I have not posted any work log for long. Even worse, I have not fully kept up with the log. So what’s going on? Is rsyslog development pace going down?
No, definitely not! I am just currently concerned with helping users, doing some labs myself, working a bit on phpLogCon and thinking about how to best do the scripting engine.
Oh yes, the scripting engine… I am promising and promising it. The reason it takes so long is that I want to have a perfect solution on the first shot (well, as close to perfect as possible…). And I want to keep it backwards-compatible to what we currently have. This involves a lot of things under the hood, and it requires very careful planning to do it right. In fact, I am even re-evaluating some theoretical concepts for the scripting language, maybe it makes sense to adopt a few more thoughts. But, even though you do not see it, I am still very focussed on it. It is the next big thing to happen.
Just to get you an example of the subtle problems: how to *best* handle the “*.*” PRI selectors that are in the current config at the rule of almost all filters. If that is not mapped properly, it can have an enourmous performance hit. To support it, I may need to do a constant folding phase in my interpretative language, something that I so far did not consider as needed. And I am still trying to avoid this (it offers little benefit for almost all other cases).
Whith that said, on to the (partial) work log:
Past day’s rsyslog work log:
– created autconf support for a rsyslog testbench
– created the first very rough beginnings of a RainerScript testbench
– disabled warnings caused by third-party libraries
– enhanced gtls driver to support fully async operation (there was
a bug before if gtls did issue EGAIN without getting any data)
– bugfix: comments after actions were not properly treated. For some
actions (e.g. forwarding), this could also lead to invalid configuration
applied to beta branch
… something missing, see git log ;) …
– released 3.19.8