Proposed plan

[Dmitri Pal]

Hi,

So where we are and what are the next steps?

1) There is a question about glibc and integration of the ul library
into it. I will take an action item to investigate.
2) I sent a header to review. Please review and provide feedback. Is it
the right API? It is reasonable or this is completely off? The file is
attached yet again.
3) Are we satisfied with the latest XML spec Keith published? Please
review and ack/nack.
4) Do we need to have a call exposed from the syslog implementation that
would tell the library a preferred encoding (JSON, BSON, XML etc.). Let
us have a thread about it. Rainer what is your take on this?
5) Is there anything else that we need to discuss or review?
6) Do we need another call?

Sorry for jumping in. I just need to get organized myself to better plan
my time and creating a list and plan helps a lot.

[William Heinbockel]

I would reprioritize them:

1. Finalize the umberlog API and JSON-Syslog compatibility/support.
Test. Write tutorials/guidance. Push to apt and yum repos. Work with
Linux vendors and upstream package maintainers to adopt new syslog
interface.
If we are going to push something to glibc, this will probably be the
basis for it.

2. Identify additional syntax features and minimal set of types &
event message requirements (e.g., XML Schema structures)

3. Define basic API requirements to align the various API interfaces
This is probably a combination of Dmitri's selog header, Rainer, and
Gergely's input.

4. Add structured log support to log4j, log4c, and other logging libraries

I will work with Keith to shore up the XML Schema and propose an
alignment of JSON and XML event structures. Let's delay any further
format discussions until we have a more mature JSON and XML support.

On Tue, Mar 27, 2012 at 12:51 PM, Dmitri Pal  wrote:
> Hi,
>
> So where we are and what are the next steps?
> There is a question about glibc and integration of the ul library
>
> into it. I will take an action item to investigate.
> 2) I sent a header to review. Please review and provide feedback. Is it
> the right API? It is reasonable or this is completely off? The file is
> attached yet again.
> 3) Are we satisfied with the latest XML spec Keith published? Please
> review and ack/nack.
> 4) Do we need to have a call exposed from the syslog implementation that
> would tell the library a preferred encoding (JSON, BSON, XML etc.). Let
> us have a thread about it. Rainer what is your take on this?
> 5) Is there anything else that we need to discuss or review?
> 6) Do we need another call?
>
> Sorry for jumping in. I just need to get organized myself to better plan
> my time and creating a list and plan helps a lot.
>
> --
> Thank you,
> Dmitri Pal
>
> Sr. Engineering Manager IPA project,
> Red Hat Inc.
>
> Looking to carve out IT costs?
> www.redhat.com/carveoutcosts/
>
> lumberjack-developers mailing list
> [e-mail]
> https://fedorahosted.org/mailman/listinfo/lumberjack-developers

[Dmitri Pal]

On 03/27/2012 02:35 PM, William Heinbockel wrote:
> I would reprioritize them:
> Finalize the umberlog API and JSON-Syslog compatibility/support.
>
> Test. Write tutorials/guidance. Push to apt and yum repos. Work with
> Linux vendors and upstream package maintainers to adopt new syslog
> interface.
> If we are going to push something to glibc, this will probably be the
> basis for it.
> Identify additional syntax features and minimal set of types &
>
> event message requirements (e.g., XML Schema structures)
> Define basic API requirements to align the various API interfaces
>
> This is probably a combination of Dmitri's selog header, Rainer, and
> Gergely's input.

Rainer and Gergely, can I have you input?

Bill,
I really need to untangle myself from this project in a long run. I am
not a developer any more. To do that I need to do as much as possible
with API and pass it to you or someone else. I think I have a lot of
pieces to create the first cut of the interface from the ideas that are
discussed on this list and already implemented in ELAPI and collection.
It will be a pity to waste all this and start from scratch. But I can't
just "do it some day later". IMO there is a momentum now.  I just need a
green light here.
This is why I am asking for some guidance and review.
It might not be a priority for you but it is definitely a priority for me.
> Add structured log support to log4j, log4c, and other logging libraries

