Add new PyChop version and create model versioning #16

RastislavTuranyi · 2025-01-17T16:34:52Z

Resolves the PyChop issue documented at Mantid#38591

Simultaneously, applies new versioning scheme for models, where all models are versioned with _v{int} at the end of their names to keep record of model changes, and aliases are created without the version to provide seamless access to the latest version of models.

Tests will be failing until the changes are also applied to https://github.com/mducle/pychop

The model properties will not work yet

src/resolution_functions/instrument.py

ajjackson · 2025-01-20T11:06:21Z

Can we xfail the appropriate tests so we get a green tick and verify nothing else was broken?

RastislavTuranyi · 2025-01-20T12:56:41Z

I've changed the tests so that the old tests test the PyChop_fit_v1 version specifically, and added new tests for testing PyChop_fit_v2 which are mostly marked as XFAIL.

src/resolution_functions/instrument.py

oerc0122 · 2025-01-22T12:16:17Z

src/resolution_functions/instrument.py

+    longest = np.array(longest) + padding
+    out = []
+    separator = '|' + '|'.join(['-' * (i + 1) for i in longest]) + '|'
+
+    for i, line in enumerate(contents):
+        new_line = '| '
+        for val, length in zip(line, longest):
+            if val is None:
+                new_line += ' ' * length + '| '
+            else:
+                new_line += val + ' ' * (length - len(val)) + '| '
+
+        if line[0] is not None:
+            if i == 1:
+                out.append(separator.replace('-', '='))
+            else:
+                out.append(separator)
+
+        out.append(new_line)
+
+    out.append(separator)
+
+    return '\n'.join(out)


Maybe it's just my functional-programming brain, but you can avoid a lot of this fuss by using Python's format functions. It needs some hackery to get the same table, but a similar table could be done without incrementing padding and preproc, different formats, etc. sep_nl is just because you can't have \ in an f-string.

Suggested change

longest = np.array(longest) + padding

out = []

separator = '|' + '|'.join(['-' * (i + 1) for i in longest]) + '|'

for i, line in enumerate(contents):

new_line = '| '

for val, length in zip(line, longest):

if val is None:

new_line += ' ' * length + '| '

else:

new_line += val + ' ' * (length - len(val)) + '| '

if line[0] is not None:

if i == 1:

out.append(separator.replace('-', '='))

else:

out.append(separator)

out.append(new_line)

out.append(separator)

return '\n'.join(out)

padding = padding + 1

preproc = [["-" if elem is None else elem for elem in row] for row in contents]

sep_line = f"|{'|'.join('-' * (i+padding) for i in longest)}|"

header_sep_line = sep_line.replace("-", "=")

sep_nl = "\n" + sep_line

data_line = "|" + "|".join(f"{{:^{length+padding}}}" for length in longest) + "|"

formatted = map(lambda x: data_line.format(*x), preproc)

out = f"""\

{sep_line}

{next(formatted)}

{header_sep_line}

{sep_nl.join(formatted)}

{sep_line}

"""

return out

These are not quite equivalent (though I am open to discuss the best way to format this nicely) - I expect it will be possible to achieve similar formatting (if desired) using this new method as well, but I am unsure I will be able to get there easily by myself.

Old implementation (using a fake example):

|------------------|------------------|--------------------| | MODEL | ALIAS FOR | CONFIGURATIONS | |==================|==================|====================| | PyChop_fit | PyChop_fit_v2 | | |------------------|------------------|--------------------| | PyChop_fit_v1 | | chopper_package | |------------------|------------------|--------------------| | PyChop_fit_v2 | | chopper_package | |------------------|------------------|--------------------| | test_model | test_model_v4 | | |------------------|------------------|--------------------| | test_model_v1 | | config1 | | | | config2 | |------------------|------------------|--------------------| | test_model_v2 | | config1 | | | | config2 | |------------------|------------------|--------------------| | test_model_v3 | | config1 | | | | config2 | |------------------|------------------|--------------------| | test_model_v4 | | config1 | | | | config2 | | | | config3 | | | | config4 | |------------------|------------------|--------------------|

