# ADL 1.5 Workbench beta 4 release - major new features **Category:** [Announcements (archive)](https://discourse.openehr.org/c/announcements-archive/155) **Created:** 2011-09-08 23:59 UTC **Views:** 7 **Replies:** 9 **URL:** https://discourse.openehr.org/t/adl-1-5-workbench-beta-4-release-major-new-features/15096 --- ## Post #1 by @thomas.beale A new beta release of the ADL 1.5 Workbench is now available [here](http://www.openehr.org/svn/ref_impl_eiffel/BRANCHES/adl1.5/apps/adl_workbench/doc/web/index.html). Quite a lot of new features have been added, as described in the [release notes](http://www.openehr.org/svn/ref_impl_eiffel/BRANCHES/adl1.5/apps/adl_workbench/doc/web/release_notes.html). Please note that this release only includes a Windows build; Linux and Mac builds will follow in the next couple of weeks. (I would be interested to know if anyone uses the Linux build...). It's still a beta, and there are bugs and deficiencies - please take the trouble to report them [here](http://www.openehr.org/issues/browse/AWBPR). Some highlights: - [New docking](http://www.openehr.org/svn/ref_impl_eiffel/BRANCHES/adl1.5/apps/adl_workbench/doc/web/images/docking_side_by_side_definitions.png) UI, enabling multiple Archetype and Class Tools at once; - [Reference Model browsing](http://www.openehr.org/svn/ref_impl_eiffel/BRANCHES/adl1.5/apps/adl_workbench/doc/web/images/rm_schema_tool.png), including any reference model, and supporting [cross-model comparisons](http://www.openehr.org/svn/ref_impl_eiffel/BRANCHES/adl1.5/apps/adl_workbench/doc/web/images/rm_schema_tool_duplex_classes.png); - Greatly enhanced [class viewer](http://www.openehr.org/svn/ref_impl_eiffel/BRANCHES/adl1.5/apps/adl_workbench/doc/web/images/explorer_class_view.png), enabling not only exploration of the static type model, but also dynamic replacements at runtime, providing a way of understanding realistic data structures; - [Search bar ](http://www.openehr.org/svn/ref_impl_eiffel/BRANCHES/adl1.5/apps/adl_workbench/doc/web/images/address_bar.png)allowing you to search for classes or archetypes; - [CKM icons](http://www.openehr.org/svn/ref_impl_eiffel/BRANCHES/adl1.5/apps/adl_workbench/doc/web/images/apgar_RM_icons.png) are now switchable in to replace the standard icons in the Archetype tool and Archetype Catalog; - beta version of [XML OPT output](http://www.openehr.org/svn/ref_impl_eiffel/BRANCHES/adl1.5/apps/adl_workbench/doc/web/images/xml_serialisation.png), based on a rule set to control details of XML serialisation (it's not yet conformant with either ADL 1.4 or 1.5 XSD - but will be soon); - the RM schema files now include proper package structures, and the [openEHR RM shcema](http://www.openehr.org/svn/knowledge2/TRUNK/rm_schemas/openehr_ehr_102.bmm) is now a faithful representation of the published specifications; - the [EN13606 model schema](http://www.openehr.org/svn/knowledge2/TRUNK/rm_schemas/cen_EN13606_0.95.bmm) has been substantially improved, although it is still not complete. Any offers to help with this are welcome! As you might guess, these features presage more powerful functionality still. In the next weeks and months, the following will be added: - archetype editing based on RM instance structure building and constraining (this is not meant to compete with other tools but rather allow people to use the AWB, which already does validation, to correct faults in archetypes, and also create new archetypes for RMs that currently have no proper editor); - single-file ADL 1.5 templates; - an ADL/AOM 1.5 XSD; - the tool XML output will be improved to support ADL 1.4 and 1.5 level XSD; - a JSON OPT generator; - an example instance generator, with dADL, JSON and XML output; - remaining ADL 1.5 semantics, including grouping and node-level documentation; - extract RM documentation from PDFs into XML DITA, and link with the RM schemas, so that the entire documentation for any class is available in one place; - connection to Seref Arikan's Bosphorous