Add to Watchlist

State of GeoServer


Citation of segment
Embed Code
Purchasing a DVD Cite video

Formal Metadata

Title State of GeoServer
Title of Series FOSS4G Bonn 2016
Part Number 68
Number of Parts 193
Author Garnett, Jody
Aime, Andrea
License CC Attribution 3.0 Germany:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
DOI 10.5446/20424
Publisher FOSS4G, Open Source Geospatial Foundation (OSGeo)
Release Date 2016
Language English

Content Metadata

Subject Area Computer Science
Abstract State of GeoServer provides an update on our community and reviews the new and noteworthy features for the Project. The community keeps an aggressive six month release cycle with GeoServer 2.8 and 2.9 being released this year. Each releases bring together exciting new features. This year a lot of work has been done on the user interface, clustering, security and compatibility with the latest Java platform. We will also take a look at community research into vector tiles, multi-resolution raster support and more. Attend this talk for a cheerful update on what is happening with this popular OSGeo project. Whether you are an expert user, a developer, or simply curious what these projects can do for you, this talk is for you.
Keywords Boundless

Related Material

again we had to start so please welcome well Jody ends Andrea and you the stated GeoServer 2016 the energy and I were just talking and it actually appears without a fair bit
of work this year so this is our chance to tell you all about it and is a technical solutions and he's been a long standing by leader in on Server project was charter member of the UN was a long-standing member of the that GeoServer steering committee and you might see me a few of the talks so GeoServer
2016 what have we accomplished for
those of you who knew to choose GeoServer project it's a Java application server that really focuses on making sharing of geospatial data easy to make use of a wide range of standards have publicized today in terms of the health check communities been
fairly static this year we have increased the number of commuters from 28 to 37 but our mailing lists and Interpol requests are holding steady at you know around 600 around 500 pull requests a year we do have a slightly smaller active base of contributors this year the pull requests are still going strong our base is really active and very healthy and you in the has also stepped down as always studio project officers and parsimonious volunteer to take on that we are keeping up an aggressive Release dental for those of you new to the project we have a little bit of a staggered release quite so each release spends 6 months as the stable release and then it receives a further 6 months of maintenance upgrades we do have 1 good which In release schedule this year we had about a two-month delay and I'll talk about the reasons for that is a little bit later so GeoServer 2 9 0 release was delayed by about 2 months but also in terms of the project we have a long-standing overhead where a lot of the pull requests that that come in we actually have to reject and say hey you know about those headers communities updated it's been a significant 7 . 7 % of our feedback this year is just as the developers please try again we talked to the always legal advice thanks to the OSG Award for making that happen and the practice and on updating your head is we can actually relax a little bit ah practices were set before the before the US joined the Greek convention and with that in mind
we have slightly updated this
is actually something we're not gonna talk about very much but it has been the major work that's gone into the project this here is focused on maintenance and technical debt the GeoServer project 10 to 12 years old quality probably more and more OK uh and over that time we developed a large codebase that we've had to well will always have to work on bringing it into the future 1 of the things
we've done is it's become impractical rely solely on volunteer time in order to keep track of the all the issues are coming from the community and where it's kind experiment of setting up a monthly kind of bugs storm and in order to gather some of our new developers and teach them how to look into the issues as they come in and see we can be a bit more timely we try this for the 1st time in July and that wouldn't be trying it again this weekend at the Cold Spring for
our first one hour we managed to close a large number of like feature requests and data and wishes for new features that have been open for like over 5 years I think in a feature request hasn't attracted funding in 5 years it's safe to let it to say nobody is interested in some yeah so this is an experiment to try the keen maintenance activity we took on this year was upgrading our our user interface from which at 1 point 4 no but it 7 and this was actually a major undertaking flow are GeoServer development team from all over the world to Victoria Canada and we took care of we had a really organized and where you where was successful in making this a great the other kind of challenge I guess as we get updated Javadi this year and that's simply because our customers ask us to Java 7 no longer receives kind of free security updates and so our customers are migrating to generate as part of that we had a surprise this was responsible for that too much delay the library the Spring framework that we used to G blue GeoServer together was incompatible with chapter 7 so we needed to update the spring force it's going to have this I think you so just a quick
enumeration of the various features that we are working on or improvements 1st that you package model is a communicable right now we are going to push it to to to support this status by increasing its testing and making sure it's compatible with their what is the best way we updated the
cascading the professor skating that is the ability for the users to act as a client to another level of the server we have an old Kobe is showing its age we made a number of improvements and tested it further and accelerated for consumption terms are after the
sources that we improved our must be support supporting vector masks and roster masks so that you can cut the rest of the time I don't know what kind of plants and stuff like that that's we added the we made
them all I more powerful in order to support that was Ike's made of a mix of hormones like great collected in RGB in all in the same was like uh we also added support for multiple projections so all
you can actually mixed together images in different projections and the same was like that's gonna be a reason just sort the context that next a also the
level of the image was like we are going to start working soon like next week on access from the removal of say you have a very deep IMO's Iike with the many images overlapping each other and your controlling 2nd order by recency of a solution core vs. gray whatever so you have about dynamic 2nd order and you wanted to avoid opening of images that are not putting not contributing to the output that's what the excess growth
what's going do we're going to support the multiple colleges for them was like we already do that for the CDF and the that's something we
