Discussions

News: IBM Contributes Code to Apache Muse

  1. IBM Contributes Code to Apache Muse (2 messages)

    IBM Corp. is contributing more code to the open-source community based on the WSDM, pronounced as "Wisdom", (Web services distributed management) standard, according to a <a href="http://www.sda-asia.com/sda/news/psecom,id,9402,srn,4,nodeid,1,_language,Singapore.html" "target="_blank"" rel="nofollow">report on SDA Asia Magazine. IBM is making its implementation of the Common Event Base framework for WSDM freely available to the open-source Apache Muse project. The vendor previously contributed WSDM development tools to the Eclipse Foundation. IBM's software contributions are expected to spur further adoption of Web Services Distributed Management (WSDM), an industry standard that makes technology easier to manage. IBM's code provides a framework for building WSDM interfaces, making it easier for businesses to incorporate the standard into their IT infrastructure. The code is available as part of the open source Apache Muse Project, where developers contribute ideas and improvements to the code through open source collaboration. Do you believe open sourcing frameworks/contributing code to an open source framework will encourage other vendors in the same space to stop duplicating development efforts and work towards embracing one single, effective approach - in this particular case, the IBM approach to building WSDM interfaces? Message was edited by: joeo@enigmastation.com

    Threaded Messages (2)

  2. might breathe some life in it[ Go to top ]

    I just checked using svn log, no activity on the repository since February until last week. The handful of commits in February were the only activity so far this year. This seems to be the fait of many of the web services projects under the apache umbrella. The few projects that are alive (mainly axis and related projects) are of course quite important to the Java community If you check here: http://svn.apache.org/viewvc/webservices/ you can see that there is quite a few apache service related projects that have seeen no development activity whatsoever for several months. A simle svn log -r{20060202}:HEAD http://svn.apache.org/repos/asf/webservices// will confirm that these projects are not being worked on. It's an interesting way to check livelyness of a project before wasting an afternoon evaluating its releases. Many apache projects are dead as a doornail. So there may be two things going on here: 1) IBM has been working on this for several months, all along with the intention of committing to apache. 2) IBM has worked on this and decided to dump their code here because they are no longer willing to invest in it and are hoping that others will jump in. As outlined above it is quite common that this doesn't happen for some projects. Some of this code may later be reused in one of the other projects of course. Now as to the standards themselves. I've had some experience with the Globus wsrf implementation. Enough to avoid anything WSRF related for the next few years. The technology is really immature at this stage, as are most WS-* implementations except for the really big ones (e.g. WS-security). Anyway, the globus implementation was being worked on by IBM people and I suspect this is a newer version of the same stuff. WSDM and related standards are really aimed at the big toolvendors and their large customers (e.g. banks). Open sourcing the technology will generate little interest outside this community but may provide a nice code sharing platform for IBM and its competitors (most of whom actively contribute to a number of OSS projects).
  3. patience is a virtue[ Go to top ]

    I just checked using svn log, no activity on the repository since February until last week. The handful of commits in February were the only activity so far this year. This seems to be the fait of many of the web services projects under the apache umbrella.
    It might appear that the Apache Muse project was in a limbo until June but that has changed considerably since IBM joined the project. The new project lead is from IBM and many companies including Compuware and Cisco have actively increased their contribution, participation and commitment. There have been 2 milestone releases and 1 major release of Muse with a completely new improved programming model and full support for J2EE and OSGi runtime deployments. To the poster of this premature remark - open source projects take time to build a community and develop. For the small number of developers in the Muse team, the amount of work that has been delivered is quite remarkable. There is no other open source project that delivers real code for web services based distributed management (be it WSDM or WS-RT).
    The few projects that are alive (mainly axis and related projects) are of course quite important to the Java community

    If you check here: http://svn.apache.org/viewvc/webservices/ you can see that there is quite a few apache service related projects that have seeen no development activity whatsoever for several months. A simle svn log -r{20060202}:HEAD http://svn.apache.org/repos/asf/webservices// will confirm that these projects are not being worked on. It's an interesting way to check livelyness of a project before wasting an afternoon evaluating its releases.
    Sometime its better to err on the side of caution. While Axis had a lot of activity, the latest release was delayed multiple times. Muse might not have shown a lot of activity but delivered on time without forsaking quality. While your mechanism for evaluating a project's value is statistically interesting it is useful only to bean counters. IT architects and developers who are looking to be relevant in the future will be judging strategy and involvement by looking at a project across releases. And typically Apache will nix projects that are irrelevant - its of no use to anyone to have irrelevant projects lurking around.
    Many apache projects are dead as a doornail.

    So there may be two things going on here:
    1) IBM has been working on this for several months, all along with the intention of committing to apache.
    2) IBM has worked on this and decided to dump their code here because they are no longer willing to invest in it and are hoping that others will jump in. As outlined above it is quite common that this doesn't happen for some projects. Some of this code may later be reused in one of the other projects of course.
    Maybe you should add the real reason to your list that currently contains only shortsighted speculation: that IBM wanted to support an open standard by an open source reference implementation so that participating companies have an open way to collaborate on real software. IBM wouldn't be investing some of their best talent in open source projects just to have a code sharing platform or with the "hope that others would jump in". Its good to be cynical but without real facts to back you up its just useless.

    Now as to the standards themselves. I've had some experience with the Globus wsrf implementation. Enough to avoid anything WSRF related for the next few years. The technology is really immature at this stage, as are most WS-* implementations except for the really big ones (e.g. WS-security).
    I bet this is what people said when WS-Security was first proposed/worked on. I guess real specs just have to find a way past cynical spectators and move forward with the resolve of dedicated developers who real want to change lacking standards rather than just complain about them
    Anyway, the globus implementation was being worked on by IBM people and I suspect this is a newer version of the same stuff. WSDM and related standards are really aimed at the big toolvendors and their large customers (e.g. banks). Open sourcing the technology will generate little interest outside this community but may provide a nice code sharing platform for IBM and its competitors (most of whom actively contribute to a number of OSS projects).
    More speculation. IBM does not need to go open source to get a code sharing platform - that is really a naive shot in the air. There has been substantial interest and adoption of the Apache Muse runtime from many core IT companies, system integrators and integrated software vendors alike. Real product development does not happen in a snap - over the next couple of years, you will see a number of products both from IBM and outside supporting a common distributed management standard be it WSDM or an evolved form of WSDM. The only way to promote a new standard is by complementing it with real code. Code talks and that's exactly where IBM is putting its talent (and money).