Fwd: DoD, Intel Officials Bullish On Open Source Software; Government-wide Software Foundation In The Mix

[John Scott III]

http://defense.aol.com/2012/10/30/dod-intel-officials-bullish-on-open-source-software-government/



-----------------------------------------------------------
John Scott
240.401.6574
< jms...@gmail.com >
http://powdermonkey.blogs.com
@johnmscott

[Joshua L. Davis]

it is time gentlemen

--
Josh
678.831.0182
@ijoshuadavis

--

[C Drew Thornton]

I just joined today...hello!

 

Might Codice be the cross be the cross-government open source foundation that Dan Risache mentioned in the article?

 

...putting the pieces together....

 

Drew

[John Scott III]

yes...

:-)

we are trying to slowly piece it together 

--
 
 

[Kit Plummer]

Wow, that was nice of Dan.  :)

Yep, we have a few projects we're lining up, and trying to meet all of their infrastructure, license and service needs.  

More news to come soon.

Kit

--
 
 

[C Drew Thornton]

Alright, another question...has any effort been made regarding outreach to other *branches* of the government? I'm willing to commit some effort towards that, to broaden the focus away from just the executive side of things.

Drew

[John Scott III]

so far we've been focused on 2 code-bases to transition into it, we had a larger meeting with other rep's from parts of the gov. 

We think the best thing to do now is focus on specific code bases, either getting them OSS and having a place for them to go or config's of OSS that need a home and are meant for gov use

Any ideas?

[C Drew Thornton]

I definitely think that's a good idea at this time. Providing case studies that are also live...because that will be the point of Codice, correct? To be a public repository for GOSS and guidance? And help to reduce duplication of GOSS projects across agencies/departments/branches? (I can hear myself upspeaking a lot in my head)

 

At this point, though, I can only be an advocate for getting OSS into the natural procurement process. And that's where I want to help... I don't work on any code bases--I'm more of a sysadmin--but I'll be leveraging OSS in my next three projects (I'm seriously new to GOSS. I'll be using OWF for the first time soon). Another project that I'm supporting will (amazingly) be introducing the entire concept of OSS from scratch.

 

What I also wonder is if Codice will be something like Apache or Apache Jakarta or Apache Commons, with an oversight function that allows for duplication and eventual integration of subprojects, in addition to accepting subproject proposals...

[Kit Plummer]

A few comments inline below:

On Oct 31, 2012, at 8:28 AM, C Drew Thornton <drew.t...@gmail.com> wrote:

I definitely think that's a good idea at this time. Providing case studies that are also live...because that will be the point of Codice, correct? To be a public repository for GOSS and guidance?

We don't intend on replicating existing OSS repositories (Github, Bitbucket, Sourceforge, et al.).  DVCS changes the game there.  We will provide the governance that dictates there be a "master" repo (that Codice controls), which is the one that flows to releases (and to the high side).

We'll definitely provide guidance around managing source in a DVCS way, and how to manage the continuous delivery flow.

And help to reduce duplication of GOSS projects across agencies/departments/branches? (I can hear myself upspeaking a lot in my head)

This is a growing concern/need.  Discovery is definitely a problem.  Codice isn't meant to be THE repository for government-developed projects.  There are many options that already exist for that.  While we work with Mil-OSS to help the advocacy function, Codice will be selective (in the Apache way), working with projects that have high community potential.  E.g. Ozone Widget Framework

 

At this point, though, I can only be an advocate for getting OSS into the natural procurement process. And that's where I want to help... I don't work on any code bases--I'm more of a sysadmin--but I'll be leveraging OSS in my next three projects (I'm seriously new to GOSS. I'll be using OWF for the first time soon). Another project that I'm supporting will (amazingly) be introducing the entire concept of OSS from scratch.

Interesting.  We, and Mil-OSS, have lots of resources in that area…don't hesitate to reach out.

 

What I also wonder is if Codice will be something like Apache or Apache Jakarta or Apache Commons, with an oversight function that allows for duplication and eventual integration of subprojects, in addition to accepting subproject proposals…

Yes, Codice will be an Apache-like organization.  You can see/follow our early efforts here: http://codice.org

The idea is to provide top-level governance to cover projects, and internal-to-project governance in the form of project management committees.

Codice will provide infrastructure (continuous delivery tools), guidance, and various levels of advocacy to the community (government, academia and industry).

--
 
 

[C Drew Thornton]

BTW. Anyone going to the Raytheon Cross Domain Tech Exchange tomorrow? http://www.trustedcs.com/newsandevents/TechExchangeEvent2012.html

I have yet to meet up with any one from Mil-OSS (which is how I got to here).

This is all good, because being able to tell the parts of the govt that are new to OSS of the existing support will be important.

A couple things inline:

On Wednesday, October 31, 2012 11:49:25 AM UTC-4, Kit Plummer wrote:

A few comments inline below:

On Oct 31, 2012, at 8:28 AM, C Drew Thornton <drew.t...@gmail.com> wrote:

I definitely think that's a good idea at this time. Providing case studies that are also live...because that will be the point of Codice, correct? To be a public repository for GOSS and guidance?

