Hi Niklas!

Your feedback is much appreciated.  Thanks very much.

On Mon, Jul 14, 2008 at 3:33 PM, Niklas Gustavsson <[EMAIL PROTECTED]> wrote:
> On Mon, Jul 14, 2008 at 8:57 PM, Les Hazlewood <[EMAIL PROTECTED]> wrote:
>> Just out of curiosity though, how come you don't seem to mind the
>> tight-coupling to the SLF4J API as much as I do?  In general,
>> tight-coupling of _any_ 3rd party API really really irritates me.
>
> As a fan of the pure SLF4J approach I see SLF4J as the very extension
> to the JRE, something that's always around. One would have hoped that
> JUL would have been just that, but since it's not, SLF4J has had to
> play that roll. I'm thinking that it will be around in the large
> majority of cases where JSecurity is used anyways and therefore see a
> custom logging API as more burden than using SLF4J directly.

I agree with you, but the sad reality is that SLF4J is not embedded
into the JRE, so therefore I must view it as a proprietary 3rd party
library.  And as to its ubiquity, I haven't yet needed to use SLF4J in
any of my most recent projects (and I have a lot of them), nor do any
of the open source projects I use currently require SLF4J (Spring,
caching frameworks, etc).

Now of course, the next version of Hibernate will require it, but not
too many others - the huge majority of frameworks still require
Commons Logging. I do feel that this will change over the next year
though, since SLF4J adoption is occurring quite rapidly - which is why
SLF4J is my preferred logging API.  JCL will be replaced soon enough
across the board.

>
>> There's nothing wrong with it if the code is so trivial that a monkey
>> could understand it.  I'm not asking anyone to do ANY effort.  I'm not
>> asking anyone to maintain it.  I've even said that if we actually have
>> a problem with it - that is, something _actually_ comes up _in
>> practice_ that indicates it is not working as expected, that I'll do
>> the 5 minute code change it takes to force the SLF4J dependency.
>
> This is not a compelling argument for me. The code on this project is
> shared among all its commiters/contributors and any code therefore
> must be understood and possibly maintained by any of those. Besides,
> the project might be around for longer than any particular commiter
> (just see MINA for a prime example). To me, that's part of why I value
> the "Community over code" tagline.

I agree with this too - my point was only that it is so easy that
anyone with an IDE can search-and-replace the relevant code in about 5
minutes - not just me.  So even if I wasn't maintaining it, it would
still be a trivial exercise.

Thanks for chiming in!

- Les

Reply via email to