On 02/12/2015 03:55 AM, ianseeks wrote:
why do you use "grep" instead of editing the files manually? It makes life easier which is one of the uses of computers. How do you interpret MariaDB database files, do you grep them?
Perhaps SUN and its acquisition, never mind IBM coming up with AIX (and then Linux) and HP putting HP/UX on its Big Iron was an indicator. UNIX aka Linux as its present incarnation is for Serious Business not for the home computer. Many of the complaints against, for example, systemd, are to do with how it a tool for managing a large business environment or possibly a shared resource such as a service provider or "agency" business model. All geared towards resource management. So the mention of system logging and database in the same message make me recall my experiences there. I worked a few years ago at a first tier bank here in Canada. The first tier ones are BIG even by US standards. Because out government won't let them merge, they have grown by acquisition, but even so, they are BIG. My work brought me in contact with the guy in operations who managed system logging. Every machine on the corporate WAN sent syslog to the central server. This included every terminal server, every 'mainframe', every brokerage, every branch. He mentioned there were over 20,000 potential inputs, though not all were active at the same time. Some were though consolidators; for example Windows PCs logged in to servers and the servers sent the syslog. All this went into a huge database. All queries were on the database. The issue wasn't to 'grep' but to 'relate'. The database allowed the easy tracing of paths of activity, looking for who did what when and how. Yes, there were other automated triggers for such things as disk decay, QoS issues as well as suspicious activity. Most of these traces were beyond what might come from a human being simply grepping though a text file. Many were beyond what even 'swatch' could do. This was stuff we now call "Big Data". There were, I was told, many hundred of canned and custom searches and triggers that were all active in parallel. Some watched for activity spread over many days. To do this with a relational database was, well, not exactly trivial, but not difficult either. Being an indexed database the queries came back in reasonable time, something that grepping though a comparative text file could not. A lot of the criticism of systemd is really a 'that was then', googling for out of date articles and issues that have long since been addressed or made irrelevant. Nothing new here, we've seen it so many times before. KDE4 was a good example. The very nature of OpenSource is that in order to get the feedback developers need they have to release an immature product. Its not like a commercial firms such a Microsoft or IBM where there is a budget for and separate group for testing. So the immature release is taken to be a final one. Not so. As Ian and others are pointing out, systemd and its tool set are a moving target, responding to issues such as this. As Ian says:
Maybe you don't know but when developing software, it takes time and not all functions are ready on day one.
-- A: Yes. > Q: Are you sure? >> A: Because it reverses the logical flow of conversation. >>> Q: Why is top posting frowned upon? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org