(11:12) Jesse_Eichar77: 1) OpenMI
(11:12) Jesse_Eichar77: 2) Summer of Code
(11:12) Jesse_Eichar77: 3) Code Sprint
(11:12) Jesse_Eichar77: anything else?
(11:13) rgould: hard coded source path lookups in .classpath files?
(11:13) Jesse_Eichar77: sure but good luck
(11:13) rgould: (i am against, not for
(11:14) Jesse_Eichar77: ok lets start with that its quick
(11:14) Jesse_Eichar77: 0) .classpath files.
(11:14) Jesse_Eichar77: starting with 0)
(11:14) rgould: if you look at .classpath in libs
(11:14) rgould: lots of hard coded source lookup paths exist
(11:14) rgould: (they have been committed)
(11:15) rgould: I have had my eclipse freak out about this a couple times
(11:15) Jesse_Eichar77: I know.
(11:15) rgould: anything to be done?
(11:15) Jesse_Eichar77: but how to avoid it.
(11:15) Jesse_Eichar77: we can make a source plugin for the GT jars
(11:15) Jesse_Eichar77: and then the paths would be relative
(11:15) rgould: that could work
(11:15) Jesse_Eichar77: but then we'd have to keep it in sync
(11:15) rgould: hmm
(11:16) rgould: sounds worse than just erasing the paths from the classpath
(11:16) chorner: hmm... this again
(11:16) chorner: we tried to add an svn:ignore to the classpath file, but subversion didn't work as expected
(11:16) Jesse_Eichar77: this again.
(11:16) Jesse_Eichar77: What the issue is richard is that I need the source code and so I assign it
(11:16) Jesse_Eichar77: and then accidentally commit it
(11:16) rgould: yea
(11:16) Jesse_Eichar77: we tried to do a svn:ignore so it doesn't commit it but it doesn't work
(11:17) rgould: I wonder if there is a different way to specify where it looks for source code
(11:17) Jesse_Eichar77: plugin is the only way that I know of.
(11:17) rgould: (why doesn't the svn ignore work? because it is already in version control?)
(11:17) Jesse_Eichar77: is there a maven command to bundle the code up and put it in the maven repo?
(11:17) chorner: yes, they didn't consider this case
(11:17) rgould: and we want to leave them in version control?
(11:18) Jesse_Eichar77: no geotools code into the .m2 repo
(11:18) chorner: "bundle the code up" ?
(11:18) Jesse_Eichar77: then we can do the same thing as libs for those files.
(11:18) Jesse_Eichar77: just make a zip or jar containing just the source for a maven project.
(11:18) chorner: oh... a src jar?
(11:18) Jesse_Eichar77: yes
(11:18) chorner: we should be able to do that
(11:19) Jesse_Eichar77: lets look into it.
(11:19) rgould: what about a classpath variable?
(11:19) Jesse_Eichar77: good one
(11:19) rgould: if you look, there is already a JRE_SRC andJUNIT_SRC_HOME
(11:19) Jesse_Eichar77: I like that.
(11:19) Jesse_Eichar77: Yes lets do that instead.
(11:19) Jesse_Eichar77: Cory?
(11:20) Jesse_Eichar77: sorry times up we have a tight schedule today
(11:20) Jesse_Eichar77: 1) OpenMI
(11:20) Jesse_Eichar77: moovida where do you want to start?
(11:20) moovida:
(11:20) moovida: I want to start to code
(11:20) moovida: I just want to know if the way is ok for the udig community
(11:21) * jgarnett has joined #udig
(11:21) Jesse_Eichar77: In short yes
(11:21) moovida: we need to use the udig for our modelling
(11:21) moovida: that is enough for today
(11:21) moovida: the yes I mean
(11:21) Jesse_Eichar77: Just a few comments:
(11:21) Jesse_Eichar77: 1) It should be dependent on Catalog but not Project
(11:21) moovida: since there will be lot we will have to do and we will expose the steps to the community
(11:22) moovida: (first enumerate or explain step by step?)
(11:22) jgarnett: hello - sorry I am late (on the bright side I have looked at OpenMI now)
(11:22) Jesse_Eichar77: 2) There is a lot of duplication between it and GeoAPI so we eventually (not necessarily immediately) want to create adapters between the OpenMI and GeoAPI interfaces
(11:22) Jesse_Eichar77: ok go on moovida
(11:23) moovida: ok, to 1
(11:23) moovida: depend on catalog? What do you mean by that?
