Add SAMM - Semantic Aspect Meta Model exporter #394

Kostadin-Ivanov · 2024-08-15T09:26:48Z

This is new functionality to allow for conversion (export) of VSS specification into .ttl formatted Semantic Aspect Meta Models, which can be further used in the ESMF tools like Aspect Model Editor and SAMM CLI - for further application integrations.

sschleemilch

A lot of code to be reviewed (+2k lines). Not finished yet but first feedback provided 👯

sschleemilch · 2024-08-19T07:35:35Z

src/vss_tools/vspec/vssexporters/vss2samm/helpers/fileHelper.py

+
+ log.debug("Writing graph to \n -- file: '%s' \n -- location: '%s'", file_name, path_to_file)
+
+ log.debug("Current working directory: '%s'", os.getcwd())


vss_tools tries to use the more modern Pathlib instead of os.* methods

In this case: Path.cwd()

Can we use the Pathlib to create folders and files dynamically, because I use it further in this script.

sschleemilch · 2024-08-19T07:38:01Z

src/vss_tools/vspec/vssexporters/vss2samm/vss2samm.py

+ minor = 0
+ patch = 0
+
+ if vss_tree and hasattr(vss_tree, "children") and len(vss_tree.children) > 0:


It is save to call just if vss_tree.children for an Anytree node

sschleemilch · 2024-08-19T07:46:15Z

src/vss_tools/vspec/vssexporters/vss2samm/vss2samm.py

+# Exporter specific options
+
+@clo.option(
+ '-sigf', '--signals-file', 'signals_file', show_default=False,


Until now we followed the posix strategy of single dash options being single characters. Not sure if we should change that here. The signals_file parameter is not needed. show_default is not needed if there is no default

Since the input is a file, check how to define it propery with cli_options as a reference using Pathlib.
It has the benefit that existence and permissions can be checked on cli level

sschleemilch · 2024-08-19T07:48:29Z

src/vss_tools/vspec/vssexporters/vss2samm/vss2samm.py

+)
+@clo.option(
+ "--split-depth", "-spld", "split_depth",
+ required=False, default=1, show_default=False,


required=False is the default, so not needed. split_depth not needed. type should be set to int

sschleemilch · 2024-08-19T07:49:13Z

src/vss_tools/vspec/vssexporters/vss2samm/vss2samm.py

+)
+@clo.option(
+ "--split", "-spl", "split",
+ required=False, is_flag=True, default=False, show_default=False,


For flags it can be useful to define them with option("--split/--no-split", default=False). It has is_flag implicitly and provides two quite concise options

sschleemilch · 2024-08-19T11:20:50Z

src/vss_tools/vspec/vssexporters/vss2samm/helpers/sammConcepts.py

+ return f'{self.ns}{self.value}'
+
+
+class SammCConcepts (Enum):


Looks pretty similar if not the same as SammConcepts Maybe define a common parent class?

The SammConcepts and SammCConcepts use different uri's (namespaces):

SammConcepts has: self.ns = f"{samm_base_namespace}:meta-model:{cnfg.SAMM_VERSION}#"

where

SammCConcepts has: self.ns = f"{samm_base_namespace}:characteristic:{cnfg.SAMM_VERSION}#"

I will see how can I combine these and override just their initialization part with these namespaces.

Thanks for the catch :)

sschleemilch · 2024-08-19T11:25:26Z

src/vss_tools/vspec/vssexporters/vss2samm/helpers/vssHelper.py

+# which then can be further manipulated.
+#
+# For example: filter the children of cloned VSSNode and add them to its cloned (copied) version.
+# REASON FOR THIS is that:


I don't think this is the case when e.g. using copy.deepcopy(). However, the performance of deepcopy is not great since it will essentially copy the complete tree including all parents and therefore all children etc. This is one of the reasons I implemented an own copy method: https://github.com/COVESA/vss-tools/pull/382/files#diff-156f359c987ddf2af79c8b9bf50be3c1899d7a3e389865b0fe7b89812d62bc51R69.
Maybe we will merge the refactoring this week. That will simplify a lot of things but will cause conflicts with this one unfortunately

I was not aware of this.
Reason for this copy was explained in the comment above this: clone_vss_node function.
This was needed when this exporter was created as external tool which was using provided by vss-tools functionality so I am open for improvements if there is a way to reduce this effort for filtering of selected signals / branches etc.