don't do yet so that we can make the enemies of complex structures we're going to optimize the coverage of user so that awareness of the chooses to only affects 1 and over the coverage of you we are not very good the orders from that the 5 so speed up and we are adding more support for repeated projections Ben current bet is that this and it's nice because to supported is then actually have to contribute to projects to the next CDF Java libraries 1st sort the Bayes library that we use and then to just over 2 other support for this projection which is used in the weather forecast
style Our without support for finally for the perpendicular offset for line and polygon stuff that you can efficiently offset in line that's part of so the 1
1 about of course we supported also and this is the 1 of we are adding some insulin like extensions to get much so that you can just stick it pilot wouldn't you just an enumeration of course and then have just about dynamic can apply the pilot on them on your roster based on the minimum and maximum values we added support for each frame Animations you can tell just over to support that out there in the logarithmic scaling instead 1 he's very nice and good for methodological literally that's why there's some
good work from during up from and this basically you know that the CSS stage has this little play in preview mode that it's nice make nice to edit we have are moving back to the main result page so you can actually type in your style and preview directly in the page without having to save the go to the previous on so it's much because we have some CSS features from users as feature coming so Jesus as the support and the transformation that we are adding it and I'm also working on the rule nesting of which helps to make the site of a more compact we also have a new language
for a standard which is why so the it's sort of derivative well it's like a so the body wire mounts also lots of boilerplate removal much more compact and it's one-to-one compatible with so the meaning that you can take so the and terminating pulled wire so the automatically which is nice
and we had the support for customizing graphics you know that just has to get alleging graphics support maybe use the legend for you but sometimes the job is not done very well how or you might want to have your own custom legend you can now popular a static image to do that but in terms of
the we are adding for over 2 9 of the 10 new J peg or PNG format for the majority that needs to have transparency so right now you're stuck to PNG which is being this format basically checks if you're image has actually any 1st currency in it and it will encode in PNG or that depending on whether there is transparency we have
support for everybody in that the mass so my book societies have agreed with a bit more flexibility and compared to the basic so we support the numbers of times and that it's available in projections of course it's up to the point of the sites that they want to support the sort not we have a community
modules generating vector times but those can be cast in the UK we could be cast as well so you can do let's say my books without the mob books by using this operation and you'll catch and the last these shows that's can become an extension for 210 right that's great
but only in terms of legend and graphic we also added the
ability from declined to to control the layout of the legend so that if you wanted to find a single column in a single rule 2 columns and rows and so on so you can now specify that in terms of the professor we
breaking the limits of the file sizes of someone tries to make a diamond-shaped file from the professor be given to the device you're asking for trouble right now the cold recognizes that and will just south of Beijing and generate multiple that are we have With this is not exactly just over by the government project the pages out there is a mapping tool is destined to and we worked on I'm making it the support that our schema mapping phi so that you don't have to get book crazy and adding tags amount and all day to get a complex feature support both in terms of complex feature support when we are joining tables in the database to build a complex feature of 3 we are now able to also send down filters to the database are for joining tables which we weren't able to do before in terms of type
catching we have added the support from the tiles you can use a single single and time like totally compliant but we also added a few extras such as having multiple MB tiles and storing the unknown lubricator projections in story more more foreigners than just the endures a bag of course if you use these extensions during ongoing able to use those and because the other systems and the government DSL service now has a configu page which is nice because it means we can add inspired extensions to it uh and that which means that we finally can do Dublin PS inspired compliant induced arm we are also working on a and discovery extension for multidimensional data sometimes when you have so that the data or data that has dimensions which are related with each other like in the forecast time and and prediction time are related with each other it's difficult to find combinations of values of the police provide you with an output we are working on a little specification to allow clients to explore their n-dimensional space we have a link there please have a look and tell us if you like that of yes reading have done some work in this area the group by process now and then what you have in the aggregation process now can do by and then you must say well this is boring yeah I know but it allows to a power of nice diagrams and the client from 8 in the previous request we improve the control we know can tell apart the queuing time and user cution time for asynchronous processes so we have a connection pool there and
shop processes can be stuck in the queue for for 1 so now we can control better that we made some improvements to the download process which is our community module so all of our we can now have it was already there since last year we now we can select the bands when we are extracting the cost of data and what's more interesting and this is the next slide if you're using
the process hunger to make very large destruction with a very large number of clients are we made a number of organizations to improve their ability to extract large images of scalar on on the machines like we made best Commissioner have for the cost of and 28 you bytes and we found the solemn scalability issues that you will not see on a 4 or 6 core machine so
configuration management won't if you want to call it certainly as this is some work that that we worked on it
boundless path with fields from Belgium this is the kind of a long term strategic way for us to server for the longest time has been tied to having a file system storage configuration and we introduced an API last year in order to allow GeoServer to store all the light imagery and fonts and small configuration files and have
a choice of where to store that and so we added this to the code base last year and this year we were able to implement JDBC store which lets us store icons and so on in the database and so JDBC what about 1 thing that's really nice well that is there is a REST API if you can manage your icons and fonts finally using the REST API which is great automation but there is also no and doing that currently is a community module we hope to add that into the mainline program later but this allows you to manage your icons from the Kuwaiti if you don't have direct access to the to the machine you'd use service 1 the other thing that we entered
