Hi all, First, a bit of introduction to myself. I'm a software engineer working at Google and was introduced to Tajo project by my friend, Keeyong Han, when I was looking for an open-source project to contribute to. My current role at Google is writing a REST API server in Java and Android client. But I also have a fair amount of experience with MapReduce and FlumeJava, so I'm hoping I can be of help in improving and enhancing the project.
As a newbie to the open source world and Hadoop, I decided to pick up the easiest tasks, and this one caught my eye. https://issues.apache.org/jira/browse/TAJO-967 I took a quick look at the source code, and it looked like a simple task, indeed. However, I also discovered tajoCli's history behavior was slightly different from others like bash and mysql. When the user enters multiple commands in a single line using semicolon and looks it up using up arrow key later, tajoCli shows each individual command separately, whereas bash and mysql shows all the commands in a single line as they were entered. For example, if the user enters two commands in a single line like this, \version; \d tajoCli shows \d first and then \version as the user presses up arrow key twice. But, bash and mysql show \version; \d together again with a single up action. I don't think bash and mysql's history behavior is necessarily better than tajoCli or vice verssa, but, IMHO, it'd be better if tajo provides user experience that users are used to and naturally expect. WDYT? If you agree, I can make the change myself. Thanks, Jaewoong