We don't intend on replicating existing OSS repositories (Github, Bitbucket, Sourceforge, et al.).  DVCS changes the game there.  We will provide the governance that dictates there be a "master" repo (that Codice controls), which is the one that flows to releases (and to the high side).

We'll definitely provide guidance around managing source in a DVCS way, and how to manage the continuous delivery flow.

And help to reduce duplication of GOSS projects across agencies/departments/branches? (I can hear myself upspeaking a lot in my head)

This is a growing concern/need.  Discovery is definitely a problem.  Codice isn't meant to be THE repository for government-developed projects.  There are many options that already exist for that.  While we work with Mil-OSS to help the advocacy function, Codice will be selective (in the Apache way), working with projects that have high community potential.  E.g. Ozone Widget Framework

To clarify for my sake... I wasn't suggesting that Codice be the place where all the work and commits are done, but what you mean is that the cumulative "releases" are what gets set back to Codice for the public's (and other govt) benefit, right? 

 

 

At this point, though, I can only be an advocate for getting OSS into the natural procurement process. And that's where I want to help... I don't work on any code bases--I'm more of a sysadmin--but I'll be leveraging OSS in my next three projects (I'm seriously new to GOSS. I'll be using OWF for the first time soon). Another project that I'm supporting will (amazingly) be introducing the entire concept of OSS from scratch.

Interesting.  We, and Mil-OSS, have lots of resources in that area…don't hesitate to reach out.

 

What I also wonder is if Codice will be something like Apache or Apache Jakarta or Apache Commons, with an oversight function that allows for duplication and eventual integration of subprojects, in addition to accepting subproject proposals…

Yes, Codice will be an Apache-like organization.  You can see/follow our early efforts here: http://codice.org

The website is wonderfully succinct, but I joined the list hoping to get more involved, and I'm not sure how open you are to letting others in yet (which I totally get; I'm not trying to butt my way in). Plus, I can tell from the Mil-OSS group that you guys are also some of the main driving forces behind it, so Codice is a long time coming for you all, and it seems kind of important! I can be a stenographer or a historian...like an RFC editor...something that is missing even from opensource.org. For example, I had a terrible time last night finding release dates for the Open Source Definition version 1.1-1.9. (Mostly failed, OSI seems to have a private mail list for those functions).

The idea is to provide top-level governance to cover projects, and internal-to-project governance in the form of project management committees.

Codice will provide infrastructure (continuous delivery tools), guidance, and various levels of advocacy to the community (government, academia and industry).

My final question regarding your last two comments is this: how does this extra layer appeal to the project managers? Will PMs be happy that upstream support is required for their projects and not feel like it is an another burden? Will they feel the same way that they do when a FOIA request is submitted? I am supposing that there is pressure to get the software out into the open, but you tell me!

[Kit Plummer]

Good stuff.  And the dialog is a great form of history…especially in its transparent state.  Keep asking questions!

More below.

Kit

On Oct 31, 2012, at 10:35 AM, C Drew Thornton <drew.t...@gmail.com> wrote:

BTW. Anyone going to the Raytheon Cross Domain Tech Exchange tomorrow? http://www.trustedcs.com/newsandevents/TechExchangeEvent2012.html

I have yet to meet up with any one from Mil-OSS (which is how I got to here).

This is all good, because being able to tell the parts of the govt that are new to OSS of the existing support will be important.

A couple things inline:

On Wednesday, October 31, 2012 11:49:25 AM UTC-4, Kit Plummer wrote:

A few comments inline below:

On Oct 31, 2012, at 8:28 AM, C Drew Thornton <drew.t...@gmail.com> wrote:

I definitely think that's a good idea at this time. Providing case studies that are also live...because that will be the point of Codice, correct? To be a public repository for GOSS and guidance?

We don't intend on replicating existing OSS repositories (Github, Bitbucket, Sourceforge, et al.).  DVCS changes the game there.  We will provide the governance that dictates there be a "master" repo (that Codice controls), which is the one that flows to releases (and to the high side).

We'll definitely provide guidance around managing source in a DVCS way, and how to manage the continuous delivery flow.

And help to reduce duplication of GOSS projects across agencies/departments/branches? (I can hear myself upspeaking a lot in my head)

This is a growing concern/need.  Discovery is definitely a problem.  Codice isn't meant to be THE repository for government-developed projects.  There are many options that already exist for that.  While we work with Mil-OSS to help the advocacy function, Codice will be selective (in the Apache way), working with projects that have high community potential.  E.g. Ozone Widget Framework

To clarify for my sake... I wasn't suggesting that Codice be the place where all the work and commits are done, but what you mean is that the cumulative "releases" are what gets set back to Codice for the public's (and other govt) benefit, right? 