(11:23) * Jesse_Eichar77 changes topic to 'Agenda: 0) .classpath files 1) OpenMI 2) Summer of Code 3) Code Sprint'
(11:23) moovida: not project = not map?
(11:24) moovida: however I think we are thinking the same
(11:24) moovida: I will need to chose maps when I do modelling
(11:24) moovida: so were to chose from? The catalog
(11:24) Jesse_Eichar77: It would be nice that it doesn't require the project plugins in order to create the operations.
(11:24) Jesse_Eichar77: all views and editors in eclipse provide a "selection"
(11:24) moovida: ahhhh... I'm wrong
(11:25) moovida: now I get it
(11:25) moovida: code dependency
(11:25) Jesse_Eichar77: yes
(11:25) Jesse_Eichar77: that's right
(11:25) moovida: alright then, should be ok
(11:25) Jesse_Eichar77: cool.
(11:25) Jesse_Eichar77: Do you want to lay out your plan?
(11:26) moovida: what do you mean?
(11:26) moovida: ok
(11:26) Jesse_Eichar77: do you have a plan for how you are going to integrate it. I saw a Java project that has openMI.
(11:27) Jesse_Eichar77: Are you going to borrow code from that?
(11:27) Jesse_Eichar77: what technologies are you going to use...
(11:27) moovida: you mean the one I attached?
(11:27) Jesse_Eichar77: and so on.
(11:27) Jesse_Eichar77: There was one on source forge.
(11:27) Jesse_Eichar77: that I saw
(11:27) moovida: alright, that should be the same I have
(11:27) moovida: so:
(11:28) moovida: what I would like to do (design will start in twoo weeks)
(11:28) moovida: is to use the openmi chaining
(11:28) moovida: but put no JNI in the core plugins
(11:29) Jesse_Eichar77: (yay)
(11:29) moovida: (you will tell me where to put that one part)
(11:29) Jesse_Eichar77: I'll address it after you finish
(11:29) moovida: we want to convert all the JGrass applications with openmi interfaces
(11:29) jgarnett: OpenMiGrass
(11:29) jgarnett: :-D
(11:30) moovida: yes, kinda like that. In order to be able to use them in timedependent environments and not
(11:30) moovida: also they will be linked in a gef environment
(11:31) moovida: instead of having tousand toolbars
(11:31) moovida: at the begin I wanted to create the engine
(11:31) moovida: and use the GEF stuff also for doing simple operations
(11:31) moovida: example#:
(11:31) moovida: I want to create a map of flowdirections from a dtm
(11:32) moovida: 1) I take a openmigrassudig module
(11:32) moovida: 2) I link a map datastore to it
(11:32) moovida: 3) I link an output map datastore to it
(11:32) moovida: and launch the thing
(11:32) moovida: what happens is:
(11:32) moovida: the links are checked for consistency
(11:33) moovida: every object, model or whatever will have parameter arguments for a proper initialization
(11:33) moovida: which will have to be set (prior to launch)
(11:34) moovida: the environment will be saved and can be reopened and reexecuted
(11:34) moovida: so you can supply modeling schenarios
(11:34) * moovida beeing to confusing?
(11:35) Jesse_Eichar77: nope
(11:35) Jesse_Eichar77: makes sense to me.
(11:36) moovida: good! The same way many models can be chained and what I hoped is that every single model can be launched also standalong
(11:36) moovida: I mean
(11:36) Jesse_Eichar77: so you could pull a resource from the catalog or a layer from a map for example if it can
(11:36) Jesse_Eichar77: adapt to the correct paramter for that type then it will be adapted.
(11:36) Jesse_Eichar77: (just brainstorming)
(11:37) Jesse_Eichar77: is that kind of what you are thinking w.r.t the UI component?
(11:37) moovida: that is something I'm not used to
(11:37) moovida: let's talk with an example
(11:37) moovida: I do one:
(11:38) moovida: because I'm hoping that with the same interfaces we are able also to create buttons
(11:38) moovida: that execute those models
(11:38) moovida: again flowdirections:
(11:38) moovida: I have a button on the toolbar
(11:38) moovida: I click it and a gui pops up
(11:39) moovida: which two textfields and two browse buttons
(11:39) moovida: I will have to browse for raster maps and then execute
(11:39) moovida: how do I do for the browse?
(11:40) moovida: so that a user doesn't take a feature map?
(11:40) Jesse_Eichar77: that is where I was thinking Drag and Drop would be useful.
(11:40) Jesse_Eichar77: But you could just browse the catalog.
(11:40) Jesse_Eichar77: or possibly the selected map.