and

|------------------|------------------|--------------------|-------------------------------| | MODEL | ALIAS FOR | CONFIGURATIONS | OPTIONS | |==================|==================|====================|===============================| | PyChop_fit | PyChop_fit_v2 | | | |------------------|------------------|--------------------|-------------------------------| | PyChop_fit_v1 | | chopper_package | SEQ-100-2.0-AST | | | | | SEQ-700-3.5-AST | | | | | ARCS-100-1.5-AST (default) | | | | | ARCS-700-1.5-AST | | | | | ARCS-700-0.5-AST | | | | | ARCS-100-1.5-SMI | | | | | ARCS-700-1.5-SMI | |------------------|------------------|--------------------|-------------------------------| | PyChop_fit_v2 | | chopper_package | SEQ-100-2.0-AST | | | | | SEQ-700-3.5-AST | | | | | ARCS-100-1.5-AST (default) | | | | | ARCS-700-1.5-AST | | | | | ARCS-700-0.5-AST | | | | | ARCS-100-1.5-SMI | | | | | ARCS-700-1.5-SMI | |------------------|------------------|--------------------|-------------------------------| | test_model | test_model_v4 | | | |------------------|------------------|--------------------|-------------------------------| | test_model_v1 | | config1 | A (default) | | | | | N | | | | | | | | | config2 | B (default) | | | | | C | | | | | D | |------------------|------------------|--------------------|-------------------------------| | test_model_v2 | | config1 | A (default) | | | | | N | | | | | | | | | config2 | B (default) | | | | | C | | | | | D | |------------------|------------------|--------------------|-------------------------------| | test_model_v3 | | config1 | A (default) | | | | | N | | | | | | | | | config2 | B (default) | | | | | C | | | | | D | | | | | E | |------------------|------------------|--------------------|-------------------------------| | test_model_v4 | | config1 | A (default) | | | | | N | | | | | | | | | config2 | B (default) | | | | | C | | | | | D | | | | | | | | | config3 | X (default) | | | | | Y | | | | | | | | | config4 | W (default) | | | | | R | |------------------|------------------|--------------------|-------------------------------|

This implementation:

|------------------|------------------|--------------------| | MODEL | ALIAS FOR | CONFIGURATIONS | |==================|==================|====================| | PyChop_fit | PyChop_fit_v2 | - | |------------------|------------------|--------------------| | PyChop_fit_v1 | - | chopper_package | |------------------|------------------|--------------------| | PyChop_fit_v2 | - | chopper_package | |------------------|------------------|--------------------| | test_model | test_model_v4 | - | |------------------|------------------|--------------------| | test_model_v1 | - | config1 | |------------------|------------------|--------------------| | - | - | config2 | |------------------|------------------|--------------------| | test_model_v2 | - | config1 | |------------------|------------------|--------------------| | - | - | config2 | |------------------|------------------|--------------------| | test_model_v3 | - | config1 | |------------------|------------------|--------------------| | - | - | config2 | |------------------|------------------|--------------------| | test_model_v4 | - | config1 | |------------------|------------------|--------------------| | - | - | config2 | |------------------|------------------|--------------------| | - | - | config3 | |------------------|------------------|--------------------| | - | - | config4 | |------------------|------------------|--------------------|

and

|------------------|------------------|--------------------|-------------------------------| | MODEL | ALIAS FOR | CONFIGURATIONS | OPTIONS | |==================|==================|====================|===============================| | PyChop_fit | PyChop_fit_v2 | - | - | |------------------|------------------|--------------------|-------------------------------| | PyChop_fit_v1 | - | chopper_package | SEQ-100-2.0-AST | |------------------|------------------|--------------------|-------------------------------| | - | - | - | SEQ-700-3.5-AST | |------------------|------------------|--------------------|-------------------------------| | - | - | - | ARCS-100-1.5-AST (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | ARCS-700-1.5-AST | |------------------|------------------|--------------------|-------------------------------| | - | - | - | ARCS-700-0.5-AST | |------------------|------------------|--------------------|-------------------------------| | - | - | - | ARCS-100-1.5-SMI | |------------------|------------------|--------------------|-------------------------------| | - | - | - | ARCS-700-1.5-SMI | |------------------|------------------|--------------------|-------------------------------| | PyChop_fit_v2 | - | chopper_package | SEQ-100-2.0-AST | |------------------|------------------|--------------------|-------------------------------| | - | - | - | SEQ-700-3.5-AST | |------------------|------------------|--------------------|-------------------------------| | - | - | - | ARCS-100-1.5-AST (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | ARCS-700-1.5-AST | |------------------|------------------|--------------------|-------------------------------| | - | - | - | ARCS-700-0.5-AST | |------------------|------------------|--------------------|-------------------------------| | - | - | - | ARCS-100-1.5-SMI | |------------------|------------------|--------------------|-------------------------------| | - | - | - | ARCS-700-1.5-SMI | |------------------|------------------|--------------------|-------------------------------| | test_model | test_model_v4 | - | - | |------------------|------------------|--------------------|-------------------------------| | test_model_v1 | - | config1 | A (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | N | |------------------|------------------|--------------------|-------------------------------| | - | - | - | - | |------------------|------------------|--------------------|-------------------------------| | - | - | config2 | B (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | C | |------------------|------------------|--------------------|-------------------------------| | - | - | - | D | |------------------|------------------|--------------------|-------------------------------| | test_model_v2 | - | config1 | A (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | N | |------------------|------------------|--------------------|-------------------------------| | - | - | - | - | |------------------|------------------|--------------------|-------------------------------| | - | - | config2 | B (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | C | |------------------|------------------|--------------------|-------------------------------| | - | - | - | D | |------------------|------------------|--------------------|-------------------------------| | test_model_v3 | - | config1 | A (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | N | |------------------|------------------|--------------------|-------------------------------| | - | - | - | - | |------------------|------------------|--------------------|-------------------------------| | - | - | config2 | B (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | C | |------------------|------------------|--------------------|-------------------------------| | - | - | - | D | |------------------|------------------|--------------------|-------------------------------| | - | - | - | E | |------------------|------------------|--------------------|-------------------------------| | test_model_v4 | - | config1 | A (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | N | |------------------|------------------|--------------------|-------------------------------| | - | - | - | - | |------------------|------------------|--------------------|-------------------------------| | - | - | config2 | B (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | C | |------------------|------------------|--------------------|-------------------------------| | - | - | - | D | |------------------|------------------|--------------------|-------------------------------| | - | - | - | - | |------------------|------------------|--------------------|-------------------------------| | - | - | config3 | X (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | Y | |------------------|------------------|--------------------|-------------------------------| | - | - | - | - | |------------------|------------------|--------------------|-------------------------------| | - | - | config4 | W (default) | |------------------|------------------|--------------------|-------------------------------| | - | - | - | R | |------------------|------------------|--------------------|-------------------------------|

