Nov-18-2013-WPS-IRC-Meeting

[17:08] -asimov.freenode.net- * Looking up your hostname...
[17:08] -asimov.freenode.net- * Checking Ident
[17:08] -asimov.freenode.net- * Couldn't look up your hostname
[17:08] -asimov.freenode.net- * No Ident response
[17:09] -asimov.freenode.net- * Looking up your hostname...
[17:09] -asimov.freenode.net- * Checking Ident
[17:09] -asimov.freenode.net- * Couldn't look up your hostname
[17:09] -asimov.freenode.net- * No Ident response
[17:10] -asimov.freenode.net- * Looking up your hostname...
[17:10] -asimov.freenode.net- * Checking Ident
[17:10] -asimov.freenode.net- * Couldn't look up your hostname
[17:10] -asimov.freenode.net- * No Ident response
[17:10] * LogBot_52n (~PircBot@80.156.182.234) has joined #52north
[17:10] * Topic is 'NOW: WPS IRC meeting | 52°North Discussion Group | Geospatial Open Source Software, geo-hacking, lunch planning. | About this channel: http://wiki.52north.org/bin/view/Documentation/IRC'
[17:10] * Set by daniel_52n!~Daniel@193.175.4.220 on Mon Nov 18 16:59:58 CET 2013
[17:10] * bot_52n sets mode +o LogBot_52n
[17:10] <benjamin_52n> there it is
[17:10] <benjamin_52n> ok, it seems, there are only people from the geoprocessing dev-list
[17:11] <daniel_52n> do we have an agenda?
[17:12] <benjamin_52n> sort of
[17:13] <benjamin_52n> As I wrote in the email, I want to use these meetings to create a roadmap for the WPS
[17:14] <joejoe> may be that I've to leave early - sorry - hope @matthias-mueller can give me an non-digital report tomorrow ;)
[17:14] <daniel_52n> ok. so how many of the active WPS developers are represented here? we have CIDA, TUD, 52N.
[17:14] <IvanSuftin-CIDA> What is TUD?
[17:14] <daniel_52n> are there any TU Dresden :-)
[17:14] <benjamin_52n> actually thats a good question
[17:14] <joejoe> Technische Universität Dresden
[17:15] <daniel_52n> An as far as I can tell that covers the contributors > https://github.com/52North/WPS/graphs/contributors
[17:15] <joejoe> the Geoinformation Systems lab to be precise
[17:15] <daniel_52n> so that's a good thing - the organisations who contribute are in the meeting!
[17:16] * carsten_52n (~Thunderbi@p3EE205E2.dip0.t-ipconnect.de) Quit (Quit: carsten_52n)
[17:16] <jiwalker> yea!
[17:16] <daniel_52n> benjamin_52n: can you quickly outline why you'd like to have a roadmap?
[17:16] <benjamin_52n> sure
[17:16] <benjamin_52n> roadmaps are essential for every project, i guess
[17:17] <benjamin_52n> so, i'd like to know, what people are working on right now in terms of WPS
[17:18] <jiwalker> we are doing some R processes over here at the moment
[17:18] <jiwalker> still maintaining the GeoDataPortal processes as well
[17:18] <matthias-mueller> I have not something on my task list right now; my last commit was to the SVN I guess
[17:19] <benjamin_52n> ok, cool. but CIDA is not working on something in the WPS code right now?
[17:19] <daniel_52n> I'm not working on anything in particular, but can report that christian is working on streaming in his theses (w/ Jordan Read)
[17:20] <daniel_52n> correction: thesis.
[17:20] <jiwalker> no, the Database addition Ivan did was the last focused effort we had
[17:20] <IvanSuftin-CIDA> We've recently completed a few commits back to the WPS codebase. Personally I am not actively working on anything WPS related at the moment.
[17:20] <benjamin_52n> ok
[17:20] <daniel_52n> so no work on WPS core.
[17:20] <benjamin_52n> i just did some minor stuff regarding compliance tests
[17:21] <joejoe> me neither - busy with my PhD right now and most likely the core WPS will be the major tool used (but not necessarily developed further)
[17:22] <daniel_52n> so we've all come users :-)
[17:22] <daniel_52n> become users!
