[slf4j-user] Recommendation for library writers: use thiscopy and paste wrapper around SLF4j to avoid default static initialization

[Adam Gent]

I concede on the performance. It is been a while since I have
attempted micro benchmarks particularly with the modern logging
backends. Its also not my area of expertise.

But the chicken/egg problem is real albeit solvable particularly with
more powerful logging backends.

My little wrapper proposition seems like a pretty crappy idea now...
oh well. It did help us with some issues (chicken/egg and some fringe
case unit testing issues).



On Sat, Mar 18, 2017 at 5:17 PM, Joachim Durchholz <jo at durchholz.org> wrote:
> Am 18.03.2017 um 21:26 schrieb Adam Gent:
>>
>> I mean I generally agree my arguments are fairly weak because I'm
>> defending a small minority of libraries. That is part because
>> libraries that have these problems (chicken/egg, performance issues)
>
>
> I still need to see how these problems even exist.
>
>> just avoid logging all together or write their own (e.g.
>> https://github.com/EsotericSoftware/kryo has its own).
>
>
> Kryo does its own logging not due to chicken/egg problems but because Nathan
> believes the complexity of an extra logger implementation is totally worth
> it if you can optimize out the logger.trace calls statically.
> It's a minority position, and one that I don't share because if we really
> start to count function calls then Java isn't the right tool for the job
> anyway.
> What happened is that people stopped arguing with Nathan and started writing
> an SLF4J adapter for Kryo, which seems to be working fine.
> _______________________________________________
> slf4j-user mailing list
> slf4j-user at qos.ch
> http://mailman.qos.ch/mailman/listinfo/slf4j-user



-- 
CTO
SnapHop (snaphop.com)
(twitter) @agentgt (linkedin) http://www.linkedin.com/in/agentgt
(cell) 781-883-5182