| 02:23:41 | kmq: | good day everyone, I am karim, your friendly summer of code student for ajax administration improvements |
| 02:24:28 | kmq: | Testbaudson: just to make sure mail didn't get lost: you recieved my mail from last week ? |
| 02:28:48 | Testbaudson: | hi kmq |
| 02:28:55 | Testbaudson: | thanks for dropping in |
| 02:29:02 | Testbaudson: | yes I got your mail |
| 02:29:46 | Testbaudson: | I wanted to get back to you |
| 02:29:54 | Testbaudson: | however, I didn't find the time |
| 02:30:00 | kmq: | no hurry |
| 02:30:02 | Testbaudson: | anything urgent on your behalf? |
| 02:30:52 | kmq: | no |
| 02:31:13 | Testbaudson: | I'll be on holidays next week |
| 02:31:23 | Testbaudson: | but I'll contact you this week for sure |
| 02:31:28 | Testbaudson: | talk to you soon |
| 02:32:31 | kmq: | ok, excellent |
| 05:47:46 | : | * seven_ might be a few minutes late for the CSW Chat. |
| 05:47:47 | seven_: | CU |
| 05:52:47 | Testbaudson: | Hi andreas |
| 05:53:01 | Testbaudson: | I copied Mifan's proposal to the wiki |
| 05:53:03 | Testbaudson: | http://www.mapbender.org/GsoC2009_CSW_proposal |
| 05:53:04 | sigq: | Title: GsoC2009 CSW proposal - MapbenderWiki ( at www.mapbender.org ) |
| 05:53:28 | Testbaudson: | will have to contact the OSGeo admin about your mentor status |
| 05:57:03 | andreas: | Hi Christoph, hello everyone |
| 05:57:20 | Testbaudson: | Hi Andreas |
| 05:57:26 | Testbaudson: | Arnulf will be a little late |
| 05:57:36 | Testbaudson: | I will try and contact Armin Retterath via phone |
| 05:57:44 | Testbaudson: | sorry for the delay |
| 05:58:03 | andreas: | Christoph, o.k. I will have a look at the wiki later |
| 05:58:40 | andreas: | ... and I will be glad, if you care about my mentor status - thanks |
| 05:59:33 | andreas: | I will stay tuned, till A. Retterath and Arnulf will join us |
| 06:00:45 | Testbaudson: | Hi tbaschetti |
| 06:00:58 | Testbaudson: | there will be a delay by a few minutes |
| 06:01:00 | tbaschetti: | hello testbaudson |
| 06:01:18 | tbaschetti: | no problem, i'll get another espresso meanwhile... |
| 06:01:57 | andreas: | hello Thomas, nice to "see" you afer FOSSGIS |
| 06:03:21 | Testbaudson: | Armin Retterath's phone is still busy... |
| 06:03:31 | Testbaudson: | but Arnulf will be here any minute |
| 06:04:30 | tbaschetti: | hi andreas, nice to meet you here |
| 06:05:04 | : | * seven is here now. |
| 06:05:17 | andreas: | Hi Arnulf |
| 06:06:16 | Testbaudson: | seven: Armin Retterath's phone is busy... |
| 06:06:25 | seven: | no wonder... :- ) |
| 06:06:48 | seven: | Did you ever see him on IRC here? |
| 06:06:54 | Testbaudson: | seven: here's the proposal: http://www.mapbender.org/GsoC2009_CSW_proposal |
| 06:06:55 | sigq: | Title: GsoC2009 CSW proposal - MapbenderWiki ( at www.mapbender.org ) |
| 06:06:59 | Testbaudson: | maybe you can take a look |
| 06:06:59 | seven: | Do we have an agen... |
| 06:07:00 | seven: | OK. |
| 06:08:34 | Testbaudson: | basically, we just want to discuss what we want Mifan to do |
| 06:08:57 | seven: | good |
| 06:09:03 | Testbaudson: | we have several interested parties |
| 06:09:08 | : | * seven is still reading... |
| 06:09:21 | Testbaudson: | Andreas Fischer |
| 06:09:25 | Testbaudson: | Armin Retterath |
| 06:09:31 | Testbaudson: | WhereGroup |
| 06:09:51 | seven: | The GeoNetwork project will also be interested. |
| 06:10:04 | Testbaudson: | we need to find a lowest common denominator |
| 06:10:58 | seven: | First question is whether we qurey remote servers or just harvest them and then do all else inside Mapbender. |
| 06:11:03 | Testbaudson: | please let me know when you have all read the proposal... |
| 06:11:11 | : | * seven reads on... |
| 06:11:43 | andreas: | still reading ... |
| 06:11:44 | Testbaudson: | Armin Retterath's phone is still busy :-( |
| 06:12:07 | seven: | send him an email |
| 06:12:19 | Testbaudson: | already did so |
| 06:14:28 | : | * seven is done reading. |