[17:22] <matthias-mueller> How do you want to drive the roadmap, benjamin? By a wish list or something? Or are there some stability issues left?
[17:22] <benjamin_52n> ok, so we have a document that was created for a code sprint (mainly involving CIDA and 52n)
[17:22] <benjamin_52n> https://docs.google.com/document/d/1pXzrt9a1QlZM5Dy7FxrzQp8UV1LkSShzkEL1bD9-mT8/edit
[17:23] <benjamin_52n> should be open for everyone
[17:23] <joejoe> beside the core developments I plan to integrate WPS processes into Scientific Workflow engines (most likely Vistrails)
[17:23] <daniel_52n> these are concrete tasks, which is good.
[17:24] <daniel_52n> do we need a more high-level roadmap? or is that going to be (a) not helpful and/or (b) soon outdated anyway?
[17:24] <jiwalker> I know there has been some work with vistrails calling GDP WPS processes
[17:24] <matthias-mueller> can I add something to the document shall we have a new one?
[17:24] <IvanSuftin-CIDA> I thought #2 was closed?
[17:25] <daniel_52n> to me a main issue with a roadmap is that all our work is project driven, and we rarely have funds to just "improve the codebase"... the code sprint proofed to be very useful to that.
[17:25] <benjamin_52n> matthias. please go ahead and add stuff
[17:25] <daniel_52n> so there is a conflict between what is on the roadmap and what is likely to be implemented.
[17:26] <benjamin_52n> daniel. you are right.
[17:26] <joejoe> @jiwalker: gdp? sorry - not my brightest day today
[17:26] <daniel_52n> benjamin_52n: if I understood you correctly you would capture long-term goals on the roadmap, things that are important but not necessarily driven by any projects today?
[17:26] <IvanSuftin-CIDA> joejoe: Geo Data Portal. We are Center for Integrated Data Analytics. A group within the US Geological Survey
[17:26] <IvanSuftin-CIDA> Geo Data Portal is a project which uses the WPS framework for processing
[17:26] <jiwalker> http://cida.usgs.gov/gdp/process/WebProcessingService?Service=WPS&Request=GetCapabilities
[17:27] <benjamin_52n> daniel, yes, i'd like to define a short term, medium and long term roadmap
[17:27] <joejoe> thanks
[17:28] <IvanSuftin-CIDA> benjamin, do you have an idea for time in mind? What is short term? What is long term? etc
[17:28] <daniel_52n> ok. let's define what these levels mean?
[17:29] <benjamin_52n> yes, so i thought about this with regard to time and also workload
[17:29] <benjamin_52n> short term means, the tasks can be done rather quickly with not much effort
[17:30] <daniel_52n> long term = things to consider when looking for new funding? or which require funding? larger tasks to be covered in research project proposals (e.g. versioned algorithms)
[17:32] <matthias-mueller> What about differentiating between features and maintenance?
[17:32] <matthias-mueller> Features are developed by projects, maintenance is a contribution from the developers beyond a particular project
[17:32] <daniel_52n> does feature imply a certain time scale?
[17:32] <matthias-mueller> Typical example is stability.
[17:33] <matthias-mueller> Not really
[17:33] <daniel_52n> or could a new feature be short term, because it is quickly done?
[17:33] <benjamin_52n> i think so
[17:33] <matthias-mueller> I think feature involves less developers
[17:34] <daniel_52n> while we're collecting "tags" for roadmap items, I think it would be important to capture the interested parties.
[17:34] <matthias-mueller> +1
[17:35] <benjamin_52n> i guess 52north is interested
[17:35] <daniel_52n> ok, so "interested" is not a really good word.
[17:36] <matthias-mueller> concerned?
[17:36] <matthias-mueller> contributing?
[17:36] <IvanSuftin-CIDA> Concerned but not sure about contributing due to time constraints
[17:37] <IvanSuftin-CIDA> Unless it specifically coincides with features we require in the immediate future
[17:37] <daniel_52n> if a feature is a requirement by, say, CIDA, they will implement that before fixing unrelated bugs. correct?