IMO this support should be on top of the new library however last time I
checked log4j and log4c did not render itself well to the ideas we have
about how the structure should look like. It will be really a challenge.
> I will work with Keith to shore up the XML Schema and propose an
> alignment of JSON and XML event structures. Let's delay any further
> format discussions until we have a more mature JSON and XML support.
>
> On Tue, Mar 27, 2012 at 12:51 PM, Dmitri Pal  wrote:
> > Hi,
> >
> > So where we are and what are the next steps?
> > There is a question about glibc and integration of the ul library
> >
> > into it. I will take an action item to investigate.
> > 2) I sent a header to review. Please review and provide feedback. Is it
> > the right API? It is reasonable or this is completely off? The file is
> > attached yet again.
> > 3) Are we satisfied with the latest XML spec Keith published? Please
> > review and ack/nack.
> > 4) Do we need to have a call exposed from the syslog implementation that
> > would tell the library a preferred encoding (JSON, BSON, XML etc.). Let
> > us have a thread about it. Rainer what is your take on this?
> > 5) Is there anything else that we need to discuss or review?
> > 6) Do we need another call?
> >
> > Sorry for jumping in. I just need to get organized myself to better plan
> > my time and creating a list and plan helps a lot.
> >
> > --
> > Thank you,
> > Dmitri Pal
> >
> > Sr. Engineering Manager IPA project,
> > Red Hat Inc.
> >
> > Looking to carve out IT costs?
> > www.redhat.com/carveoutcosts/
> >
> > lumberjack-developers mailing list
> > [e-mail]
> > https://fedorahosted.org/mailman/listinfo/lumberjack-developers

[Rainer Gerhards]

> -----Original Message-----
> From: [e-mail]
> [mailto:[e-mail]] On Behalf
> Of Dmitri Pal
> Sent: Wednesday, March 28, 2012 1:20 AM
> To: William Heinbockel
> Cc: lumberjack logging
> Subject: Re: [lumberjack] Proposed plan
>
> On 03/27/2012 02:35 PM, William Heinbockel wrote:
> > I would reprioritize them:
> > Finalize the umberlog API and JSON-Syslog compatibility/support.
> >
> > Test. Write tutorials/guidance. Push to apt and yum repos. Work with
> > Linux vendors and upstream package maintainers to adopt new syslog
> > interface.
> > If we are going to push something to glibc, this will probably be the
> > basis for it.
> > Identify additional syntax features and minimal set of types &
> >
> > event message requirements (e.g., XML Schema structures)
> > Define basic API requirements to align the various API interfaces
> >
> > This is probably a combination of Dmitri's selog header, Rainer, and
> > Gergely's input.
>
> Rainer and Gergely, can I have you input?

Digesting the volume of mail. Have to admit I am a bit overwhelmed (you know
that feeling ;)). So far, I prioritized the discussion on the "next thing",
that is umberlog, much less the type ect stuff, somewhat in line with Bill's
thoughts. Will try to do a decent review and provide more feedback today.
> Bill,
> I really need to untangle myself from this project in a long run. I am
> not a developer any more. To do that I need to do as much as possible
> with API and pass it to you or someone else. I think I have a lot of
> pieces to create the first cut of the interface from the ideas that are
> discussed on this list and already implemented in ELAPI and collection.
> It will be a pity to waste all this and start from scratch. But I can't
> just "do it some day later". IMO there is a momentum now.  I just need
> a
> green light here.
> This is why I am asking for some guidance and review.
> It might not be a priority for you but it is definitely a priority for
> me.

I understand your discussion. From the glimpse of casually reading the mail
traffic, I tend to agree to what you propose. But I share Bill's view that we
should probably let mature some of the early work before we take additional
steps. It may be somewhat of a turn off if we now do a full blown API that we
offer ... and experience will later tell us we need to make larger changes.

I actually have experience in this regard, because libee followed exactly
this approach roughly 1 to 2 years ago, based on the CEE ideas at that time.
I have now a hard time not to break anything I built on top of it. Even more,
the interface considerably changes and this is hard to explain to existing
users (thankfully only few in libee's case and only by proxy of liblognorm,
which makes things easier).

It's a hard decision for me if it is a good idea to try to do all of this in
parallel, on unproven ground, but with big momentum rather than use an
evolutionary approach and lose some momentum by this...

Rainer