| 06:14:56 | Testbaudson: | ( as far as I understand, tbaschetti is just an observer...but please feel free to share your opinion anytime ) |
| 06:15:23 | tbaschetti: | right, i'm just curious |
| 06:16:20 | andreas: | will be back in a minute - too much coffee ;- ) |
| 06:16:38 | Testbaudson: | seven: about your first question: can you elaborate? |
| 06:17:19 | seven: | Yes. |
| 06:18:05 | seven: | Will MB query remote services - a list of them and then wait for the answer? |
| 06:18:22 | seven: | I think this is a bad idea, from experience with catalog services. They are dead slow. Or dead. |
| 06:18:43 | seven: | Therefore it might be more interesting to harvest catalog on a regular bassis |
| 06:18:47 | Testbaudson: | ok..so Mapbender acts as a cache? |
| 06:18:51 | seven: | and store everything inside MB |
| 06:18:53 | seven: | Yes. |
| 06:19:14 | seven: | Then we can comfortably query using standard PHP, SQL and our data structure. |
| 06:19:33 | seven: | The focus would then also lie on import filters for different metadate formats. |
| 06:20:05 | seven: | This is just an idea. Maybe the project was planned differently. |
| 06:20:13 | andreas: | another opinion: I think MB should not be a datastore, so in GDI GeoNetWork could be |
| 06:20:19 | seven: | Maybe catalogs are so big that we do not want / can cache them. |
| 06:20:38 | seven: | GeoNetwork also has its drawbacks... |
| 06:20:57 | seven: | I am not sure whether I want Mapbender to alsways have to rely on a GeoNetwork instance to run. |
| 06:21:12 | seven: | it is a dependency that comes with Java, Tomcat, etc. |
| 06:21:24 | seven: | Up to now we only need PHP and a database. |
| 06:21:27 | andreas: | I was thinking about getting just an interface, maybe provide a picklist, which catalog a user wants to query |
| 06:21:29 | seven: | ...and a webserver and so on. |
| 06:21:42 | seven: | Yes, but exactl that is a pain the ass. |
| 06:21:56 | seven: | Because all those servers are really odd and slow to query and use. |
| 06:22:03 | Testbaudson: | maybe we can do the caching in a second step? |
| 06:22:35 | Testbaudson: | I think we should have both options anyway |
| 06:22:42 | seven: | Hmmm. |
| 06:22:53 | seven: | I have not seen one( ! ) CSW that performs. |
| 06:23:10 | seven: | They are all dead in the water. It is amazing, I know... |
| 06:23:14 | andreas: | I known one |
| 06:23:40 | andreas: | ... there is - for example - geocatalog |
| 06:24:12 | seven: | URL |
| 06:24:14 | seven: | ? |
| 06:24:23 | andreas: | just a second ... |
| 06:24:53 | andreas: | www.geocatalog.de :- ) |
| 06:25:45 | andreas: | what I intended to get developed is very similar to the Geodatenserver of RVR ... |
| 06:26:33 | andreas: | .. user can click on "Get another theme" and a very simple ( ! ) dialog opens |
| 06:27:46 | seven: | ...waiting for geocatalog... |
| 06:28:12 | andreas: | from this it is possible to pick a catalog, e.g. geocatalog, and fill in a keyword to retrieve a list a wms that fit |
| 06:30:06 | andreas: | you might take a lokk at http://www.rvr-online.de/karten_geodaten/geodienste/Geodatenserver.php - start LUKAS and pick "Themen hinzuladen", then go to "Katalog-Suche" |
| 06:30:07 | sigq: | Title: RVR Online - Geodatenserver ( at www.rvr-online.de ) |
| 06:32:24 | Testbaudson: | seven: do you know how this all relates to opensearch? |
| 06:32:37 | seven: | No, not really. |
| 06:32:37 | andreas: | right now only 'geocatalog' can be choosen but my idea is to provide something like a picklist, so the user can get the requested search done by more than one at the same time |
| 06:32:37 | Testbaudson: | is opensearch some kind of csw wrapper? |
| 06:32:45 | seven: | CSW is old compared to OpenSearch. |
| 06:33:08 | seven: | They are evaluating whether OpenSearch would be something to extend or replace CSW |
| 06:33:14 | seven: | They = OGC |
| 06:34:37 | seven: | Moin armin |
| 06:35:01 | seven: | armin: http://logs.qgis.org/mapbender/ |
