Hi Emmanuel ,
Emmanuel Lecharny wrote:
John E. Conlon a écrit :
Hi Alex,
Thanks for the comments. See inline for responses.
Alex Karasulu wrote:
Doing this means committing to OSGi and I'm not going to be too
comfortable with doing this until I see:
We can look at this two ways. From one perspective we would not be
committing 'fully' to OSGi, as the decorated metadata would not
affect the behavior of the jars as they operate today.
One of the big issues for the OSGi community today is to convince
library providers to add a few lines of metadata to their jars
artifacts so they could also be used in OSGi projects. So from this
perspective I guess I am trying to convince us to 'at least' to make
our jars OSGi friendly.
...
Hi John, Alex,
First of all, I have to say that OSGi effort is valuable, and we have
to thank you, John, for you dedication and continuous effort regarding
it during those last months.
I want to address some of Alex concerns, and also give you my opinion
on the next few steps, and months. During last Apache Conference, we
have announced the first release of Apache Directory Server (1.0) and
Mina has become a TLP, separated from Directory. We also included
three sub-projects into Apache DS :
- triplesec,
- mitosis,
- ldap studio
This was a lot of work, and a lot of little things to manage (and you
know that it takes more time to manage many little things that one big
thing).
We are currently preparing a 1.0.1 release of ADS (a bug fix), quickly
followed by a 1.5.0 release, with a new Schema Management feature
(plus Java 5 support). We also have to make Mitosis stable for may (or
even better, before may !). The site has been totally redesigned -
thanks to Ersin, pam and many contributors-, and the documentation has
been improved - thanks to Stefan and Christine -. Again, this was a
lot of work.
As Alex stated, I think this time for dust to settle a little bit. We
have a damn loaded roadmap since ApacheCon EU (may 1st), and I'm not
sure we may dedicate enough attention to OSGification of ADS before
this date.
What I would suggest is that we should put this OSGification on the
roadmap starting on may, 4th, when we will be able to back it. This is
really true that if you don't get committers attention, then it will
be hard time (many complaints to be expected ...).
I'm sorry to admit that I didn't had a minute to follow up your work,
and I didn't had time to read anything about OSGi during last year. My
bad... I really want to jump into it, as I think it will help us on
many aspect.
So, can't we wait 3 more months before we jump into OSGi for ADS? If
you feel being ready, and that we should at least fulfill some first
steps that won't harm the project, then I will be pleased to follow
your instructions (like adding meta-info into the jars), but you will
have to be very explicit about what we must do, because we will be
like babies doing our first steps : keep the way really clear of any
bump and traps :)
Yes a few baby steps.
1. The first should be painless. If permitted I would begin by adding
the plugin to our pom.xml files (going one by one) to generate the jars
with the metadata.
2. The second step I think is something that everyone will (should)
applaud. As a matter of fact it was a subject of a recent thread.
Want to collaborate with someone ( Ole would you be interested in
working with me on this??) on an effort to start documenting all our
packages in all our projects with package-info.java files. This is the
bottom up documentation effort that I was referring to when I responded
to Alex. With this we should gain greater insight into the workings of
ADS down to the packages. As this happens I can begin to tune the
metadata to approaprately take advantage of what the module offers to
and requires of other packages. (Again this wont affect jar users, but
it will increase the decoupling of the osgi bundles.)
I don't want to slow you down, I just want to express my concern to be
unable to dedicate enough time for this task in the next few months.
And I guess this is also a concern for many others ADS committers...
You can ignore the OSGi side of the house. But you might be asked some
questions from time to time though, regarding the functionality of
packages (this in order to support the package doco effort).
warm regards,
John
