Re: [science-iwg] Triquetrum conf call : Thursday June 18th, 17:00 UTC/G

[Christopher Brooks <cxh@xxxxxxxxxxxxxxxxx>]

Hi Jay,

Right, sorry about the misstep about openness.  Over time my works seems to have moved away from newsgroups and large public mailing lists to smaller lists with smaller communities.  I'm fine with having the conversation be public. 

The control issue is not a huge issue, and it certainly has come up in the past.  Just something to watch for is all.

Part of my email really was about reusing the code in the ptII core.  These would be the classes in the kernel, actor, actor.lib and other packages. We don't have to reuse this code, but making these classes available could be a win.  The issue is that proving that the code is not encumbered and getting assignments of copyright could be tricky, but is doable.

I'll post to the forum momentarily.

_Christopher

On 6/9/15 8:39 AM, Jay Jay Billings wrote:

Christopher,

Thanks for the letter. Its great to meet you. I went ahead and CC'ed the Science Working Group list on this since it has turned into a technical discussion. Setting up meetings is OK to do in private, but we need to keep technical discussions in the open.

I absolutely love the name Triquetrum. I'm an astrophysicist, so I know it well.

I am very glad to hear that you are joining the Foundation. That is really great.

I wanted to address the issues of openness and access that you bring up. First, Eclipse projects are required to use the Eclipse infrastructure, even from the the very beginning, and to have all lists, forums, and bug reports out in the open. Repositories can be on either Eclipse.org servers or Github and most new projects are using the latter. Private communication can happen of course, but the largest part of the discussion must be public. We will find a very cold reception from the community if we are not open.

As far as access to the code goes, the only people who will have commit privileges will be people working on the project, which will most likely be only the people on this list. All contributions from other sources will have to pass through a contribution mechanism such as a pull request or bug report, which requires review by committers and the IP team. So, I wouldn't worry about updates to the core from the perspective of outside developers.

Actual project committers might change things in the code contributed from Ptolemy, - 'Ptolemy core' - but that's their job. Most likely we will have our own parts of the project - even our own high-level cores - that we are developing though. For example, I most likely won't be working on any pieces of Triquetrum contributed from Ptolemy because ICE doesn't use them and I don't know how they work; I'll be working on the service layer and any workflow components above it that directly relate to ICE, like our Item and ItemManager infrastructure if I add that as part of the initial contribution.

I was very interested in the last part of your email about the different pieces of Ptolemy and how it works. I think it will be easier to list these components of the initial contribution and others on our Forum, so I started a thread. (I personally have trouble reviewing this kind of thing over email.)

https://www.eclipse.org/forums/index.php/m/1697926/#msg_1697926

This is very exciting!

Jay

On Tue, Jun 9, 2015 at 10:53 AM, Christopher Brooks <cxh@xxxxxxxxxxxxxxxxx> wrote:

June 18th, 17:00 GMT would work for me.  14:00 is a little early, but doable if necessary.

Let me know if you want me to host the call.  We have ReadyTalk, which provided toll free numbers and app sharing if we want it.

A quick introduction about me might be helpful: I'm a software engineer, I've worked on Ptolemy II for many years.  I've met Erwin and Matt in person.  I'm hoping to see the Ptolemy code evolve and be useful in the future, so having an Eclipse project seems like a good idea.  One of my minor concerns is about control over making changes to the core.  The core has not changed that much over the years, but it has evolved.  We've worked hard to keep backward compatibility and have a good test suite, so the changes have been fairly easy to manage.  However, we see Ptolemy II as a research laboratory and sometimes it is necessary to make changes.  We've rarely had branches and strongly encourage the head to always be working.

Edward is the primary author of Ptolemy II and actively involved in the day-to-day development.  Edward is taking a year long sabbatical to work on software, but he has encouraged me to represent him in this effort while keeping him in the loop.  Edward and I both use Eclipse as a software development tool, though we have not done much coding using Eclipse classes.

A few words/a reminder about the triquetrum might be in order.  The triquetrum is the device the Mr. Ptolemy is holding below.  The triquetrum is an instrument for astronomical observations, it predates the astrolabe.  I like the idea of the triquetrum because it reminds me of model-view-controller.  The short name "triq" is fairly unique and could be used for a product.  The name evokes the tricycle, which could be bad, but I don't mind.  Edward owns have http://triquetrum.org.  We are offering this name for use for this project.