| 06:35:02 | sigq: | Title: Index of /mapbender ( at logs.qgis.org ) |
| 06:35:05 | seven: | Pick today |
| 06:35:15 | seven: | To read up on what was already discussed. |
| 06:37:49 | armin: | test |
| 06:38:02 | Testbaudson: | http://logs.qgis.org/mapbender/%23mapbender.2009-04-29.log |
| 06:38:03 | sigq: | Title: IRC Log - #MAPBENDER ( at logs.qgis.org ) |
| 06:38:58 | : | * seven tries to enable Pop-Ups for LUKAS... |
| 06:39:27 | Testbaudson: | you might take a lokk at http://www.rvr-online.de/karten_geodaten/geodienste/Geodatenserver.php - start LUKAS and pick "Themen hinzuladen", then go to "Katalog-Suche" |
| 06:39:28 | sigq: | Title: RVR Online - Geodatenserver ( at www.rvr-online.de ) |
| 06:39:56 | andreas: | for sure, not erverything is fine at Geodatenserver :- ) |
| 06:40:48 | andreas: | It's a little hard to express my idea in this medium, but I will try ... |
| 06:41:22 | andreas: | ... MapBender could be MapClient, UserDB etc. |
| 06:41:24 | seven: | http://217.78.131.130/lukas/sdiWMSCatSearch.do;jsessionid=7D423CF12DA01D3BF8AA10403438829E |
| 06:41:25 | sigq: | Title: Add a Web Map Service ( at 217.78.131.130 ) |
| 06:41:27 | seven: | Ahrgl. |
| 06:42:28 | seven: | First click: Waiting for http://217.78.131.130/... |
| 06:42:28 | andreas: | ... and GeoNetWork could be the DB to store metadata in and provide a catalogService |
| 06:42:41 | armin: | we need a html interface for do the search |
| 06:43:14 | andreas: | ... so the user is not interested in metadata for itself ( ? ) but wants to search for geodata |
| 06:43:14 | seven: | GeoNetwork does not really have a database. It only stores the XML - if I remember correctly. |
| 06:43:48 | armin: | it has a db |
| 06:44:04 | andreas: | ... as armin says, we need a html interface, so the use can do simpleSearch ( ! ) out of the mapClient and can add wms, that were found |
| 06:44:11 | armin: | but the xml data is stored in a single field |
| 06:45:17 | armin: | we implemented a opensearch interface for portalu |
| 06:45:58 | armin: | the opensearch urls are stored in the mb database and are requested by some php scripts |
| 06:46:42 | seven: | This is more like my suggestion. So it seems that we have different options to debate. |
| 06:46:48 | armin: | have a look at http://www.geoportal.rlp.de and search after 'wasserschutzgebiete saarland' |
| 06:46:50 | sigq: | Title: GeoPortal Rheinland-Pfalz - Startseite ( at www.geoportal.rlp.de ) |
| 06:46:52 | seven: | Lets find the consensus solution. |
| 06:47:35 | armin: | the different search interfaces should be handled like the wms or wfs services - update exchange and so on |
| 06:47:41 | andreas: | armin: o.k. about GeoNetWork, but I mean any kind of datastorage - so MapBender mightnot have to be a storage for metadata to |
| 06:48:00 | armin: | the<y should have owners who register them in the mb-db |
| 06:48:57 | armin: | in our case mb is no storage, the md records are only saved in the tmp by session |
| 06:49:58 | seven: | Mapbender already stores metadata for all its services. |
| 06:50:25 | seven: | Should Mapbender also have a CSW Server intterface? |
| 06:50:27 | armin: | there are 2 different scripts: the first makes the request and handles the first resut, the second makes the detailed view |
| 06:50:41 | seven: | Or update another server using CSW? |
| 06:50:55 | andreas: | I know, about the storage of some metadata in MB |
| 06:51:12 | armin: | we have a interface for DSC - thats the way portalu do the exchange |
| 06:51:24 | andreas: | ... but its not all of metadata INSPIRE efforts |
| 06:51:41 | seven: | armin: expl. DSC pls |
| 06:51:49 | seven: | Hmm. Gone. |
| 06:51:58 | seven: | He probably typed too fast. |
| 06:52:17 | andreas: | so we have CSW with other software, so why do we need this in MB? |
| 06:52:26 | Testbaudson: | dsc = data source client |
| 06:52:38 | seven: | armin: can you please explain what DSC means? |
| 06:52:42 | seven: | OK. |
