< Back to Blogs
Interoperability Customer Council by admin on June 21, 2006 01:02PM

Last week we announced the formation of an Interoperability Customer Executive Council.  This is something many folks here at Microsoft have been working on and thanks go to a wide variety of people.  I was involved from an OSS perspective and will continue to be active with both our internal groups and this customer council to help this mission succeed.  As you can imagine, interoperability is a very wide and loosely defined word – it can mean many things to many people.  Particularly when you produce a lot of software!  Rather than try to ‘pick’ certain areas to focus on (believe it or not, Microsoft is constrained by time and money like all other businesses) we decided the most effective and beneficial path would be to have our customers drive this conversation. Thus was born this council, formed by customers of different sizes, business types, and geographic location, in order to provide guidance on interoperability issues that are most important to customers, including connectivity, application integration and data exchange. 

Interoperability in a heterogeneous environment doesn’t happen accidentally, nor does having your code open make everything ‘just work’ together.  It takes thoughtful design and architecture; it takes time, effort and engineering discipline.  It requires working closely with partners, competitors and customers.  It also requires mature understanding that all things don’t necessarily require interoperability, and frequently those things that do, sometimes require different types of interoperability.  Point being: it takes focus, energy and commitment.  This council is a great step and example of this commitment from Microsoft. 

I’m very excited by this step forward, below are some further news stories about this:

CNET News.com

Clients to advise Microsoft on software linking

http://news.com.com/Customers+to+advise+Microsoft+on+software+linking/2100-1001_3-6083664.html

eWeek

Microsoft Customer Council to Focus on Interoperability

http://www.eweek.com/article2/0,1895,1976394,00.asp

-Bill

Comments RSS
  1. fluke said:

    "Technical work on interoperability issues around Windows, Linux, Unix and open-source software is also taking place in Microsoft's open-source software lab, and the company has set up a community Web site, Port 25, about the issue."

    I think this goes to the heart of one of the key problems with Port 25.  For the most part, it seems like Port 25 has been peddle commerical "technical" solutions.  This even has go so far as to claim PowerShell (for Windows 2003) as the answer to a question involved Windows CE!  It also seems like we are dealing with a sales associate for "answers" instead of fellow developers.

    Not once in the life of Port 25 so far has the question come up what the community would like to see different about MS licensing of SDK, DDK or protocol/file documents.  Instead, we are given JDZ telling us that if a plug-in to Visual Studio is important to us then we should just write one.   He disregards that the license for Visual Studio SDK compromises creating such a plug-in under a FOSS license.

    I would like to see Port 25 live up to it's name.  Let's see interoperatablity of MS Sender-ID as a standard on all things that speak on "Port 25" (SMTP).  When the council is able to sit down with Eric Allman, Wietse Venema, D. J. Bernstein and a member of the Apache Spamassassin group to come up with an agreeable license for Sender-ID, then I will believe that Port 25 and the Interoperability Customer Executive Council are serious about making a difference.  Until then, why even continue discussing "industry wide" and "technical solutions to interoperability" at Microsoft while the same exact company artifically creates legal obstacles.

    If MicroSoft owes it to the share holders to maximize profits through such legal hoops for FOSS, then so be it.  But if that is the case, please stop being two faced and claiming that your seeking interoperability while keeping the road blocks in place.

    posted at 04:16PM 06/21/2006
  2. liquidat said:

    May I comment to these articles?

    "(But) we haven't had a formalized process where we pull people together and think about the various scenarios."

    This sounds as forming a real process with defined ways and clear structures for talking about such topics. That is a nice idea and should be followed since interoperability is a problem in today real life situations and the customers running into trouble need ways to talk with both parties which are causing the problems. The way for OSS is most often more direct (e-mail lists, bugzilla, etc.) and Microsoft can learn a lot by looking at such feedback-systems - and by listening to this feedback.
    Therefore I think this council is a good step.


    "open to ways of working with the open-source community broadly, and even in the GPL space we are trying to find ways in which we can build bridges to GPL, but the bridge has to be carefully constructed."

    This troubles me a bit: why do you care about the license? I don't think that anyone at Microsoft thinks about building kernel modules or linking to specific GPL-only libraries. I can't see any need for that at the moment because the software level where GPL software and Microsoft's software work together is never touching these areas.
    Better said, the main problems are far away from licensing interoperability as far as we talk about heredity of licenses.


