JSP, ASP, PHP pick one) generally some frameworks (taglibs, struts, tiles), SQL, Web servers, applications servers, Database servers, CMS servers that double as a bad application server, the list sprawls out from here. Developers had to master all of this just to put a page together. Technology soup can bloat to the point where it is unmaintainable. When you get into the enterprise you can add MQ, LDAP, SAP, Peoplesoft and a slew of others, that require integration. SOA and ESB's are doing a lot to clean it up and give a single implementation technology but it is still a mess. Jaguar once had the distinction of being the automobile with the most moving parts, they also had the distinction of being one of the most unreliable vehicles. Technology for technologies sake is always a project killer. I have seen it too many times. When I take over a troubled project (one of my specializations is troubled project rescue) one of my first efforts is to reduce the amount of technology load the project carries. That coupled with a focus on MVP saves a good deal of projects.