| 06:52:53 | seven: | Not okay? What is that? |
| 06:53:17 | seven: | andreas: Performance. |
| 06:53:31 | seven: | So maybe Testbaudson is right and |
| 06:53:38 | seven: | we should first implement the client |
| 06:53:48 | seven: | and then check whether we make it a cache. |
| 06:53:57 | andreas: | Performance is one aspect |
| 06:54:45 | armin: | the information modell of portalu is enough for inspire |
| 06:54:45 | armin: | the dsc is a catalog server component which gives information in the information model of portal u |
| 06:54:45 | armin: | it is the most important aspect i think |
| 06:55:30 | armin: | the problem of csw is, that there is no topology or broker technology |
| 06:56:06 | armin: | geonetwork harvest all md records and create new indices |
| 06:56:32 | andreas: | ... but storing redundant data might bi a problem to, get it update ... |
| 06:57:06 | andreas: | ... so I was thinking first of a component-model of GDI with interfaces |
| 06:57:28 | seven: | Update using GeoRSS and maybe AtomPub |
| 06:57:34 | armin: | thats the way portalu acts |
| 06:57:54 | andreas: | AtomPub? |
| 06:57:55 | armin: | now idea about georss or atompub |
| 06:58:11 | seven: | andreas: Where in that component model is the unique id of a WMS? |
| 06:58:31 | : | * seven starts to understand the differences... Interesting. |
| 06:58:38 | armin: | or the unique id of a layer |
| 06:58:46 | andreas: | do we need a unique id? |
| 06:58:51 | seven: | the world needs it |
| 06:58:56 | andreas: | ? |
| 06:58:58 | seven: | Thats the way the Web works.. |
| 06:59:04 | armin: | yes to lock many clients to this id |
| 06:59:07 | seven: | Everything has ann ID. Its a URL. |
| 06:59:41 | armin: | but the url doesnt change when the attributes of one layer changes |
| 06:59:49 | seven: | That URL has a life time. |
| 06:59:53 | seven: | It says in the HEADER |
| 07:00:07 | seven: | and it can be updateed automatically, when its life time ends. |
| 07:00:09 | seven: | Done. |
| 07:00:12 | seven: | Very simple. |
| 07:00:15 | andreas: | When I search for geodata e.g. in geocatalog and I find a WMS I#m interested in ... |
| 07:00:42 | andreas: | ... I just get the onlineResource ( unique ID! ) and do a getCapabilities |
| 07:01:18 | armin: | the problem is, that the services should be connected to clients like qgis or some other |
| 07:01:39 | andreas: | as a user I don't want to leave my lovely MBClient for searching proposal |
| 07:01:41 | armin: | the searching is only one component - for human users |
| 07:02:05 | armin: | the searching can be done in the client |
| 07:02:28 | armin: | we only disconnect it cause we use typo3 |
| 07:03:20 | armin: | the biggest problem is that there are many catalogs and the user must know in which catalogue he should search |
| 07:03:38 | andreas: | armin: yes I understand, so a searchingInterface should be implemented in MB |
| 07:03:44 | seven: | therfore I thought that... |
| 07:03:59 | seven: | all catalogs are connected by an MB admin... |
| 07:04:15 | seven: | and then store the relevant information in the MB storage.... |
| 07:04:17 | andreas: | catalogs could be choosen by user from list or - right - the admin could do this job |
| 07:04:38 | seven: | and only when it is queried, ask the remote catalog server. |
| 07:04:53 | seven: | I think it is necessary that the admin does that because |
| 07:05:00 | seven: | there are diffferent CSW schema |
| 07:05:00 | armin: | thats the way we do it - but for csw it is very slow |
| 07:05:29 | seven: | which need to be modified to fit with the single HTML user search interface. |
| 07:05:35 | andreas: | o.k. - it might be to difficult for common users anyway ... |
| 07:05:40 | armin: | thats the way |
| 07:05:50 | andreas: | ... so the admin might be the right person |
| 07:06:09 | armin: | we need a definition for a good search interface |
| 07:06:13 | seven: | The job of customizing the search only needs to be done once for each catalog. |