with regards to possible conflicts, no worries, I will deal with them :)

sschleemilch · 2024-08-19T11:25:54Z

src/vss_tools/vspec/vssexporters/vss2samm/helpers/vssHelper.py

+ # which was used for the creation of vss_node,
+ # i.e. its corresponding name in the used for creation of VSSNode, dictionary.
+ properties_to_clone = {
+ "$file_name$" : "$file_name$", # noqa: E203


Might be simpler to specify things not to clone? But yeah, see comment before

As per my previous comment I was trying to keep only the fields which my script was going to use, instead of copying whole VssNode, but again as I said I am OK for further refactoring of this and if possible to build "filtered" tree in different way I can even remove it.

The "cloning" and related logic is now removed.

Done

sschleemilch · 2024-08-19T11:26:30Z

src/vss_tools/vspec/vssexporters/vss2samm/helpers/vssHelper.py

+ or (vss_node.is_leaf
+ and hasattr(vss_node, 'datatype')
+ and vss_node.datatype is VSSDataType.BOOLEAN
+ and (vss_node.name.startswith('Is') or vss_node.name.startswith('is'))


vss_node.name.lower().startswith("is")

sschleemilch · 2024-08-19T11:30:08Z

src/vss_tools/vspec/vssexporters/vss2samm/vss2samm.py

+
+# TODO: Currently this is a workaround to read the Vehicle.VersionVSS, which is provided from COVESA/VSS
+# and provides the supported VSS version.
+# In best case this functionality should be provided by the vspec tool, loaded in this script.


You mean by a method in the common part of vss_tools right? Something like:

def get_vss_version(root: VSSNode) -> tuple[int]: ... return major, minor, patch

Yes.
As per my previous comments, this converter was built as external to vss-tools script and I could not find way to get version of current VSS tree like: vss_node.get_vss_version(), which then will return either string in the form: major.minor.patch or a tuple as you suggest.

sschleemilch · 2024-08-20T14:27:58Z

The major refactoring got merged. Let me know if you need help adapting to it

SebastianSchildt · 2024-08-27T14:32:25Z

Meeting 08/27

needs reabse
rdflib already in pyproject.toml (generally all deps?)

Kostadin-Ivanov · 2024-08-27T15:35:30Z

I was on holidays, and now I am constantly checking for master updates and rebasing if there is any.
Sorry for the delayed response on this.

I am now keep updating the code as per your feedback.

SebastianSchildt · 2024-09-02T21:44:50Z

You seem to be mostly dying by our linter enforcing stricter code quality checks. Note the section about pre-commit install in the main Readme: https://github.com/COVESA/vss-tools?tab=readme-ov-file#pre-commit-set-up

I found the easiest way is, installing pre-commit, "touching" the file CI complains about by adding a space somewhere (i.e. so you do have something to add/commit), do the commit, and likely the check will fail AND it will have repaired the file for you. Check whether changes are ok, add the file again, and commit...

sschleemilch · 2024-09-03T04:13:07Z

You seem to be mostly dying by our linter enforcing stricter code quality checks. Note the section about pre-commit install in the main Readme: https://github.com/COVESA/vss-tools?tab=readme-ov-file#pre-commit-set-up

I found the easiest way is, installing pre-commit, "touching" the file CI complains about by adding a space somewhere (i.e. so you do have something to add/commit), do the commit, and likely the check will fail AND it will have repaired the file for you. Check whether changes are ok, add the file again, and commit...

Or run pre-commit run --all-files

Kostadin-Ivanov · 2024-09-03T08:34:48Z

Thank you both :)

I was trying to reuse some of these pre-checks locally, but for some reason they were passing OK on my side and then failing here.

Most likely, I was missing some pre-check locally on one side and on the other, I wanted to do it in smaller updates, just not to mess something else.

