Validation data from VocBench showing in public ShowVoc
Karen Goes
Karen Goes
Roland Wingerter
Dear Karen,
Apparently, you have set up VB3 and Showvoc to use the same instance of Semantic Turkey. This can be convenient for practical reasons (e.g. you want to have a DEMO pack on your laptop with both systems, and you don’t want to setup different projects for them with the same content), but it is not recommended in a production environment.
In a production environment, VocBench and ShowVoc should use separate instances of ST.
For example, here is how I install VB3 and SV on my notebook.
1. Create separate folders for each app, for example:
c:\lorestar
c:\lorestar\vocbench
c:\lorestar\showvoc
2. Move the downloaded VB and SV zip files to their respective folders and unzip them there.
3. Create two links on the desktop for starting VB3 and SV by running semanticturkey.bat in their respective bin folders.
4. Turn your browser to localhost: 1979/vocbench3/ to start VB3 and localhost:1983/showvoc/ (or the server port you configured) to start SV.
Finally, when running VB3 and SV locally, I use different browsers (Chrome, Firefox, etc.) to circumvent an issue caused by fact that both SV and VB store session data in the browser. (Cf. https://groups.google.com/g/vocbench-user/c/RjFPnzaFHDI/m/XfgKDlidBAAJ).
Karen Goes
Karen Goes
.png?part=0.2&view=1)
.png?part=0.1&view=1)
stel...@uniroma2.it
Dear Karen,
my first guess was the same as Roland (thanks for that Roland). Now, if there are different instances, the next question would be “how did you load the data into ShowVoc?” and, to have a bigger picture, “how did you create the projects in the two applications?”.
Let’s start from a fact: there’s no way that, with a clean and independent installation of ShowVoc, the system writes validation data into a repository.
So, if you are having separated installations, a scenario for your case could be this one:
- VB and SV are separated, but you are sharing the same instance of remote triple store
- Indeed, when you create projects, if you give the same name to their respective projects, they would use (by default) the same repository name. This, however, would send an error msg to the user saying that there’s already an existing repository with that name, so the user may change project name or simply customize the repository name for that project
- So, if we are in this scenario, the probable hypothesis is that you explicitly chose to use the same repository upon project creation
Could you let us know about the above?
Kind Regards,
Armando
P.S: for simplicity, you could run two instances of your triple store as well, so to be sure to keep the two systems completely isolated. It’s not something inconvenient, and if you were running a production server it would even be a good practice, using a triple store for potentially heavy usage but accessed by less users (VocBench’s case) from another one that could be accessed by many more users but mostly with read operations (ShowVoc’s). So the two would even require different optimizations.
P.P.S: technically speaking, it’s not VB that writes validation data. VB prepares, only in case History or Validation are being used, the triple store (by using one of the plugins deployed within it, the st-changetracking-sail-<version>.jar developed by us) to use validation for that repository. Anyhow, this is a detail not influencing the above, where it seems pretty clear that the two applications are seeing the same data, so if it’s not two clients sharing the same system, it’s probably the two systems sharing the same repository
From: vocben...@googlegroups.com <vocben...@googlegroups.com> On Behalf Of Karen Goes
Sent: Wednesday, November 19, 2025 9:11 PM
To: vocbench-user <vocben...@googlegroups.com>
Subject: [vocbench-user] Re: Validation data from VocBench showing in public ShowVoc
Dear Roland,
First of all I think I might have sent this reply before to you directly and I don't know how to undo that, so my apologies for that.
I have tried running SV and VB3 in different browsers as well as different terminals (which I also did in my oringal post). But I still see the unvalidated data in SV. Is there any way to make sure I run different Semantic Turkey instances to prevent this from happening?
Kind regards,
Karen Goes
On Wednesday, November 19, 2025 at 8:43:21 PM UTC+1 Roland Wingerter wrote:
Dear Karen,
Apparently, you have set up VB3 and Showvoc to use the same instance of Semantic Turkey. This can be convenient for practical reasons (e.g. you want to have a DEMO pack on your laptop with both systems, and you don’t want to setup different projects for them with the same content), but it is not recommended in a production environment.
In a production environment, VocBench and ShowVoc should use separate instances of ST.
For example, here is how I install VB3 and SV on my notebook.
1. Create separate folders for each app, for example:
c:\lorestar
c:\lorestar\vocbench
c:\lorestar\showvoc2. Move the downloaded VB and SV zip files to their respective folders and unzip them there.
3. Create two links on the desktop for starting VB3 and SV by running semanticturkey.bat in their respective bin folders.
4. Turn your browser to localhost: 1979/vocbench3/ to start VB3 and localhost:1983/showvoc/ (or the server port you configured) to start SV.
Finally, when running VB3 and SV locally, I use different browsers (Chrome, Firefox, etc.) to circumvent an issue caused by fact that both SV and VB store session data in the browser. (Cf. https://groups.google.com/g/vocbench-user/c/RjFPnzaFHDI/m/XfgKDlidBAAJ).
Kind regards
Roland
Karen Goes schrieb am Montag, 17. November 2025 um 10:57:19 UTC+1:
Good morning,
We are looking into ShowVoc to use for displaying our taxonomy terms online. I've been testing with it in combination with VocBench and with the history and validation option turned on it seems that even though the data isn't available in the storage before it has been validated, it can be seen in the public ShowVoc instance. This is not what we would like to happen in a public environment, as the data has not been validated yet, is there a setting I'm missing to turn this off?
Kind regards,
Karen Goes
--
You received this message because you are subscribed to the Google Groups "vocbench-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vocbench-use...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/vocbench-user/aeefa965-eb4c-4e9e-b157-b8abf551cb7an%40googlegroups.com.
Karen Goes
Karen Goes
stel...@uniroma2.it
Dear Karen,
well the idea is that VocBench is the editor, and there you have an always evolving status of the dataset, which might be unstable. Once you reach a point in which you want to deploy a stable version, you export your data into ShowVoc, while VB continues to edit its own evolving version.
How to export?
- Simple procedure is you export your data and then load it into ShowVoc (ShowVoc’s admins and dataset managers can do that). https://showvoc.uniroma2.it/doc/user/datasets_manager.jsf#loading_data
- Or you can use the ShowVoc deployer: https://vocbench.uniroma2.it/doc/user/ioext/deployers/
There will major updates to all the data architecture in the next release and a pervasive presence of metadata through the metadata registry (already available in the current version, but with limited effects on the rest of the application). Furthermore, there will be support for OpenID.
The metadata revision in new version will also be based on a revised extension of the DCAT3 vocabulary, with the possibility to define abstract datasets (e.g. AgroVoc, as a thesaurus in general), versions of the datasets, deployments of versions that can be coexisting on the same project or across different projects, and VB will allow for multiple editable forks of a same dataset.
With such major evolutions, we will try to keep up with a progressively stronger integration between the two systems (e.g. by allowing to export a dataset’s metadata specified in VB3 together with the dataset itself, onto SV, or to have the same user being authenticated through OpenID when and using their credentials for deploying from VB to SV without specifying further connection configurations).
That is a massive work and we are doing lot of effort on reaching these objectives; so, I take the occasion to apologize with all the community for the longest time ever in the story of VocBench in between two subsequent releases :-) . To our excuse, there have been several other factors that concurred to this extended development time but…more details will follow with the next release! What I assure is that it will be worth the wait.
Kind Regards,
Armando
To view this discussion visit https://groups.google.com/d/msgid/vocbench-user/d343d168-74be-4c92-9667-b3bf580254cen%40googlegroups.com.