| 07:06:14 | andreas: | let's keep it simple for the user |
| 07:06:24 | seven: | the user only sees one interface. |
| 07:06:43 | armin: | but we dont have a ranking!!! |
| 07:06:48 | seven: | Hehe. |
| 07:07:14 | armin: | there is no one when using csw |
| 07:07:17 | seven: | And we can only do our own ranking when we have a big community of people who vote on it. |
| 07:07:19 | andreas: | in my experience, a simple userIntrerface is VERY important... |
| 07:07:22 | seven: | By using it or not using it. |
| 07:07:30 | andreas: | ... take a lool at Google... |
| 07:07:38 | seven: | lol |
| 07:07:40 | andreas: | ... or GoogleMaps |
| 07:07:52 | andreas: | there is just one field for searching |
| 07:08:07 | seven: | www.geoportal.rlp.de |
| 07:08:47 | armin: | please test the search for 'wms saarland' and add the service to mb |
| 07:08:58 | andreas: | this is very simple and one reason for the popularity of this application |
| 07:09:47 | seven: | Cannot add. Its secured. |
| 07:09:49 | seven: | :- |
| 07:09:59 | armin: | the biggest problem is the definition of the categories for the different resultsets |
| 07:10:19 | armin: | seven: choose the one in metadata categorie ;- ) |
| 07:11:30 | Testbaudson: | maybe we should start to find a lowest common denominator |
| 07:11:34 | Testbaudson: | I think one point is |
| 07:11:47 | Testbaudson: | to start to add the geoportal.rlp functionality to Mapbender trunk |
| 07:12:15 | Testbaudson: | I think this could be merged with Mifan's work |
| 07:12:21 | armin: | but we need a concept to add the service not by write the url to the session |
| 07:13:28 | armin: | we need an abstract class for the search interfaces |
| 07:13:46 | armin: | and a database schema |
| 07:14:36 | Testbaudson: | ok, I assumed we have new table for catalogue services |
| 07:14:46 | Testbaudson: | similar to wfs, wms |
| 07:14:51 | armin: | yes |
| 07:14:53 | andreas: | yes |
| 07:15:17 | andreas: | and this tabel should provide information about the CSW Schema too |
| 07:15:53 | seven: | andreas: Why? |
| 07:16:45 | andreas: | I think this is important to comunicate with different catalogs, isn't it? |
| 07:17:09 | armin: | what is the schema? - the url |
| 07:17:19 | seven: | Yes, but if we store the schema what are we supposed to use it for? |
| 07:17:48 | seven: | armin: Yes. But then what do you have to know in Mapbender in order to query that catalog? |
| 07:18:08 | andreas: | My idea was, too store all the information that is nessesary to get talked to different CSW |
| 07:18:44 | seven: | I think it willlook like the current WFS interface. Pretty complex, messy thing. |
| 07:18:45 | armin: | but you only need the spec id like CSW 2.0.2, csw 2.0.1 .... |
| 07:19:02 | seven: | Yes, plus the specifics of every single service. |
| 07:19:28 | armin: | it is easier than wfs, cause there is only one schema for csw - as i know |
| 07:19:59 | armin: | we dont need to support the older ones |
| 07:20:25 | andreas: | couldn't do the MB-CSW-Client a capabilitiesRequest on the CSW-Server, to gain information? |
| 07:20:26 | armin: | lets begin with csw 2.0.2 and portalu's opensearch |
| 07:20:57 | seven: | andreas: Yes. Then you have the capabilities doc in Mapbender. What do you do with it? |
| 07:21:33 | : | * Testbaudson has not fully understood the relation between csw and opensearch |
| 07:21:44 | armin: | csw is wfs |
| 07:21:44 | andreas: | me not eather |
| 07:21:45 | seven: | armin: Can you elaborate a bit? |
| 07:21:55 | armin: | yes... |
| 07:22:29 | armin: | csw is like a profile of wfs - but has specific search fields and a specific schema - iso19139 |
| 07:22:51 | andreas: | seven: I thought the capabilities of the CSW could provide important information to communicate with different CSW |
| 07:22:54 | armin: | opensearch is a simple xml search interface |
| 07:23:37 | armin: | andreas: you need it to find the bindings |
| 07:24:05 | armin: | and to see if it supports transaction |