(11:40) moovida: drag and drob is nice for visualization
(11:40) Jesse_Eichar77: I agree.
(11:40) Jesse_Eichar77: scenario:
(11:41) moovida: but when you have a textfield, no one will drag a file into it
(11:41) moovida: they expact a browse button
(11:41) moovida: go
(11:41) Jesse_Eichar77: (this is another vision)
(11:41) Jesse_Eichar77: 1) You press the create work flow button or menu item
(11:41) Jesse_Eichar77: 2) A editor opens
(11:42) Jesse_Eichar77: (editor would be for workflow
(11:42) Jesse_Eichar77: - Editor would have a toolbox containing all the operations
(11:42) Jesse_Eichar77: - Tools for maybe moving items, possibly linking items
(11:42) Jesse_Eichar77: 3) drag a operation on to editor
(11:42) Jesse_Eichar77: - it will add the operation to the work flow
(11:43) Jesse_Eichar77: 4) (two ways to do this
(11:43) Jesse_Eichar77: - Drag a resource from the catalog on to the editor
(11:43) Jesse_Eichar77: - or drag a layer from a map (in projects view) onto the editor
(11:43) Jesse_Eichar77: 5) select link tool
(11:44) Jesse_Eichar77: 6) draw link from resource to operation
(11:44) moovida: in 4 you mean drag a resource from the catalog on the operation to define it as input or output?
(11:44) Jesse_Eichar77: - maybe a dialog will open asking if it is an input or output
(11:44) moovida: ok, sorry, that was it
(11:44) Jesse_Eichar77: - it also may ask which input it is (if there are multiple inputs)
(11:45) Jesse_Eichar77: 7) do the same for the output
(11:45) Jesse_Eichar77: another way is drop resource directly on operation
(11:45) Jesse_Eichar77: and the menu would open asking if it is an input or output.
(11:45) Jesse_Eichar77: Just ideas.
(11:45) Jesse_Eichar77: what do you think?
(11:45) moovida: the first part is what I was talking about when talking about GEF, the second part is a very nice part that I like!!
(11:46) moovida: drag the resources into the editor and then link! Great!
(11:46) Jesse_Eichar77: cool.
(11:46) moovida: what is not resource
(11:46) moovida: can be put into the palette
(11:46) moovida: operations
(11:47) moovida: values
(11:47) Jesse_Eichar77: that's right.
(11:47) moovida: database connections for dischare data (example)
(11:47) moovida: etc. etc
(11:47) moovida: dischare = discharge
(11:47) Jesse_Eichar77: to clarify about the values for example
(11:47) Jesse_Eichar77: what exactly would be on the palette again?
(11:48) Jesse_Eichar77: a DB Connection?
(11:48) Jesse_Eichar77: then you'd drag it to the editor
(11:48) moovida: values = thresholds, particular parameters of the model
(11:48) Jesse_Eichar77: and fill in the exact values?
(11:48) moovida: alright
(11:48) Jesse_Eichar77: it was a question. It that kind of what you were thinking?
(11:49) moovida: in the palette we have the operations and everything not in the catalog
(11:49) moovida: so a model could need value, and the palette will have a scalar object
(11:49) Jesse_Eichar77: I understand that. But we wouldn't have "concrete" values would we? IE the value 14.
(11:50) moovida: a model will need a database connection to get the meterological data
(11:50) jgarnett: Hey guys 10 mins left ...
(11:50) moovida: so you drag this non -spatial object into it
(11:50) jgarnett: (The last two should go quick )
(11:50) moovida: and in a properties tab define the connection
(11:50) moovida: Jody ok, sorry
(11:50) Jesse_Eichar77: ok that was what I was thinking too. We're talking about the same thing.
(11:50) Jesse_Eichar77: Good.
(11:50) moovida: GREAT!
(11:51) Jesse_Eichar77: Do you have any other questions you need resolved right now moovida?
(11:51) Jesse_Eichar77: (this is your chance)
(11:51) moovida: No, I have a lot to do on it before it comes to problems
(11:51) moovida:
(11:51) Jesse_Eichar77:
(11:51) Jesse_Eichar77: ok sounds good.
(11:51) moovida: I mean udig related ones
(11:51) Jesse_Eichar77: Next topic
(11:51) Jesse_Eichar77: 2) Summer of Code
(11:51) Jesse_Eichar77: Cory? Jody?
(11:52) chorner: oh
(11:52) chorner: yes, see e-mail – we need ideas for the summer of code
(11:52) moovida: until when is time to contribute ideas?
