SEMS web page

The wordpress site has been serialised to static web content and is now available through sems.bio.informatik.uni-rostock.de. All sems.uni-rostock.de/.* links should send proper HTTP redirects, so your browser will automatically guide you to the new location.

Trac

Everything trac-related is now available through GitHub. There are plenty of repositories at github.com/SemsProject, the trac-wikis and tickets are converted to GitHub pages at semsproject.github.io/.
The webserver should be configured to properly redirect individual projects, wikis, tickets etc… Thus, bookmarks and scripts should be working fine.

Compilation Products

During SEMS’ life time we created quite some binaries. People and pipelines are relying on their availability. However, they also needed to move and are now available here:

• Maven artifacts moved to mvn.bio.informatik.uni-rostock.de (used to be mvn.sems.uni-rostock.de)
• Binaries moved to bin.bio.informatik.uni-rostock.de (used to be bin.sems.uni-rostock.de)
• JavaDocs moved to doc.bio.informatik.uni-rostock.de (used to be jdoc.sems.uni-rostock.de)

Again, webservers are configured to redirect people and pipelines automagically!

Maven

We continue developing our tools and we will provide them through Maven. As described above, our Maven repository has moved to mvn.bio.informatik.uni-rostock.de.
But even better: we’re now distributing our tools through Maven central! No need to integrate a third party repository anymore :)
However, we switched the group id. It is now de.uni-rostock.sbi – so update your POMs. But don’t worry, you don’t need to update all the package names, the classes are still packaged as de.unirostock.sems.*. Find our artifacts published under de.uni-rostock.sbi at Maven Central.
Maven central will only contain new artifacts. The repository at mvn.bio.informatik.uni-rostock.de will still deliver previous artifacts, and mirror Maven central for the new artifacts. Thus, feel free to keep our third party repository integrated in your workflows.

Tools

Online tools, demos and prototypes have moved as well:

• The BiVeS web application is now running at https://bives.bio.informatik.uni-rostock.de. We installed a Proxy at the old URL (bives.sems.uni-rostock.de) that will invisibly hand over incoming tasks, so whoever integrated BiVeS does not need to worry. You should still switch to the new URL, just to make sure… ;-)
• The CombineArchive Web interface already moved to https://cat.bio.informatik.uni-rostock.de/ a year ago. Projects, cookies, and files have been migrated and the webserver properly redirects traffic.
• The CaRo interface is now available at https://caro.bio.informatik.uni-rostock.de/. A proxy is forwarding jobs, similar to the BiVeS web application above.
• BudHat moved to https://budhat.bio.informatik.uni-rostock.de/. The dev version is available at https://budhat-dev.bio.informatik.uni-rostock.de/. Redirects are installed. However, there are already many real-life tools that showcase the power of BiVeS, so you BudHat only survived because of paper-links etc..
• COMODI moved to https://comodi.bio.informatik.uni-rostock.de/latest/ already some time ago.
• MoSt is now available at https://most.bio.informatik.uni-rostock.de.
• M2CAT is still an issue, as it strongly relies on Masymos and Morre. However, there is no documentation on how to properly setup those tools :(
  As soon as I find time to reverse engineer the stuff, M2CAT will properly be available at https://m2cat.bio.informatik.uni-rostock.de/.

By the way, did you recognize that we also deployed proper SSL certificates for all services? :)

Cool, but how…

..did I just do it? Maybe it’s interesting for one or the other, or maybe you need to archive/migrate your $stuff as well? Therefore a quick overview on how I’m handling different things.

WordPress

This website used to be a WordPress instance. It has been serialised to static HTML pages using HTTrack and a bit manual fine-tuning – see also Archiving a Website.