| 07:24:10 | andreas: | opensearch and CSW: I think we should provide a searchInterface in order to find data |
| 07:24:58 | andreas: | ... if we can assume that catalogs become more important the interface should be able to ask catalogs ... |
| 07:24:58 | Testbaudson: | as far as I understand |
| 07:25:04 | Testbaudson: | Mifan should focus on CSW |
| 07:25:13 | Testbaudson: | and we should leave the design open to other services |
| 07:25:17 | Testbaudson: | like opensearch |
| 07:25:22 | Testbaudson: | is this ok? |
| 07:25:30 | armin: | ok |
| 07:26:07 | andreas: | if we do assume that searching will be done in other ways too, be should enable this as well - because the user wants to find the data and doesn't care about the technique |
| 07:26:19 | andreas: | ok |
| 07:26:43 | Testbaudson: | I think the challenge in this project is |
| 07:26:50 | Testbaudson: | to find a starting point |
| 07:26:55 | Testbaudson: | Mifan can't do all we want |
| 07:26:56 | armin: | we need a concept for n catalogues or serachintefaces - but a human must use it |
| 07:27:05 | Testbaudson: | but what he builds has to be extensible |
| 07:27:11 | andreas: | keep it simple! |
| 07:27:25 | armin: | the starting point is the html interface and the things we want to find! |
| 07:27:34 | andreas: | right! |
| 07:27:55 | Testbaudson: | ok, so the client side is fairly easy |
| 07:28:06 | Testbaudson: | maybe we should take about the admin side |
| 07:28:31 | Testbaudson: | about modeling the database for example |
| 07:28:31 | armin: | there must be some categories to classify the interfaces |
| 07:28:58 | andreas: | yes I think so too, in my opinion a search field for keyworlds and a filter for the bounding box as the current map extent could be enought for first |
| 07:29:02 | armin: | maybe one first search after the interfaces from different institutions |
| 07:29:46 | armin: | we need a filter for the requested resource types |
| 07:30:23 | andreas: | wms and wfs? |
| 07:30:42 | armin: | what we search for: layers, objects, maps, applications, wmc!( views ) |
| 07:31:17 | armin: | no one knows wms and wfs |
| 07:31:19 | andreas: | I assume the user works with a mapclient ... |
| 07:31:34 | andreas: | ... and suddenly he is missing some data ... |
| 07:31:42 | andreas: | so what he wants is .. |
| 07:31:48 | seven: | Maybe she just needs a KML. |
| 07:31:53 | andreas: | a searching tool for this data .. |
| 07:32:20 | andreas: | ... and after he got the results a simple click to add this data to his map. |
| 07:32:52 | Testbaudson: | andreas: so sth. similar to geoportal.rlp.de would be ok for you? |
| 07:32:54 | andreas: | In RVR GeodatenServer this is done in steps |
| 07:32:57 | armin: | but think about the wfs - it is a little bit complicated |
| 07:33:12 | andreas: | Christoph: yes |
| 07:33:19 | Testbaudson: | ok |
| 07:34:50 | andreas: | About the steps: find the wms ( wfs ) - offer them in a list - choose a wms/wfs - offer a picklist for the layers to be added |
| 07:35:16 | armin: | but you dont search for wms! |
| 07:35:29 | armin: | you search for layers and objects |
| 07:35:32 | seven: | armin: We are talking about different work flows here. |
| 07:35:37 | andreas: | armin: I don't understand |
| 07:35:51 | seven: | andreas is talking about gis users. |
| 07:35:59 | andreas: | do I? |
| 07:36:02 | seven: | who need daa in everyday work. |
| 07:36:10 | seven: | daa = data |
| 07:36:23 | armin: | but also them search for layers |
| 07:36:40 | seven: | armin is talking more about a geoportal product that can be used by professionals |
| 07:36:50 | seven: | but also turns up things for other user groups. |
| 07:37:12 | : | * seven might also have interpreted this wrongly... :- ) |
| 07:38:22 | seven: | Testbaudson: You suggestion to merge geoportal.rlp into MB is ok. |
| 07:38:36 | seven: | But that is not really a job for Mifam, is it? |
| 07:38:52 | Testbaudson: | no, but we should be able to use Mifan's work... |