(11:53) chorner: when can go on indefinitely
(11:53) chorner: *we
(11:53) chorner: but google will starting looking on monday
(11:53) moovida: alright
(11:53) chorner: students will start looking a few weeks after that
(11:54) moovida: so if we of the JGrass part add a topic... what will happen next?
(11:54) moovida: what will we have to do?
(11:54) chorner: we will need a mentor to lead the student, if we get a student
(11:54) moovida: how much to write?
(11:54) chorner: realistically, we'll probably have a few mentors helping out
(11:55) chorner: someone on the uDig side, some on the JGrass side
(11:55) chorner: whoever is willing to help students out with jgrass (a few hours per week) can be an mentor
(11:56) moovida: but the students?
(11:56) moovida: do we have to supply them?
(11:56) chorner: no
(11:56) chorner: the students will come from all over
(11:56) chorner: and google will pay them
(11:57) chorner: we will probably just get one
(11:57) moovida: so "just" write down a project of three months and hope that someone will chose it?
(11:57) chorner: yes
(11:57) moovida: and that google will chose it?
(11:57) chorner: you can also nudge some students and say "hey! look! isn't this cool"
(11:58) chorner: google doesn't choose a particular idea – they will just look to see that we have good ideas floating around
(11:58) moovida: so if I find someone that does it, get the project?
(11:58) moovida: or is google filtering first?
(11:58) chorner: there is a complicated process which determines who gets a student
(11:58) chorner: as mentors we will have a small say
(11:58) chorner: in who is selected
(11:59) moovida: alright, I got it
(11:59) jgarnett: Um sanity check here
(11:59) chorner: http://udig.refractions.net/confluence/display/HACK/Summer+of+Code+2007
(11:59) jgarnett: what we really need is a list of potential mentors
(12:00) chorner: i think that is jesse, jody, richard, me
(12:00) jgarnett: Based on our experience last time I would like to see more information on each idea (so students get off on the right foot).
(12:00) chorner: and a jgrass folk or two
(12:00) jgarnett: moovida ?
(12:01) moovida: yes, I would candidate as mentor, that was obvious for me if I contribute an idea
(12:01) jgarnett: For reference here is the page for GeoServer (GEOSDEV/GeoServer+Summer+of+Code+Ideas)
(12:01) jgarnett: bah - http://docs.codehaus.org/display/GEOSDEV/GeoServer+Summer+of+Code+Ideas
(12:01) jgarnett: actually we had lots of ideas with no mentor last year
(12:01) jgarnett: there is some responsibility in being a mentor (and some money)
(12:02) moovida: I made mentor for different tesises... kinda the same?
(12:02) chorner: potential mentors is a new thing
(12:02) chorner: i think they are pre-screening
(12:02) * FrankW has joined #udig
(12:03) jgarnett: Hi Frank - we are talking about getting a list of potential mentors for you
(12:03) chorner: k... potential mentors need to put up their name and google account
(12:03) jgarnett: would Friday be okay?
(12:03) chorner: it will be on our wiki page jody
(12:03) FrankW: Friday is fine, with google account names.
(12:03) jgarnett: But frank needs to collapse them all into one list ...
(12:04) jgarnett: http://wiki.osgeo.org/index.php/2007_Google_SoC_Application
(12:04) FrankW: Right - they all have to be listed in the application.
(12:04) moovida: so a mentor would be related to a certain proposal, right?
(12:04) chorner: no
(12:04) jgarnett: Nope - just are you available if a student needs mentoring
(12:05) moovida: does that make sense? What if I can't help him?
(12:05) jgarnett: we can assume that for the uDig community we will match up the right two people together
(12:05) jgarnett: (so you would be our top pick for OpenMI GEF editor)
(12:05) jgarnett: Do I have this right Frank?
(12:05) moovida: alright I see, it is kinda interchangable?
(12:05) FrankW: jgarnett: yes I think so.
(12:06) moovida: or better, a net of mentors?
(12:06) jgarnett: yes we need to collect two things
(12:06) jgarnett: 1) list of mentors
(12:06) jgarnett: 2) list of ideas
(12:06) chorner: okay.. i think we've beaten this one to death
(12:06) moovida: sorry if I bother, but I have to be sure guys
(12:06) jgarnett: The list of mentors is more important needed for the weekend; the ideas can be ready end of day Monday.
(12:07) moovida: for example I'm no feature guy
(12:07) moovida: what if I get one about features?
(12:07) jgarnett: yep. To be blunt you will have a chance to say no
(12:07) jgarnett: so dont worry about getting matched up with something you know nothing about.