Services that can handle HTTP redirects

Services that just handle default HTTP traffic now reply with a permanent redirect (HTTP/1.1 301 Moved Permanently). That’s realised using tiny Nginx vhosts. For example, bin.sems.uni-rostock.de can now be found at https://bin.bio.informatik.uni-rostock.de/, and the following definition makes sure people are forwarded to the proper location:

Thus, a request on http://bin.sems.uni-rostock.de/CombineArchiveWeb/CombineArchiveWeb-0.4.9.war will be forwarded to https://bin.bio.informatik.uni-rostock.de/CombineArchiveWeb/CombineArchiveWeb-0.4.9.war. Try it yourself:

Services that can’t handle 301

The above is probably the preferred approach, as people/browsers/search-engines won’t recognize anything and the old URLs will smoothly disappear (they won’t be copied anymore etc).
However, not every tool/pipeline is able to work with 301 redirects… For example, the BiVeS web application accepts POST requests. Tools that send a POST to bives.sems.uni-rostock.de may also receive a 301 that BiVeS moved to bives.bio.informatik.uni-rostock.de, and they may look at the new URL if you’re lucky, but they (most certainly) won’t resent the POST data to the new location. Thus, all tools that integrate BiVeS will not work anymore… :(
Thus, I’m not responding with a 301, but the request will just be proxied to the new location. For example, the Nginx configuration for BiVeS looks like this:

This, however, is a bit sub-optimal as users won’t realised that BiVeS has moved…
That means I need to make sure this URL stays up and running!

Source code and wikis and tickets and stuff

Our git and trac stuff was migrated to GitHub (https://github.com/SemsProject/) and GitHub Pages (https://semsproject.github.io/). That has been made possible by Martin Peters, who developed a https://github.com/FreakyBytes/TracMigrator.

Monitoring

I’m a big fan of monitoring. Thus I started the migration with monitoring of web content using check_links. For example, I wrote a check to see if the page behind https://sems.bio.informatik.uni-rostock.de/projects/bives/ still contains the phrase highlighted reaction network. So a went around the many different sites and tools to define almost 100 different checks. This way, my monitoring solution would notify me as soon as a page/tool is not properly migrated and I can be sure that the HTTP redirects for all the URLs forward to the correct final locations (also with proper content negotiation for COMODI etc).
These checks are now properly integrated in my monitoring infrastructure, so that I can continuously be sure that links (in papers for example) stay valid.

Help from the whale

In the process of migration I encapsulated almost everything in Docker containers. This way it’s easier to handle the (legacy) content and a future migration will hopefully be much quicker.

But where is…?

Did we forget something? Is some tool not working properly anymore?
If you experience any problems please drop me an email.

SEMS peeps will continue their fights for improved management of computational models on various fronts:

• Dagmar temporarily tries the real life outside academia.
• Fabienne continues to do her PhD in Business Informatics.
• Mariam remains in Rostock, but moved on to revolutionise machine learning.
• Vasundra moved to Trondheim to work on causal statements and ontologies.
• Martin and Tom managed to land a new project at the SBI in Rostock to continue work on the versioning front in collaboration with groups from Munich and Heidelberg.
• Ron works towards finishing his PhD in a denbi project in cooperation with Heidelberg.

To keep our results and project code available in the future, we are moving most of our stuff to Github: semsproject.github.io/. Martin will try to keep most of the tools, prototypes, and demos alive, but probably under different domains. For example, the BiVeS web application already moved to bives.bio.informatik.uni-rostock.de and the Combine Archive Web tool is available from cat.bio.informatik.uni-rostock.de. We try to properly redirect browsers and tools to the new locations, but let us know if anything goes wrong.

If you have doubts/questions/miss us do not hesitate to drop SEMS people an e-mail. We’ll be more than happy to hear from you.

Also check our Outro.

To run the application just call the following command:

This will bind your host’s port 1234 to the interface of the CombineArchive Web application in the container and you’ll be able to access the web interface at http://localhost:1234. There are a few options to run and configure the container, just have a look at the binfalse/webcat Docker Hub repository.

Thus, setting up the web interface was never easier nor cleaner. Plus: You will increase the privacy of your users and may get better performance from your web servers. :)

Let us know if you have problems with the setup.