| 07:38:54 | seven: | m->n |
| 07:38:59 | armin: | the merging would not be so difficult i think |
| 07:39:00 | Testbaudson: | in a "relaunched" Geoportal.rlp |
| 07:39:04 | seven: | Good. |
| 07:39:39 | seven: | Maybe Mifan also has good ideas. From what I read he seems to have done quite a lot of interesting things already. |
| 07:40:26 | Testbaudson: | I think we should start to sum up our discussion |
| 07:40:44 | armin: | ok |
| 07:40:55 | Testbaudson: | we want an easy to use search interface |
| 07:41:09 | Testbaudson: | that queries abstract services |
| 07:41:18 | Testbaudson: | which could be CSW, OpenSearch or whatever |
| 07:41:30 | Testbaudson: | and the result would be an abstract result object |
| 07:41:38 | Testbaudson: | which could be a layer, a feature |
| 07:41:40 | Testbaudson: | etc |
| 07:41:43 | Testbaudson: | is this ok? |
| 07:41:52 | armin: | yes |
| 07:42:11 | andreas: | I think it is |
| 07:42:19 | seven: | yes |
| 07:42:19 | Testbaudson: | then we need to discuss how to do the administration |
| 07:42:24 | Testbaudson: | we need a table for services |
| 07:42:37 | Testbaudson: | and means to distribute them among applications |
| 07:42:44 | Testbaudson: | maybe like gui_wms |
| 07:43:20 | Testbaudson: | we should start with csw |
| 07:43:21 | andreas: | yes and the admin should be able to decide, which ones to offer |
| 07:43:45 | andreas: | starting with CSW is a good idea, I think |
| 07:43:46 | Testbaudson: | andreas: just like you can do with wfs gazetteer, for example? |
| 07:44:21 | armin: | with wfs_conf i think |
| 07:44:28 | Testbaudson: | armin: true |
| 07:44:33 | andreas: | o.k. |
| 07:45:00 | armin: | we need an abstract result object |
| 07:45:37 | Testbaudson: | we will have to define what results will be expected |
| 07:45:43 | andreas: | what ever that is, it should be possible to add this to the current map by click |
| 07:45:55 | armin: | the results that can be handled by mb! |
| 07:46:15 | Testbaudson: | ok, so basically wms, wfs, features? |
| 07:46:16 | armin: | or to view the bbox e.g. |
| 07:46:28 | seven: | the results that cvan be handled by the browser. |
| 07:46:33 | armin: | wfs_conf for the mb intern results! |
| 07:46:53 | Testbaudson: | ok |
| 07:47:02 | seven: | So it is just another WFS? |
| 07:47:16 | armin: | csw is a wfs |
| 07:47:31 | armin: | but not so general |
| 07:47:36 | : | * seven leans back. Go on. |
| 07:48:03 | armin: | with what |
| 07:48:07 | Testbaudson: | I think what we have created as a WFS gazetteer is a subset of this new catalogue search |
| 07:48:34 | Testbaudson: | so it would make the WFS gazetteer obsolete ( ? ) |
| 07:48:52 | armin: | the csw is a subset of o wfs - its a wfs_conf |
| 07:48:53 | andreas: | let's keep technical stuff behind first: It should be possible to query any CSW-Server, that anyone can get knowledge from |
| 07:49:42 | Testbaudson: | andreas: I guess we should have something similar to "Load WMS/WFS" for CSW |
| 07:49:48 | Testbaudson: | which reads the capabilities document |
| 07:49:56 | Testbaudson: | and stores this info in the new table |
| 07:50:39 | armin: | and reads opensearch urls and store them in the same table |
| 07:50:48 | Testbaudson: | armin: why is csw a subset of WFS? It can return more than a WFS |
| 07:50:54 | andreas: | I'm thinking in same direction |
| 07:51:05 | armin: | and what? |
| 07:51:28 | Testbaudson: | I assumed WFS only returns features |
| 07:51:44 | Testbaudson: | while a catalogue returns features or other services |
| 07:51:46 | seven: | and a feature can be pretty anything |
| 07:51:51 | armin: | and metadatarecords are simple features with a bbox |
| 07:52:09 | seven: | which is nothing but a standard geo data object. |
| 07:52:17 | armin: | the service record is only a record |
| 07:52:37 | seven: | like mapbender users wfs |
| 07:52:47 | armin: | the difference is, that the csw use a special schema |
