Hi!

Mark> Frank, Mark> Can you tell us what is the future of smilib in the sence of: Mark> - future developments Mark> - your managment of the project Mark> - Any other information Mark> And is there any adoption of the sming? , is it standartize by now?

Libsmi started as the practical part of my Diploma thesis in 1999 which focused on an `improved SMIv2' and a compiler to convert SMIv2 to `SMIng' and back. The modular concept of the software, the release under an open source license, and the tools on top of the base library (smilint and smidump) turned it into a quite successful project. Over time, we added a number of smidump drivers driven by several research needs. We added the tool smidiff driven by practical needs. Other parts are driven by projects that evolved besides libsmi, like the SNMP command line tool (scli) [1] and the Java AgentX sub-agent toolkit (JAX) [2].

Currently, we are working on stuff to convert MIB modules and MIB data to XML Schemas and XML instance documents. Another active development takes place to contribute an SPPI parser/checker and to enhance the libsmi data structures so that they can convey all information supplied in PIBs. From time to time we identify new typical errors made in SMIv2 MIB design and we try to add error checks to the parser so that the usefulness of smilint increases slowly.

The SMIng stuff that is currently part of libsmi is still that old SMIng proposal from 1999 [3]. It is quite buggy and it probably even does no longer compilable cleanly. We did not yet work on a new SMIng parser conforming to any of the (two remaining) proposals that are discussed in the IETF SMING working group these days [4,5,6]. However, when these discussions tend to come to a consensus, we hope to find time and/or students to work on an implementation as part of libsmi.

Most efforts are mainly driven by personal fun and needs of a very small group at the TU Braunschweig and the University of Osnabrueck, namely Juergen Schoenwaelder and Frank Strauss as University research staff members and Torsten Klie and Moritz Bunkus as students who spen[td] a lot of fruitful time on the project. A lot of bug fixes and enhancements have also been contributed by other people like Bill Fenner and many others (see the THANKS and ChangeLog files). Thus, I would not say that there are strong plans and deadlines for any milestones within the project. And since the software is released under an open source license, you at least don't have to fear that upcoming problems cannot be fixed, even if the project initiators would not be able or willing to help some day (which I hope doesn't happen, but which will happen some time).

[1] http://www.ibr.cs.tu-bs.de/projects/scli/ [2] http://www.ibr.cs.tu-bs.de/projects/jasmin/jax.html [3] http://www.ibr.cs.tu-bs.de/users/strauss/sming/ [4] http://www.ietf.org/html.charters/sming-charter.html [5] http://www.ibr.cs.tu-bs.de/projects/sming/ [6] http://www.ietf.org/internet-drafts/draft-bierman-sming-ds-02.txt

-frank