    It would be nice if Port 25 could summarize some of these problems (please, not in a video; figures and facts have to be on "paper" :) ) and outline where Microsoft sees problems in the interoperability with GPL software. Additionally I would like to see more information about the first aims and technical topics this new council talks about.

    liquidat

    PS: As usual: it would be nice to get an explanation or to discuss how this new council fits to strategies like the MTP license (explicitly against OSS), the fact that several of Microsoft's SMB extensions are still closed or other similar issues which prevent the OSS world to work on the interoperability problem. But I'm getting the impression that this is not what Port 25 deals with (but then: why not?)... :/

    posted at 04:35PM 06/21/2006
  3. cronos said:

    I agree with fluke in all their statements. Do you want a concrete petition? We, at the plant where I work, are having problems with samba and clients, ranging from clients that doesn't connect, to clients that crashes samba. I'm not involved personally in this problem because I only work with free software, but the people that refuses to change their Windows stations and to spend money in a Windows server, surely would appreciate your help.

    And, if you are thinking that I work for a micro enterprise of ten people, in fact I'm working for a enterprise listed in the first 30 places of fortune global:

    http://www.finfacts.ie/Private/curency/fortune.htm

    Moreover, the enterprise has a membership in the OSDL, so I'm feel confident I can get a statement from our representative at OSDL, just in case you think I'm speaking only for me and my little department.

    posted at 06:57PM 06/21/2006
  4. einhverfr said:

    Different groups mean different things when they say "interoperate."  When I worked at Microsoft, the word had a very specific meaning:  Allow other products to work with Microsoft systems so that people can migrate *to* Microsoft software.  There are *no* provisions to allow systems to interop in ways that allow people to migrate *from* Microsoft systems.

    By this measure, Microsoft is doing a pretty good job.  And I would argue that Microsoft appears to be coming around and offering the sort of interop that encourage people to actually value Microsoft solutions as a part of the ecosystem.  However, there is still a *long* way to go in this regard.

    For example, I am doing a lot of work at one of my customer's sites at the moment (they are a fortune 100 company).  They have a *large* Microsoft license, and routinely, I have to install various distros of Linux on my test system along with Windows Server 2003.  I am finding all manner of interop problems because I cannot work on the test system when it is running Windows and still connect it to the network (because it is not managed by IT).  However, I can do so with Linux without violating any policies.  Thus I do most of my work using Linux because I can actually check my email (via OWA) and access web apps that I need for most of my work with them.

    In the process I have found a large number of interop issues involving highlighting and picture positioning in OpenOffice which become unchangeable in MS Office, a few issues involving Samba, and the like.  Sometimes I wonder if these are deliberate.  But most of the time I think it is just a matter of a lack of interop testing.  Why would Microsoft want to test interop with Samba?  Isn't it easier for them to just say "It works with out systems-- it is Samba's fault?"

    But again, the situation is admittedly getting better.

    Best Wishes,
    Chris Travers
    Metatron Technology Consulting

    posted at 01:38AM 06/22/2006
  5. jcannon said:

    fluke & cronos,
    Bill may jump in, but I wanted to let you know we're listening. I spend my time figuring out how to make Port 25 better for the community, so I take your comments very seriously. I see two components to your comments - one is about the site content & delivering on our promise. The other is a policy discussion, which is not something I have any influence over :)

    On the Port 25 side - you make some good points & you'll see more over the coming months that come closer to delivering interoperable content, research from the lab & responses to customer & site visitor feedback. We're still new (which is not an excuse), but we're still building the site out - you'll see threaded comments & podcasting soon, as well as getting content much more aligned to our tagline of interop lab work. In fact, you'll see a blog the first week of July about upcoming research & have an opportunity to supply feedback on methodology.  We're also on an approach to get our first research project out in mid July. So my point...we're getting there. Yes, we have work to do & we're listening  - so keep the feedback coming and rest assured folks are taking action on it.

    posted at 06:25PM 06/23/2006
  6. http://www.serverpipeline.com/blog/archives/2006/06/fud_never_chang.html

    I dunno, it's like they say, "FUD never changes".

    posted at 10:58PM 06/24/2006
  7. billhilf said:

    cronos, I'd be happy to take your input on the specifics of the issues you describe between Linux and Windows.  We'll try our best to look into these if we can get good details.

    liquidat, licenses matter, regardless if it's free or paid software.  My next blog (or the one after) is about licensing in our shared source programs.

    fluke, interseting comments, I'd love to see the example where powershell was recommended as a solution to WinCE - please point that out for me, I missed it somehow.  Also, you're right, we are trying to provide both insight (and opinions) and technical solutions.  
    -Bill

    posted at 01:23AM 06/26/2006
  8. fluke said:

    fuffybunnies:
    It is amazing how well the popular press can take one or two lines out of context and wrap a whole new story around it.  But did expect a little better from the editors at CMP.