Fingers crossed that I will manage to clear all of these style checks soon :)

sschleemilch · 2024-09-04T15:12:34Z

src/vss_tools/vspec/vssexporters/vss2samm/helpers/ttl_helper.py

+ # NOTE: this is a special case for the OBD branch, which children are marked as deprecated from VSS 5.0,
+ # but the branch itself is not marked as deprecated.
+ deprecated_children = []
+ if hasattr(vss_node.data, "type") and vss_node.data.type is NodeType.BRANCH:


every vssnode has type

I got this error, when I removed the hasattr check:

sschleemilch · 2024-09-04T15:23:46Z

src/vss_tools/vspec/vssexporters/vss2samm/helpers/ttl_helper.py

+ # Skip parsing of VSSNode is branches, which have only deprecated children
+ # NOTE: this is a special case for the OBD branch, which children are marked as deprecated from VSS 5.0,
+ # but the branch itself is not marked as deprecated.
+ deprecated_children = []


Isn't the whole part a single liner?

deprecated_children = list(filter(lambda n: n.data.deprecation, vss_node.children))

Thanks for the hint.
I updated the code as advised.

sschleemilch · 2024-09-04T15:25:33Z

src/vss_tools/vspec/vssexporters/vss2samm/helpers/ttl_helper.py

+ node_char_uri = ttl_builder.add_node_branch_characteristic(graph, vss_node, node_uri)
+
+ if (
+ hasattr(vss_node.data, "instances")


if it is a branch it will always have instances filled therefore if vss_node.data.instances should suffice

I was getting similar to previous screenshot VSSRaw error if I've removed the hasattr check

Done.

I updated the checks as advised.

sschleemilch · 2024-09-04T15:26:02Z

src/vss_tools/vspec/vssexporters/vss2samm/helpers/ttl_helper.py

+ graph.add((node_char_uri, SammConcepts.DATA_TYPE.uri, node_entity_uri))
+
+ # Populate Entity properties if the current vss_node holds any child node
+ if hasattr(vss_node, "children") and len(vss_node.children) > 0:


just if vss_node.children

or just start the for loop. It will be a noop if there are no children

Done.
Removed this redundant check for children length.

sschleemilch · 2024-09-04T15:27:28Z

src/vss_tools/vspec/vssexporters/vss2samm/vss2samm.py

+ttl_helper = None
+
+
+def __setup_environment(output_namespace, vspec_version, split_depth: int) -> None:


that setup function looks quite suspicious. Why do we need dynamic imports here?

The goal here was to get user input, which then can be used to initialize the script config, and later to load the: samm_concepts, vss_helper and ttl_helper which are working with some of the config fields, which depend on the user input at the time of call of the script.

Again, this was just my understanding of handling this case, but if there is a better way to do it, I will be happy to do some refactoring :)

Just use function/class arguments does not work? If not, you could have a global state such as config.py that contains objects that get filled at the script start and other modules can then import and use.

EDIT: I just noticed that you have a config.py already. Then I do not get why we need dynamic imports... They can just import config and use those variables, can't they?

There is a call to init the config with user input data:

def __setup_environment(output_namespace, vspec_version, split_depth: int) -> None: # Initialize config before to load other helpers / libraries, based on defined by user input, config data cfg.init(output_namespace, vspec_version, split_depth)

without this, I was getting errors for undefined variables in the dynamically loaded _helper(s)

Sorry for the "code" but for some reason I could not upload a screenshot.

sschleemilch · 2024-09-04T15:30:26Z

src/vss_tools/vspec/vssexporters/vss2samm/vss2samm.py

+
+ if vss_tree.children and len(vss_tree.children) > 0:
+ # Get the VersionVSS node so to extract current VSS version from it.
+ vss_version_node = next(filter(lambda vs_child: vs_child.name == "VersionVSS", vss_tree.children), None)


There is also a tree method which might be what you want:

vss_tree.get_child(vss_tree.get_fqn() + ".VersionVSS")

sschleemilch · 2024-09-04T17:04:18Z

I am sorry for the opinionated Feedback 😄

erikbosch · 2024-09-05T10:55:41Z

Hello @erikbosch, I am not sure how long it might take to prepare such tests. Could we do it as a post-release update / improvement?

So far, I provided a samm.md with more information about this exporter, some user guide and examples how to validate generated aspect models.

With regards to validation - briefly, this can be done using the ESMF CLI tool with a simple command like:

samm aspect vss_ttls/com.covesa.vss.spec/5.0.0/Vehicle.ttl validate

The ESMF - Aspect Model Editor (AME) can also be used for manual validation of generated aspect models.