the Victorian Office Team we as introduced a status that extension pages for the rest API and this allows us to check the GeoServer configuration which is really helpful for automation I think we're going to do the work printed work right so parameter
configuration of it happens when you are working on just over that you have several environments in which you're working like a test environment production environments sometimes there is a lot more environments and each environment has its own set of connection parameters the databases so you are you need to connect to different also different ports with different user name and password and so on right now it's a bit of a pain because you switch directly from 1 environment to the other and then you have to to fix it somehow all this connection parameters that in just over 2 thousand and we applied to Portland tonight we're going to allow you to the set variables in the configuration and then look them up from our property fine or some sort of all the configurations or sources so that you can literally just switch the configurations around and we don't we don't having to manually change the configuration but goes hand in
hand with a new model which is of a cover of community model which is going to performance will become during the curation and restored it in a different environment take into account the variable if need be you operations are going to be fully a sink controlled by a recipe I you can do that right right so you might try to but what a story without actually doing to restore but just trying it out and see if there is any problem before actually planning it and that is a full rest API to control more and the the backups and restorers so we have a URI as usual but we also have a API for automation in terms of security we have a new held up user groups service so if you are storing your user in users and held out this is a new opportunity we already have an authentication system this is a different way to interface with a written about so it's always nice to have a choice I you should know about your friends there was a presentation this morning about it we have no way to evaluate inside just as opposed to a separate survey which makes it user and we have a simple user interface to edit the rules and control their own lives and we also added the ability to control the admiral's of minerals are those that control can manage which layers so administration control all of the control of the
school in that study so the
fungal talking a little bit about some of the R&D without going on images of a community of 1 of the projects that's near and dear to my heart is that location tech project uh uh the
due date team is actually put together a GeoServer community module which allows to getting repositories to be integrated into the application I so there's a new screen there to configure geologic repositories and then you publish those held as is as individual data sources so you can actually publisher of the same information of twice 1 on the master branch 1 on like a past branch and make that available to just normal WFS clients and there is also a graduate in our web API so you can control and interact with some of this repository management staff from for automation looking ahead to the
next year we've got some R&D that we can see into the future
and 1 of the things on the horizon which actually scheduled for the constraint this weekend was looking into German compatibility I thought we were going to have to pull this off by October but it looks like the javelin release has been delayed nevertheless if you're interested in helping out this weekend please join the GeoServer reading for that we we are always welcoming volunteers the other 1 is a little bit of of a tougher problem as you well know GeoServer is a Java application and an open source application when we initially joined the equestrian community you because the little bit of heartburn because Java was not open source Java was controlled by Oracle and open JDK did not to did not exist yet and that was troubling to some members of the Austrian community with open JDK now being free and available we can now stand up GeoServer entirely on an open source platform which is amazing except for 1 small then we make use of an image processing library called Chavez advanced imaging and this was part of the original time Oracle lost Java and it is still produced under our distributed under a binary license and that means that we are not 100 per cent on an open source platform and that's something that we'd like to fix now we do know other people that really value open Source freedom and so
we're putting together a joint initiative with location tech and was still in order to create our own Rasta processing engines and this will be a major undertaking Apple foundations and we'll be taking place over the next year so we really lean on the location i pt and to make sure that this new library is really free of any kind of incumbents and will be really working hard with the GeoTools GeoServer change in order to successfully migrate all of our projects over to this question it says it's important
to note that we have to do that because of J. I besides license so is are the great library for faster processing and supports the factor loading tile-based computationally because of its own pocket can support
concurrent a type calculation and the like which is something you normally don't find in open source Java are after processing libraries so that's why we're going to but I do this work to retain all these abilities to process very large rosters without ever having to load them into memory which is of paramount importance server which is having hundreds of concurrent requests the users who who think that's
presentation how we we doing for time we have about
6 minutes for questions law if you want to ask a question please raise your hand them some of us will come with a microphone otherwise become recorded question and people in the lives of the unit cell or any questions we might be off the hook this room is always intimidating because it's such a big room that people are scared ask questions I think it's the podium 1 is next the sexual play it's a good question it was on my slides uh with the next release because we suffer this two-month delay
nevertheless we really like to keep to arguably scheduling in part because it matches the business cycle it's very handy to have a really scheduling in February because we don't have a lot of customer work slowing us down rather than continually have this two-month delay we're going to claw back there a month at a time to the next release of GeoServer is scheduled for October I believe yes and then the following release in the spring will be back on
track again they're going to do to to short cycles to a five-month cycles so 6 the
can we thank you that's very much and please keep enjoying using thank
God you
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation


  499 ms - page object


AV-Portal 3.8.0 (dec2fe8b0ce2e718d55d6f23ab68f0b2424a1f3f)