[17:38] <jiwalker> yes, we contribute back the features or bugfixes we run into during the course of our projects
[17:39] <benjamin_52n> ivan, jordan, ok, i guess, it is not really possible to plan this, is it?
[17:39] <daniel_52n> and I think the items on the general roadmap do not always fit under (yours or anyones) "course of our projects".
[17:39] <matthias-mueller> 52N WPS has been driven a lot by feature contributions in the past
[17:40] <jiwalker> we would like to see a roadmap that prioritizes some of the core work I think
[17:40] <matthias-mueller> while it is hard to plan for features it might be easier to plan for bugfixes / refactorings?
[17:40] <joejoe> sorry gents - need to go now ... @usgs: the vistrails / WPS combination is very interesting - hope you don't mind if I get in touch with you again in the future? don't know whether you're directly involved or not ;)
[17:40] <jiwalker> yeah, I'll ask for more info on it, I've just heard a bit
[17:40] <benjamin_52n> by joejoe
[17:41] <daniel_52n> I see one more aspect of roadmap: Prioritization!
[17:41] <joejoe> ok cool. bye all!
[17:41] <daniel_52n> again: are things with a high priority always short term?
[17:41] <IvanSuftin-CIDA> matthias, I think that planning with multiple groups with different time allotments, bug fixes and smaller features are a lot easier to hit
[17:41] <jiwalker> one thing that we have been wanting on our end for a while is job management
[17:42] * joejoe (~joejoe@141.30.100.136) Quit (Quit: Leaving)
[17:42] <benjamin_52n> jordan, i'd like to have this, too
[17:42] <jiwalker> we haven't gotten around to tackling it, but it is something that would be cool if we knew where it fit on a roadmap
[17:42] <IvanSuftin-CIDA> daniel_52n: priority and time don't always coincide. time is more tightly tied in to complexity
[17:42] <daniel_52n> yes.
[17:42] <IvanSuftin-CIDA> a job management console (what is running, when it started, cancel job y/n) would be amazing
[17:43] <matthias-mueller> job management certainly ha a relation to WPS 2.0
[17:43] <daniel_52n> so the point that I'd like to make is that the roadmap could serve many purposes.
[17:43] * bot_52n (~PircBot@geoprocessing.demo.52north.org) Quit (Ping timeout: 265 seconds)
[17:44] * bot_52n (~PircBot@geoprocessing.demo.52north.org) has joined #52north
[17:44] * ChanServ sets mode +o bot_52n
[17:44] <daniel_52n> IvanSuftin-CIDA: the job management console would fit perfectly into the new UI (GSoC project)
[17:45] <benjamin_52n> ok, so you CIDA guys (and all), would you have resources to work together on this feature in the near future?
[17:45] <matthias-mueller> what would be the implications in the WPS framework?
[17:45] <daniel_52n> imho one purpose for the roadmap: collect ideas and realize when there is critical mass (= funding in between the contributors) for a feature because we have noted ideas and requirements.
[17:45] <IvanSuftin-CIDA> Also a fully featured control API for the console apart from the UI would be required
[17:46] <matthias-mueller> right
[17:46] <matthias-mueller> havin only the interfaces adjusted is probably not enough. there are processes that by nature cannot support cancelling.
[17:47] <IvanSuftin-CIDA> benjamin_52n: I think this would not be something we'd be able to handle in the short term. This would have to go into our planning process here in which we are at least a half year out on
[17:47] <benjamin_52n> ok
[17:47] <jiwalker> Ivan and I are on a project until Feb, and after that there is a possibility that we could allocate some resources
[17:48] <jiwalker> but we would have to sell it to the project and/or management
[17:48] <IvanSuftin-CIDA> The management console idea is something we've been sorely lacking and one that one of our project managers has been calling for over and over. So that would be the highest likeliness that we can work on once we do get time to do so
[17:50] <jiwalker> mostly in terms of process queuing based on resource constraints
[17:50] <daniel_52n> and for that it would be great if you can work based on the administrative backend.
[17:50] <daniel_52n> benjamin_52n: that is not yet merged, is it?