    Take a look at:  http://www.serverpipeline.com/189600903

    Once the lines are put in context, it appear to me to be observations instead of FUD.  And while "Get the facts" did contain FUD, no where in the articles does CMP claim that BH made any statements about the accuracy of the campaign.  Instead, he only praised Martin Taylor for taking MS fragement responce to Linux and providing an offical single responce.

    The article by Don St. John refers to the Paula Rooney article but misses it's main point: "Linux thrives on Web servers, DNS servers and single, lightweight appliances that do 'one thing well.'"

    Don St. John's "FUD never changes" article skips this main point because then the rest of his article just sounds silly.  Don would like use to believe that what BH said would require "all those Wall Street firms that have rolled out Linux as their base for high-transaction, 24/7 financial services platforms just ran screaming for the East River as if they were being chased by zombies?"  Instead, BH has already admitted that there are niche areas that Linux thrives at.  However, that does not mean that Linux is a threat when it comes to market share.

    The Don St. Johns article is a straw man arguement which takes things out of context to build to a different set of claims than where actually said.  Instead, what should be focused on is that the observations about the future of Linux where only made regarding desktop and server enviroments.  It might be possible that for these two areas that Linux's market share will begin to slump.  But it is still interesting that the BH article says nothing about the future of embedded Linux.

    ----------

    billhilf:
    http://port25.technet.com/archive/2006/05/24/2531.aspx asks a question about file descriptor abstraction when porting applications to WinCE and XP.  The answer then tries to peddle PowerShell which I believe is available for neither WinCE or XP.  No where in the answer does is a solution for CE or XP ever discussed and Port 25 has never provided a follow up to address the problem.

    Also of interest is: http://port25.technet.com/archive/2006/05/18/2476.aspx where Port 25 recommends a very messy work-around to address that every Windows to date (including Vista beta) still supports an outmoded concept of contextless reserved file names (instead of moving them into a dev directory or eliminating them entirely).

    It is helpful that Port 25 provides both insight and technical solutions.  But to continue to drum technical solutions that don't apply (PowerShell for WinCE?!) or have JDZ demand we use the Visual Studio SDK is not helpful.

    I have a number of projects that I would be interested in that are at a stand still because the Windows DDK/IFS license is incompatible with basic FOSS concepts.  While I have even emailed Port 25 a month ago about this issue, I have yet to see any dialog openned on what clauses present a problem to FOSS contributions to improve the Windows enviroment.

    posted at 02:01PM 06/26/2006
  9. dreamer said:

    Bill: "..nor does having your code open make everything ‘just work’ together..."

    No, but it does help people to see exactly _why_ your code isn't interoperable:

    http://validator.w3.org/check?uri=http%3A%2F%2Fport25.technet.com%2Farchive%2F2006%2F06%2F21%2F2650.aspx

    ..which is the first step to making it so.

    Perhaps you should start with the basics and come back when you've mastered them? ;)

    d.

    Oh, and while you're there:

    http://jigsaw.w3.org/css-validator/validator?uri=http%3A%2F%2Fport25.technet.com%2FThemes%2FBlogs%2Fport25%2Fstyle%2Fstyle.css&usermedium=all
    http://jigsaw.w3.org/css-validator/validator?uri=http%3A%2F%2Fport25.technet.com%2FThemes%2FBlogs%2Fport25%2Fstyle%2Fprint.css&usermedium=all
    http://jigsaw.w3.org/css-validator/validator?uri=http%3A%2F%2Fport25.technet.com%2Fcustomcss.ashx&usermedium=all

    *Sheesh!*

    posted at 08:02AM 07/03/2006
  10. cronos said:

    "cronos, I'd be happy to take your input on the specifics of the issues you describe between Linux and Windows".

    Well, there was no input at all. I asked my coworkers for details about these failures, and even where it happened again, there was no response. I think they are more interested finding who could solve their problems, instead to know what are the problems in order to solve themselves.

    In the future, I will only post facts that I know at first hand. Because things like this, I usually not work with people that uses propietary software.

    posted at 09:14AM 07/19/2006
Post a Comment
*
*