Java wizardry, enabling deployment as both a server and a web/cloud side tool for any kind of front end. The tool is in Eiffel (the most efficient language for me to develop in), and anyone who thinks they know OO properly, even if they have never seen Eiffel before, is welcome to join in. Peter Gummer, Seref and I can help them with Eiffel. (At some point, the whole thing can be fully replicated in Eclipse, Ruby, Python, etc, but for the moment, our feeling is that it is worth continuing for a while longer with a single ADL 1.5 compiler, and to develop a Java and other front-ends, while ADL 1.5 completely matures.) If there are people from national programmes and also vendors out there who find this work useful, I recommend they consider how to take advantage of the emerging new openEHR governance to help resource this kind of work better. - thomas beale --- ## Post #2 by @Seref Hi Tom, Thanks for all the hard work. I'll give it a spin under Linux and let you know how it goes. If I had a Mac (cough), I'd do the same for the sugar coated BSD ;) Regards Seref --- ## Post #3 by @tonyshannon Well done and many thanks Tom, Yes more important progress\. Thanks for all the work that is going into this key \( & indeed open source\) tool\.\. Regards Tony --- ## Post #4 by @thomas.beale you'll have to wait for the build, or else do your own\! Peter might generate one next week if he gets a minute\.\.\. \- t --- ## Post #5 by @Peter_Gummer1 Seref Arikan wrote: > Thanks for all the hard work\. I'll give it a spin under Linux and > let you know how it goes\. > If I had a Mac \(cough\), I'd do the same for the sugar coated BSD ;\) There is not much sugar on the Mac build unfortunately, Seref, because it doesn't use Cocoa\. The Mac build runs under X11 \(i\.e\., X\-Windows\) and uses GTK, similar to the Linux build\. I've built it many times on Linux and the Mac and it runs just as well as on Windows\. Apart from the fact that it looks pretty ugly under GTK on the Mac, it is very hard for users to install because they need to install X11 and MacPorts first, and then wait ages for various packages to build\. Linux users might say, "So what, that's normal", but it's completely abnormal for Mac and Windows users\. Then, having built the packages, it may not run due to the packages they installed being incompatible with the ones that I had on my machine when I built it\. DLL hell\. The best solution to this would be to forget X11 on the Mac and build it for Cocoa instead\. I did do a Cocoa build of it a few weeks ago \-\- the first time I had ever tried to do that \-\- and it does compile and run natively on the Mac\. Unfortunately, however, it didn't work well\. The EiffelVision GUI framework has been ported to Cocoa, but not completely\. The docking in particular was very bad\. It was bad enough that it's currently unusable, I'm afraid\. X11 is still the only option on the Mac\. \- Peter --- ## Post #6 by @Seref Hi Peter, We may be able to replace Eiffel Vision with something else, but that is the next step of experiments, and will take a long discussion before we get started with it\. Thanks for the explanation\! Cheers Seref --- ## Post #7 by @thomas.beale we can certainly do that \- EiffelVision is only implicated in GUI builds of the AWB\. But the AWB can be built \(as you know ;\-\) in any back\-form, e\.g\. DLL, \.so, or a running service\. The entire underlying design is predicated on this idea, so doing that is easy\. \- thomas --- ## Post #8 by @Sreedhara_R Please remove my e-mail from all your mailings lists. Thanks. Dr. Sreedhara --- ## Post #9 by @system Should be on technical list only, Sam --- ## Post #10 by @thomas.beale Er\.\.\. yes\! We were not watching the cc:s properly\. Apologies to all\. \- thomas --- **Canonical:** https://discourse.openehr.org/t/adl-1-5-workbench-beta-4-release-major-new-features/15096 **Original content:** https://discourse.openehr.org/t/adl-1-5-workbench-beta-4-release-major-new-features/15096