[17:51] <benjamin_52n> ok, sounds good, i think this could be something we could work on the in the course of a WPS 2.0 reference implementation on our side
[17:51] <IvanSuftin-CIDA> We also were bouncing around the idea of scoring jobs in order to deduce a way to not only prioritize jobs but also load balance as well as give a rough estimation to the client how long it may take
[17:51] <daniel_52n> one important task (long term) for me is a better modularization and configuration framework, i.e. it cannot be that I hardcode WPS4R features in the webapp for example
[17:51] <benjamin_52n> daniel_52n, no it is not merged
[17:53] <daniel_52n> ok, so if the process management is the biggest issue, should we focus on that for the short term?
[17:53] <benjamin_52n> daniel_52n, yes, this should be checked, when the admin app is merged...
[17:53] <daniel_52n> one more use of a roadmap: know what shoud not be worked on.
[17:54] <daniel_52n> I should have suggested that at the beginning, but I think it is a very good idea to keep these meetings under one hour and go for more regular than for doing everything at one go.
[17:55] <benjamin_52n> i think the process stuff is at least medium term
[17:55] <benjamin_52n> daniel_52n, you are right
[17:55] <daniel_52n> we are nearing that hour now... should we collect next steps?
[17:55] <benjamin_52n> yes
[17:55] <daniel_52n> I'd like to hear some ideas for a "tool" to manage the roadmap, though.
[17:55] <IvanSuftin-CIDA> Benjamin, are you guys looking to integrate Spring into the WPS framework in coming releases?
[17:56] <jiwalker> I've seen roadmap sort of things done in trello
[17:56] <jiwalker> allowing votes and comments
[17:56] <daniel_52n> I though GitHub issues at first, but maybe it is better to distinguish bug reporting and task management from higher level planning.
[17:56] <benjamin_52n> Ivan, I'd like to have that, but we have no concrete plans yet
[17:57] <benjamin_52n> ok, lets try this trello thing
[17:57] <daniel_52n> jiwalker: trello sounds good!
[17:57] <benjamin_52n> Ivan was talking about this a year ago already :)
[17:58] <daniel_52n> to me it is just important that we note down somewhere (wiki, contribute.md) what goes where (is a feature request a GitHub issue, or a trello card?)
[17:58] <benjamin_52n> i will try to move the things from the google doc to trello
[17:58] <benjamin_52n> daniel, yes, thats important, too
[17:59] <jiwalker> maybe to keep it simple using the github milestones?
[17:59] <daniel_52n> my suggestion: GitHub is public for feature requests and bug reports, and used for concrete implementation tasks. Trello for roadmap (also public if that is possible with read access only) ??
[17:59] <jiwalker> though I do like trello
[17:59] <daniel_52n> jiwalker: I'm just concerned that we mix too many things in the GitHub issues.
[18:00] <daniel_52n> but we could try it, too...
[18:01] <benjamin_52n> ok, so maybe we should discuss this next time
[18:01] <daniel_52n> that's the experienced OGC WG leader talking !
[18:01] <benjamin_52n> and i can prepare something
[18:02] <jiwalker> sounds good, thanks for setting this up
[18:02] <benjamin_52n> yes, i just had SWG telecon before this meeting, so i'm in the flow
[18:02] <benjamin_52n> so, when should we met next?
[18:02] <daniel_52n> sounds good to me.
[18:02] <benjamin_52n> in one week?
[18:02] <jiwalker> works for me
[18:03] <daniel_52n> benjamin_52n: will you also transfer the ideas from the log to the Google Doc/Trello/GitHub issues ? though you should not work on this too much before we decide what we would like to use
[18:04] <benjamin_52n> can do.
[18:04] <benjamin_52n> ok, anything else?
[18:06] <daniel_52n> one week is good.
[18:06] <jiwalker> have a good week.
[18:06] <daniel_52n> thanks, you too!
[18:06] <matthias-mueller> bye
[18:06] <benjamin_52n> ok, then thanks for joining, have a good one, stick around, if you like ;)
[18:06] * daniel_52n changes topic to '52°North Discussion Group | Geospatial Open Source Software, geo-hacking, lunch planning. | About this channel: http://wiki.52north.org/bin/view/Documentation/IRC'
[18:07] * matthias-mueller (~chatzilla@mlrn10.mw.tu-dresden.de) has left #52north