I can also host a mailing list if we want that.  We could either use the triquetrum list I have set up or we could start another list.  I see the value in not having a list because it is more apparent who is getting the email.  I have websites that allow me to set up a wiki as well.  I'm guessing we can eventually use something at the Eclipse site, but if we want non-public infrastructure, I can provide it.

I submitted the Eclipse Membership agreement to the University of California Business Contracts for us to join the Eclipse project.  This could take many months to complete.  If necessary, I could join as an individual.

In Ptolemy II, we have the notion of configurations.  The Ptiny configuration is a small, useful configuration with a minimum of third party licenses.  Ptiny includes the kernel, common polymorphic actors, common models of computation and Vergil, the UI.  Vergil is AWT and is what we are looking at replacing with something that is more based on Eclipse.

The Ptolemy II source code is primarily BSD, so contributing it should not be that difficult.  We've done a reasonably good job managing the licenses, the dependencies in the core are fairly straightforward.  A tricky part could be tracking down certain contributors and getting an assignment of copyright.  Fortunately, the core is fairly stable and not that many people have worked on it. 

Hallvard Trætteberg created some OSGi modules that use the core Ptolemy II classes.  One issue that we ran in to is that Ptolemy II uses the class path to add actors, thus we don't always know in advance the class names of all the actors that are to be used in a system.  I believe that this is a bit contrary to how OSGi works, where OSGi expects to know all the dependencies in advance.  There are workarounds to this though.

_Christopher

On 6/9/15 4:56 AM, Erwin de Ley wrote:

Great, thanks Jay.

We can/should indeed discuss the relationship with ICE and SAW (and the DAWN workbench?).
I was thinking to include this under "expectations"/requirements in a second call.

The first call could focus on the global ideas as needed/relevant to have a good project proposal and how that fits in the Science IWG.

But the more info we have from the start, the better!

cheers
erwin

Jay Jay Billings schreef op 09/06/2015 om 13:15:

Erwin,

That time works for me.

It would also be useful to discuss the relationship of this project to ICE since I see ICE as one of its clients. We should also add an item to the agenda to review Sandia's requirements with SAW.

Jay

On Jun 9, 2015 5:08 AM, "Erwin de Ley" <erwin.de.ley@xxxxxxxxxx> wrote:

Dear all,

First of all let me introduce Prof. Edward Lee and Christopher Brooks, from UC Berkeley.
They are the project leads of the Ptolemy II framework, which we've been using for many years now in Passerelle, and which will also be integrated in Triquetrum.
They will also be involved in Triquetrum (in fact they even "donated" the project name!), for which I want to thank them!

Based on the previous email-exchanges, it would seem that we should plan our first call on Thursday, June 18th.
I could be available either around 14h or 17h UTC/GMT with a preference for 17h.
Can you check if this is feasible for (most of) you?

The agenda could be :
- quick intro of everyone
- overview of the project proposal and the three proposed lines of work in triquetrum
- position of the project in the Science IWG, interactions with e.g. dawnsci and viz
- some initial technical and design choices/constraints
- initial code drop

In a second call, after everyone has had some time for reflection, I would like to discuss contributions, roles, expectations.

Please don't hesitate if you have any remarks/suggestions/...

--

Met vriendelijke groeten - Bien à vous - Kind regards

Erwin De Ley

---

Tel. +32 9 335 22 10 fax: +32 9 335 22 19 erwin.de.ley@xxxxxxxxxx

iSencia Belgium Voorhavenlaan 31 bus 11 B-9000 Gent www.isencia.be

 

-- 
Christopher Brooks, PMP                       University of California
Academic Program Manager & Software Engineer  US Mail: 337 Cory Hall
CHESS/iCyPhy/Ptolemy/TerraSwarm               Berkeley, CA 94720-1774
cxh@xxxxxxxxxxxxxxxxx, 707.332.0670           (Office: 545Q Cory)

--

Jay Jay Billings

Oak Ridge National Laboratory

Twitter Handle: @jayjaybillings

-- 
Christopher Brooks, PMP                       University of California
Academic Program Manager & Software Engineer  US Mail: 337 Cory Hall
CHESS/iCyPhy/Ptolemy/TerraSwarm               Berkeley, CA 94720-1774
cxh@xxxxxxxxxxxxxxxxx, 707.332.0670           (Office: 545Q Cory)