Yes, it could be done in a separate PR as I see it. The important part as I see it is that we long term get some test cases so we detect regressions (if any) when introduced.

erikbosch · 2024-09-05T10:58:06Z

docs/samm.md

@@ -0,0 +1,220 @@
+# Vspec - Semantic Aspect Meta Model (SAMM) exporter


Great addition - now you have a better documentation that most other tools :-)

Thank you for the hint on this :)

erikbosch · 2024-09-05T11:13:48Z

Tested the tool on one of the test vspecs we have. It gives an error because of some exception, most likely some syntax for allowed which is not supported yet. As this construct is not (yet) used in std catalog I do not see it as critical, even if error message possibly could be improved.

erik@debian6:~/vss-tools/tests/vspec/test_allowed$ vspec export samm -u ../test_units.yaml -q ../test_quantities.yaml --vspec test.vspec --target-folder xxx
[13:09:37] INFO     Loading VSS Tree...                                                                                                                                                                       vss2samm.py:182
                                                                                                                                                                                                                             
           INFO     Loaded 'VSSQuantity', file=/home/erik/vss-tools/tests/vspec/test_allowed/../test_quantities.yaml, elements=6                                                                       units_quantities.py:34
           INFO     Loaded 'VSSUnit', file=/home/erik/vss-tools/tests/vspec/test_allowed/../test_units.yaml, elements=7                                                                                units_quantities.py:34
           INFO     VSpecs loaded, amount=1                                                                                                                                                                      vspec.py:122
[13:09:38] INFO     No signals selected.                                                                                                                                                                      vss2samm.py:212
                    Creating model for the whole tree.                                                                                                                                                                       
                                                                                                                                                                                                                             
           INFO     Update output: 'xxx' with ESMF namespace: 'com.covesa.vss.spec' and VSS Version: '1.0.0'.                                                                                                 vss2samm.py:214
                                                                                                                                                                                                                             
           INFO     Generating SAMM output...                                                                                                                                                                 vss2samm.py:224
                                                                                                                                                                                                                             
           ERROR                                                                                                                                                                                              vss2samm.py:280
                    There was a problem with VSS to ESMF processing.                                                                                                                                                         
           ERROR                                                                                                                                                                                              vss2samm.py:281
                    Error: unsupported operand type(s) for +: 'int' and 'str'                                                                                                                                                
erik@debian6:~/vss-tools/tests/vspec/test_allowed$

Kostadin-Ivanov · 2024-09-05T11:14:32Z

I am sorry for the opinionated Feedback 😄

No worries, there is nothing to apologize for and thank you for the good feedback and hints :)

I am currently implementing all of your latest feedback and hope to be ready for final review by end of today :)

Kostadin-Ivanov · 2024-09-05T11:30:46Z

Tested the tool on one of the test vspecs we have. It gives an error because of some exception, most likely some syntax for allowed which is not supported yet. As this construct is not (yet) used in std catalog I do not see it as critical, even if error message possibly could be improved.

erik@debian6:~/vss-tools/tests/vspec/test_allowed$ vspec export samm -u ../test_units.yaml -q ../test_quantities.yaml --vspec test.vspec --target-folder xxx
[13:09:37] INFO     Loading VSS Tree...                                                                                                                                                                       vss2samm.py:182
                                                                                                                                                                                                                             
           INFO     Loaded 'VSSQuantity', file=/home/erik/vss-tools/tests/vspec/test_allowed/../test_quantities.yaml, elements=6                                                                       units_quantities.py:34
           INFO     Loaded 'VSSUnit', file=/home/erik/vss-tools/tests/vspec/test_allowed/../test_units.yaml, elements=7                                                                                units_quantities.py:34
           INFO     VSpecs loaded, amount=1                                                                                                                                                                      vspec.py:122
[13:09:38] INFO     No signals selected.                                                                                                                                                                      vss2samm.py:212
                    Creating model for the whole tree.                                                                                                                                                                       
                                                                                                                                                                                                                             
           INFO     Update output: 'xxx' with ESMF namespace: 'com.covesa.vss.spec' and VSS Version: '1.0.0'.                                                                                                 vss2samm.py:214
                                                                                                                                                                                                                             
           INFO     Generating SAMM output...                                                                                                                                                                 vss2samm.py:224
                                                                                                                                                                                                                             
           ERROR                                                                                                                                                                                              vss2samm.py:280
                    There was a problem with VSS to ESMF processing.                                                                                                                                                         
           ERROR                                                                                                                                                                                              vss2samm.py:281
                    Error: unsupported operand type(s) for +: 'int' and 'str'                                                                                                                                                
erik@debian6:~/vss-tools/tests/vspec/test_allowed$

@erikbosch - could you try to run it again. I removed the try/catch blocks so we should see the whole error + stacktrace.