The Docker image is available from the Docker Hub as binfalse/bives-webapp. To pull the latest version of the BiVeS web application just call

Running the tool is almost as simple pulling it:

This call will bind the container’s web server to your machine’s port 1234. Thus, if you open your web browser and go to http://localhost:1234 you will already start talking to BiVeS!
To verify that BiVeS is working correctly, you can try to compare two files using the following curl command:

Replace http://localhost:1234 with the settings of your environment if you didn’t run the above command lines on your local machine.

We updated our maven configuration to automagically create and publish new docker image with every release of the BiVeS web application.

We have created an online resource for the characterization and comparison of electrophysiological cell models in a wide range of experimental scenarios. The details of the mathematical model (quantitative assumptions and hypotheses formulated as ordinary differential equations) are separated from the experimental protocol being simulated. Each model and protocol is then encoded in computer-readable formats. A simulation tool runs virtual experiments on models encoded in CellML, and a website (https://chaste.cs.ox.ac.uk/WebLab) provides a friendly interface, allowing users to store and compare results.

Read more at: http://www.cell.com/biophysj/references/S0006-3495%2815%2904753-0

Export RO from WebCAT

We are happy to announce that the CombineArchive WebInterface is now able to export Research Objects!

You just need to click the [Export RO] button, as shown in the image.
That’s been made possible with the help of CaRo and the CaRoWeb tools.

?remote= downloads containers from other web servers

In addition to the famous HTTP-POST-conversion you can now supply a link to a container that’s to be converted, e.g.:

the value supplied to the ?remote= gets interpreted as a link to a container (COMBINE archive or Research object) on a remote server. It will be downloaded and the conversion will be triggered. Thus, you do not need to download a COMBINE archive yourself, but you can point to an archive on the Internet! :)
That means, if you want to convert the archive behind http://scripts.sems.uni-rostock.de/getshowcase.php you can simply curl for:

?redirect=checkout immediately delivers the container

We understand that people do not always want to parse needless JSON code. Thus, if you add ?redirect=checkout to your query you’ll get redirected to the actual container!

Without redirect:

you need to parse the JSON to retrieve the container from http://caro.sems.uni-rostock.de/checkout/CombineArchiveShowCase.omex4508850515417520827-converted-2015-09-11-23-17-PM+0200.ro

Using the latest version of the web interface you just need to append redirect=checkout and you’ll get the container:

Give it a try and open the link: http://caro.sems.uni-rostock.de/caro/caro/?remote=http://scripts.sems.uni-rostock.de/getshowcase.php&redirect=checkout

Read more about CaRoWeb and its options at the project page!

At gallery.sems.uni-rostock.de you’ll find images, logos, and figures that we created. If you click on an image you’ll get more information about it.
The number of images is currently fairly small, but will hopefully be increasing in the future.

Martin has already been there for two weeks and also gave a talk last week at a research seminar (slides).

BiVeS as integrated in PMR2

One of our main goals at HARMONY 2015 was to improve the version control of PMR2, as we described earlier. During HARMONY, Tommy Yu, main developer of PMR2, and Martin Scharm already achieved preliminary results.

Today Tommy sent us the first version of BiVeS integrated into the CellML model repository:

go here: http://staging.physiomeproject.org/workspace/beeler_reuter_1977/file/cb090c96a2ce627457b14def4910ac39219b8340/beeler_reuter_1977.cellml

Then click the ‘Compare …’ button. Do the same with: http://staging.physiomeproject.org/w/tommy/beeler_reuter_1977_uncertexample/file/d26c870944e07f00950fd671a8eb73663e96672c/beeler_reuter_1977.cellml

Now both files should be selected, just pick one as source and the other as target, hit ‘Compare Source with Target’ and watch magic happens courtesy of Martin’s BiVeS WS tool.

Tommy is using the BiVeS web service which he had deployed to one of his servers. This is still a feature in testing (running on staging servers), but will hopefully soon be available on the main repository.

Tommy, you did an excellent job! :)