(12:07) jgarnett: :-D
(12:07) moovida: :S
(12:07) moovida: alright, I trust
(12:08) jgarnett: 3) Code Sprint
(12:08) jgarnett: We have a day set aside at the end of the FOSS4G conference
(12:08) moovida: joyful, joyful
(12:08) jgarnett: It sounds like from email that we would like to get a hackfest in gear
(12:08) moovida: will the long codesprint be makable?
(12:08) jgarnett: (part of this is so you can enjoy the conference)
(12:09) jgarnett: Depends how big Jesse's new place is
(12:09) jgarnett: (that was a joke)
(12:09) moovida:
(12:09) chorner: i think there is interest in hacking more the monday after
(12:09) chorner: we can get a sprint on friday
(12:09) chorner: but some will want more
(12:10) jgarnett: Well we have paid for facilities on the friday - and uDig can get in on that.
(12:10) moovida: yes, one day is too short!!
(12:10) jgarnett: For more we will need to organize ourselves.
(12:10) chorner: i think sat/sun is a bad idea
(12:10) Jesse_Eichar77: we can find something. I'm not too worried.
(12:10) jgarnett: (sigh - we are all such hackers - don't know why they put up with us)
(12:10) moovida: which means Victorians organise for the whole world?
(12:10) chorner: after 5 days of conference, people will want to go surfing
(12:10) chorner: or to the pub
(12:11) * moovida wants to go to the pub...
(12:11) jgarnett: moovida I was including you as a hacker - ie 'show me the code'
(12:11) Jesse_Eichar77:
(12:11) moovida:
(12:11) jgarnett: So lets get down to biz ...
(12:11) chorner: let's leave it open then: we will have a uDig code sprint on friday
(12:11) jgarnett: moovida it makes a difference to you - having a code sprint.
(12:11) chorner: and will do some more the next week
(12:11) * sfarber has joined #udig
(12:11) jgarnett: and if you are going to fly out here you would like more then a single day hacking?
(12:12) moovida: yes, that would make my presence there much much better
(12:12) jgarnett: (Worst case Jesse and I take a day's holiday on Monday and I have wireless at my house)
(12:12) jgarnett: Jesse are we confident we can arrange something? Enough for moovida to make plans ...
(12:13) * moovida says WOW (not like in vista... ut)
(12:13) Jesse_Eichar77: Definately
(12:13) jgarnett: sweet.
(12:13) moovida: jody, thanks, I would need to be sure
(12:13) moovida: Great!
(12:13) moovida: amazing
(12:13) moovida:
(12:13) jgarnett: Okay I will sum up this kind of stuff on the sprint page; and I am going to try and contact some of the TOPP developers that have been to sprints before.
(12:13) Jesse_Eichar77: Should be fun
(12:13) moovida: I'm sure it will
(12:13) jgarnett: It will be very cool.
(12:14) jgarnett: Um but guys; to have a successful sprint
(12:14) jgarnett: you need to choose something that can be done; and is useful...
(12:14) jgarnett: aka no random hacking.
(12:14) moovida: never been at a sprint
(12:14) jgarnett: Something to think about over email; and it may be you see something at the conference and decide what to do on the spot.
(12:14) FrankW: Well, the alternative is a less structured hack-a-thon.
(12:14) moovida: but I guess soemone will have to coordinate
(12:15) moovida: hack-a-ton?
(12:15) jgarnett: FrankW++ I was hoping each developer community woudl self coordinate - but a lot of ideas cut across project groups.
(12:15) jgarnett: It may be a nicer "message" for OSGeo to promote collaboration across developer communities.
(12:15) FrankW: jgarnett: With different groups in the same place, it seems like we should contemplate cross-project efforts.
(12:16) jgarnett: I will write it on the page; this is going to be one wild crazy week ...
(12:16) jgarnett: ... I do have to warn you that the GeoTools developers last time ignored the conference and just coded most of the week
(12:16) FrankW: That is has been our inspiration!
(12:16) jgarnett: ... I would like to promiss of a sprint in order to let them be social and meet others.
(12:17) moovida: I can immagine...
(12:17) jgarnett: Cool ... meeting is over time. Shall we kick out the lights?
(12:17) Jesse_Eichar77: sounds good
(12:17) * FrankW fades...
(12:17) moovida: good for me
(12:17) * FrankW has left #udig ("Leaving")
(12:17) Jesse_Eichar77: thanks for everyone coming by
(12:17) jgarnett: thanks
(12:17) moovida: thanks guys.. ciao