I'd say that there is some problem with the test.vspec but I don't know its contents.

Kostadin-Ivanov · 2024-09-05T11:37:08Z

A lot of code to be reviewed (+2k lines). Not finished yet but first feedback provided 👯

We've been working on this script as "external" to vss-tools tool for almost year and a half :)

sschleemilch · 2024-09-05T11:51:35Z

I removed the try/catch blocks so we should see the whole error + stacktrace.

Now you know what I meant hehe

sschleemilch · 2024-09-05T11:59:05Z

src/vss_tools/vspec/vssexporters/vss2samm/helpers/ttl_builder_helper.py

+#
+
+
+def __add_node_tuple(graph: Graph, subject_uri: URIRef, predicate_uri: URIRef, object_data: Literal | URIRef) -> bool:


Might this be a candidate to be an own exporter? Like vspec export ttl?

As far as I remember, there used to be a ttl exporter in the contrib folder, but I think that this is obsolete and it was generating different formatted ttls.

Do you think it would be reasonable to separate it into an exporter that exports it correctly?
Maybe we can reduce the complexity of that exporter then a bit? Does not need to happen now of course...

Like I said, there is already this vss2ttl which seems to do this, but when I look at its code, it seems to be a different type of ttl format.

Both helpers: ttl_helper and ttl_builder_helper are quite bound to the SAMM formatted ttls. for this reason I am not 100% how easy would be to make these more generic so they can be used for more general purpose and why not as a stand alone ttl exporter.

Maybe as another improvement, we can think about such refactoring as further improvement.
It would be interesting if then we can reuse functionality of one exporter by another exporter, but we will see if we go in this direction.

By the way, just to add, I reused the try/catch, which is no longer present in the vss2samm, from there :)

Signed-off-by: Kostadin Ivanov (BD/TBC-BG) <kostadin.ivanov@bosch.com>

Kostadin-Ivanov · 2024-09-09T12:01:48Z

Hello Sebastian (@sschleemilch) and Erik (@erikbosch),
could you please have a lok at my responses to your feedback and "resolve" them if your are happy with my updates or add your additional thoughts if I've missed something.

Also, could you tell me what will be next step, if you are happy with the updates.
You will merge this PR to main branch or I should do it?

And one more thing, after we merge this PR, when it will be released in the master VSS-TOOLS and would it be updated in the COVESA VSS as well?

Thank you :)

erikbosch

I consider this one good to go now. later changes, if any, can be performed by subsequent PRs.

erikbosch · 2024-09-10T11:20:24Z

Hello Sebastian (@sschleemilch) and Erik (@erikbosch), could you please have a lok at my responses to your feedback and "resolve" them if your are happy with my updates or add your additional thoughts if I've missed something.

Also, could you tell me what will be next step, if you are happy with the updates. You will merge this PR to main branch or I should do it?

And one more thing, after we merge this PR, when it will be released in the master VSS-TOOLS and would it be updated in the COVESA VSS as well?

Thank you :)

I think this one is good to go now. Only maintainers have write rights and can merge, so I will click the merge button. Then it will show up on master branch. We are planning a 5.0 release soon, but have not yet agreed on exactly when. If needed we could update the submodule reference from vss to include this one, and then possibly integrate into the CI and Makefile for vss as well, to make sure that it still pass.

After merge it would be good if you could add at least a single test case that verifies that the tool runs and produces expected output. That would help us detect future regressions, if any

Kostadin-Ivanov · 2024-09-10T13:28:13Z

Thank you very much @erikbosch .

I will take note on the further updates / improvements / tasks, that we identified during this PR discussion and make sure that we follow on them.

With regards to the tests, I will have a look at how other tests are organized and try to add some.

Thanks again to everybody for your support :)

Kostadin-Ivanov force-pushed the add_vss2samm branch 6 times, most recently from ad4856c to 5ebbf86 Compare August 16, 2024 09:03

sschleemilch reviewed Aug 19, 2024

View reviewed changes

Kostadin-Ivanov force-pushed the add_vss2samm branch 4 times, most recently from 056488f to 85e7f70 Compare August 20, 2024 13:19

Kostadin-Ivanov force-pushed the add_vss2samm branch from 85e7f70 to 9dba80e Compare August 20, 2024 15:44

Kostadin-Ivanov force-pushed the add_vss2samm branch 3 times, most recently from b513803 to 1c3d33d Compare August 27, 2024 15:31