| 07:52:47 | seven: | it has a link to the next URL. thats it |
| 07:53:00 | andreas: | If someone has a CSW-Server like any government organisation, this one provides a CatalogInterface ... |
| 07:53:01 | armin: | yes |
| 07:53:50 | andreas: | ... so this one should ( just ) return the OnlineResources from other services, to get them adressed |
| 07:54:21 | andreas: | ... erything else, is communication to one of the offered services, the user has picked |
| 07:55:06 | armin: | maybe, but the csw 2.0.2 gives everey wms layer as single metadatarecord - with a link to its service |
| 07:55:34 | : | * Testbaudson doesn't want to spoil the party...but it's almost 18:00 |
| 07:55:44 | armin: | the problem are the services which have 200+ layers |
| 07:55:47 | Testbaudson: | maybe we should schedule a 2nd meeting with Mifan |
| 07:55:57 | armin: | we should |
| 07:56:05 | Testbaudson: | how about next week? |
| 07:56:06 | seven: | Shall we continue via mail on dev? |
| 07:56:15 | andreas: | it's a problem of that service, why does such a big think exist? |
| 07:56:18 | seven: | at least a summary protocol |
| 07:56:29 | andreas: | o.k. |
| 07:56:36 | Testbaudson: | seven: I would send the link to the log file |
| 07:56:47 | Testbaudson: | if you want more...someone would have to summarize |
| 07:57:06 | seven: | "Someone" usually is /dev/null |
| 07:57:13 | andreas: | do we have the same in mind? |
| 07:57:26 | armin: | maybe we should meet more physically |
| 07:57:39 | seven: | a work meeting |
| 07:57:44 | seven: | anybody have some funding? |
| 07:57:45 | armin: | yes |
| 07:57:48 | andreas: | yes, it is hard to talk this way |
| 07:58:03 | andreas: | no |
| 07:58:04 | seven: | we can invite you to Bonn. |
| 07:58:09 | seven: | That should be half way. |
| 07:58:11 | armin: | in may |
| 07:58:22 | seven: | Good. |
| 07:58:30 | seven: | That is next week. |
| 07:58:36 | armin: | maybe |
| 07:58:50 | seven: | So in Maybe. I like that one. :- ) |
| 07:58:53 | andreas: | No way for me next week |
| 07:59:09 | seven: | There is a Mayen in RLP... |
| 07:59:12 | : | * seven shuts up. |
| 07:59:35 | Testbaudson: | maybe we can do a doodle poll |
| 07:59:38 | armin: | use doodle? |
| 07:59:44 | Testbaudson: | we don't have to agree now |
| 07:59:58 | seven: | We can ask Mifan to integrate what we have discussed into the OSGGeo Wiki article. |
| 08:00:08 | seven: | Someone > Mifan |
| 08:00:23 | Testbaudson: | osgeo wiki article? |
| 08:00:28 | Testbaudson: | please specify |
| 08:00:29 | armin: | to much work for him, he should think more than write ;- ) |
| 08:00:30 | seven: | GSoC |
| 08:00:40 | Testbaudson: | you mean his proposal |
| 08:00:41 | seven: | He seems to be pretty organized |
| 08:00:50 | seven: | Yes. |
| 08:00:57 | Testbaudson: | ok |
| 08:01:02 | seven: | We can move it to Mapbender Wiki. |
| 08:01:23 | Testbaudson: | so I will create a poll with the 2nd and 3rd week in May...ok? |
| 08:01:31 | seven: | ok |
| 08:01:34 | Testbaudson: | it should not be later than that |
| 08:01:41 | andreas: | ok |
| 08:01:52 | armin: | ok |
| 08:02:34 | Testbaudson: | ok, thatnks for attending |
| 08:02:39 | seven: | Thanks all. |
| 08:02:40 | Testbaudson: | it has been a productive meeting |
| 08:02:48 | Testbaudson: | let's talk face to face next time |
| 08:02:53 | armin: | bye - howto log out? |
| 08:02:58 | Testbaudson: | bye and have a nice evening |
| 08:03:03 | seven: | /quit |
| 08:03:05 | Testbaudson: | armin, just close the window |
| 08:03:12 | armin: | \quit |
| 08:03:22 | andreas: | o.k. thank you for talking - bye |
| 08:03:23 | seven: | :- ) |
| 08:03:27 | Testbaudson: | thanks andreas, bye |
| 08:03:29 | seven: | Bye. |
| 08:04:28 | Testbaudson: | thanks for your help, seven |
| 08:04:42 | Testbaudson: | good to have a lot of different persoectives |
| 08:04:59 | : | * seven is happy to be able to collaborate a bit more recently. |
| 08:08:32 | Testbaudson: | have a nice evening, bye |