Yep.  We'll help the projects streamline that flow.  And, to whatever extent we can - we'll work with Forge.mil/MACE/MAX/etc. to optimize the bidirectional activities.  While we're focused on "releases" - we will govern the "authoritative" source repository as a whole, and through each PMC individually.  We just don't need to duplicate existing services.  

I'm working, literally right now, on setting up the artifact repository which will hold the "official" releases, SNAPSHOTs, and any nightly releases.  One of the things we are also looking at - is being an official/blessed repo for other, non-Codice projects.  There's a lot involved in managing/integrating the many different certification processes so we don't want to rush into that…but it's on the radar.

 

 

At this point, though, I can only be an advocate for getting OSS into the natural procurement process. And that's where I want to help... I don't work on any code bases--I'm more of a sysadmin--but I'll be leveraging OSS in my next three projects (I'm seriously new to GOSS. I'll be using OWF for the first time soon). Another project that I'm supporting will (amazingly) be introducing the entire concept of OSS from scratch.

Interesting.  We, and Mil-OSS, have lots of resources in that area…don't hesitate to reach out.

 

What I also wonder is if Codice will be something like Apache or Apache Jakarta or Apache Commons, with an oversight function that allows for duplication and eventual integration of subprojects, in addition to accepting subproject proposals…

Yes, Codice will be an Apache-like organization.  You can see/follow our early efforts here: http://codice.org

The website is wonderfully succinct, but I joined the list hoping to get more involved, and I'm not sure how open you are to letting others in yet (which I totally get; I'm not trying to butt my way in). Plus, I can tell from the Mil-OSS group that you guys are also some of the main driving forces behind it, so Codice is a long time coming for you all, and it seems kind of important! I can be a stenographer or a historian...like an RFC editor...something that is missing even from opensource.org. For example, I had a terrible time last night finding release dates for the Open Source Definition version 1.1-1.9. (Mostly failed, OSI seems to have a private mail list for those functions).

Well, as you can imagine, the skids are greases and we have movement, but the wheels aren't really turning just yet.  Yes, Codice is bit of an evolutionary step for Mil-OSS.  Rather than twist Mil-OSS into something it probably isn't suited for, legally or otherwise we've started Codice.  Obviously, there is overlap and we're considering various tactics to prevent duplication of any effort - or even structuring the organization into one form in the future (where Mil-OSS would be focused on advocacy in the mil/IC landscape), allowing other similar organization to grow under that umbrella.

There's definitely a need for a better form of collecting the relevant information.  I've always thought a wiki was the right answer and there's been some movement within Mil-OSS to support that.  I'm setting up a wiki for Codice, to home all of our documentation (including our operational ones like bylaws, contributor and other software agreements, etc.).  We definitely want to capture all of the information that will aid in both the consumption and production of open source software/hardware in the government space.  Once this is up, there'll likely be a call to arms to help populate these wikis in an organized manner.  We may even ask for your help in organizing and operating that.

The idea is to provide top-level governance to cover projects, and internal-to-project governance in the form of project management committees.

Codice will provide infrastructure (continuous delivery tools), guidance, and various levels of advocacy to the community (government, academia and industry).

My final question regarding your last two comments is this: how does this extra layer appeal to the project managers?

This is a good question.  It depends on the project.  Like Apache they tend to focus on libraries, frameworks, and other tools/servers that underpin bigger projects.  We see Codice fitting in the same way, and don't have any ambition in dealing directly with PoRs.  That said, if you look at OWF, they already have the democratic structure that makes it a perfect candidate.  We don't want to supplant that structure, just support it.  We can help reduce the resource burden for the typical engineering tools and processes as well - in a way that facilitates multiple-contractor/individual/organization operations.  One of the keys, is that by being true open source - we're discoverable in the same way that real open source is.  By creating/enforcing democratic structures around these middleware projects - it ensures potential users that they aren't controlled by any single entity and have the community to sustain it's life.

Will PMs be happy that upstream support is required for their projects and not feel like it is an another burden?

According to policy any open source used must be COTS, and by definition it is.  But, there's also the need to have a support structure around the project.  Codice isn't the support structure.  Like I said in my previous answer, Codice will just ensure that there is the potential for more than a single support channel.  It is still up to the integrator/developer to research the community and health of the project.  I feel like we're not proposing anything new in this regard.  Codice software should be considered in the very same ways as Log4J, or MongoDB.  The difference is that Codice software will generally be targeted at government/mil/IC problems.

Will they feel the same way that they do when a FOIA request is submitted? I am supposing that there is pressure to get the software out into the open, but you tell me!

Funny you should raise this question.  We're talking with a group who has some software that supports MIL-STD-2525.  Just about everybody has a need for it, and it was developed by government developers.  The problem is that it is controlled by PoR(s).  Obviously Codice shouldn't be used to undermine PoRs, so we're trying to get there via the high road, and working with people on both sides to get it open source and out to the public domain.

Our tag line "a Public Code Conservancy" is really applicable in this regard.  Stop wasting tax-payer dollars duplicating the effort n-times over is one form of pressure.

--
 
 

[John Scott III]

i can't hopefully someone will take good notes!