Kostadin-Ivanov force-pushed the add_vss2samm branch 7 times, most recently from d0b68fa to be3bdf8 Compare September 2, 2024 15:37

Kostadin-Ivanov force-pushed the add_vss2samm branch 2 times, most recently from 428119c to 1a2a13e Compare September 3, 2024 06:08

sschleemilch reviewed Sep 4, 2024

View reviewed changes

Kostadin-Ivanov force-pushed the add_vss2samm branch 3 times, most recently from c5cccbf to 1df07d4 Compare September 5, 2024 10:00

erikbosch reviewed Sep 5, 2024

View reviewed changes

Kostadin-Ivanov force-pushed the add_vss2samm branch from 1df07d4 to b185813 Compare September 5, 2024 11:27

Kostadin-Ivanov force-pushed the add_vss2samm branch from d0802d5 to d5af4ec Compare September 5, 2024 11:35

Kostadin-Ivanov force-pushed the add_vss2samm branch from d5af4ec to 0e84663 Compare September 5, 2024 11:39

sschleemilch reviewed Sep 5, 2024

View reviewed changes

Kostadin-Ivanov force-pushed the add_vss2samm branch 2 times, most recently from fa3c890 to 90c4b76 Compare September 5, 2024 12:36

Add SAMM - Semantic Aspect Meta Model exporter

dc1cfbf

Signed-off-by: Kostadin Ivanov (BD/TBC-BG) <kostadin.ivanov@bosch.com>

Kostadin-Ivanov force-pushed the add_vss2samm branch from 90c4b76 to dc1cfbf Compare September 5, 2024 14:54

erikbosch approved these changes Sep 10, 2024

View reviewed changes

erikbosch merged commit b898562 into COVESA:master Sep 10, 2024
5 checks passed


		log.debug("Writing graph to \n -- file: '%s' \n -- location: '%s'", file_name, path_to_file)

		log.debug("Current working directory: '%s'", os.getcwd())

		ttl_helper = None


		def __setup_environment(output_namespace, vspec_version, split_depth: int) -> None:

		@@ -0,0 +1,220 @@
		# Vspec - Semantic Aspect Meta Model (SAMM) exporter

		#


		def __add_node_tuple(graph: Graph, subject_uri: URIRef, predicate_uri: URIRef, object_data: Literal \| URIRef) -> bool:

Add SAMM - Semantic Aspect Meta Model exporter #394

Add SAMM - Semantic Aspect Meta Model exporter #394

Conversation

Kostadin-Ivanov commented Aug 15, 2024 • edited Loading

sschleemilch left a comment • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Kostadin-Ivanov Aug 20, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

sschleemilch commented Aug 20, 2024

SebastianSchildt commented Aug 27, 2024

Kostadin-Ivanov commented Aug 27, 2024

SebastianSchildt commented Sep 2, 2024

sschleemilch commented Sep 3, 2024

Kostadin-Ivanov commented Sep 3, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

sschleemilch Sep 5, 2024 • edited Loading

Choose a reason for hiding this comment

Kostadin-Ivanov Sep 5, 2024 • edited Loading

Choose a reason for hiding this comment

sschleemilch Sep 4, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

sschleemilch commented Sep 4, 2024

erikbosch commented Sep 5, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

erikbosch commented Sep 5, 2024

Kostadin-Ivanov commented Sep 5, 2024 • edited Loading

Kostadin-Ivanov commented Sep 5, 2024

Kostadin-Ivanov commented Sep 5, 2024

sschleemilch commented Sep 5, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Kostadin-Ivanov Sep 5, 2024 • edited Loading

Choose a reason for hiding this comment

Kostadin-Ivanov commented Sep 9, 2024

erikbosch left a comment

Choose a reason for hiding this comment

erikbosch commented Sep 10, 2024

Kostadin-Ivanov commented Sep 10, 2024

Kostadin-Ivanov commented Aug 15, 2024 •

edited

Loading

sschleemilch left a comment •

edited

Loading

Kostadin-Ivanov Aug 20, 2024 •

edited

Loading

sschleemilch Sep 5, 2024 •

edited

Loading

Kostadin-Ivanov Sep 5, 2024 •

edited

Loading

sschleemilch Sep 4, 2024 •

edited

Loading

Kostadin-Ivanov commented Sep 5, 2024 •

edited

Loading

Kostadin-Ivanov Sep 5, 2024 •

edited

Loading