That's just because I stuck in centred (^) rather than left aligned (<) in the formats

Change data_line to:

data_line = "| " + "| ".join(f"{{:<{length+padding}}}" for length in longest) + "|"

And it should do what you had before.

Ah! I see what you mean, it was grouping things previously.

So, do you want to go ahead with this? With both split_on_model and _format_table in a new utils module?

Are they likely to be re-used? If so then collecting some generic output-printing utils makes sense. This kind of modularity is good for testing.

But if they would have private names and be used in one place, I'm not sure it's so beneficial to split them into a new module.

Not at the moment, no. In the future, difficult to say, but I can't think of anything

If we're not sure, the safest option is private functions in the existing module. As soon as we want to re-use them, they can "go public" in a logical place. But the reverse maneuver is an API-breaking change.

(Ok, we're allowed those at this moment, but I suspect the timescale for "added another feature that prints tables" could overlap with the stable release schedule.)

Could you give me an example of how the split_on_model is supposed to be used inside _format_table? Can't seem to figure out how it's supposed to replace more_itertools.split_before

src/resolution_functions/instrument.py

oerc0122 · 2025-01-22T13:32:13Z

src/resolution_functions/instrument.py

+        for model_name, model_data in self._models.items():
+            length = len(model_name)
+            if length > longest_name:
+                longest_name = length
+
+            if isinstance(model_data, str):
+                contents.append([model_name, model_data, None])
+
+                length = len(model_data)
+                if length > longest_alias:
+                    longest_alias = length
+            else:
+                for i, config_name in enumerate(model_data['configurations']):
+                    if i == 0:
+                        contents.append([model_name, None, config_name])
+                    else:
+                        contents.append([None, None, config_name])
+
+                    length = len(config_name)
+                    if length > longest_config:
+                        longest_config = length


While it requires 2 runs through the elements, I would honestly not compute the length in this building-the-list stage as it makes the loop a lot more complex.

I would iterate through the compiled list and compute the longest in each column after this. The list will be short (<1k elements), so the cost of multiple runs (unless this function is called a lot) is irrelevant.

Pulling the logic out and using:

for name, alias, config in contents: longest_name = max(longest_name, len(name)) longest_alias = max(longest_alias, len(alias)) longest_config = max(longest_config, len(config))

(or something more efficient) at the end makes the purpose clear.

Maybe inside _format_table, we could do (though we would either need to do masking or change the None into e.g. "-"):

contents = np.array(contents) longest = np.max(np.vectorize(len)(b), axis=0)

Seems promising. It should be done after the None->str conversion, just in case we decide to replace None with "----------" 😉

I think it'd be better to do that upstream, in format_available_models_and_configurations and format_available_models_options, and completely eliminite the Nones as well as the pre-processing.

With those sort of details already taken care of, it becomes more practical to have a generic table formatter that e.g. supports multiple criteria of how blocks are split. (Or maybe just requires the blocks to be pre-split as well.)

The previous changes were wrong - they broke the method

ajjackson · 2025-01-24T15:04:59Z

src/resolution_functions/instrument.py

+        try:
+            model = self._models[model_name]
+        except KeyError:
+            raise InvalidModelError(model_name, self)


Would it be appropriate for these to be chained exceptions? https://peps.python.org/pep-3134/

except KeyError as e: raise InvalidModelError(model_name, self) from e

I am not sure the additional information would be useful - the main purpose of this (and similar) exceptions is to communicate to the user that their input was wrong (and how to correct it), knowing that a KeyError was involved seems like unnecessary information. I don't think there would be any harm in the chaining, per se, so I am not necessarily opposed, but I also don't see the reason for - do you have any particular reason for this?

I think it can give a cleaner output than "while handling X, Y was raised", but I'm not super familiar with it. Try raising both and see which output looks better?

WITHOUT from:

Traceback (most recent call last): File "C:\Users\dni83241\Documents\Git\resolution_functions\src\resolution_functions\instrument.py", line 466, in _resolve_model model = self._models[model_name] KeyError: 'PyChop_fit4' During handling of the above exception, another exception occurred: Traceback (most recent call last): File "C:\Users\dni83241\AppData\Local\JetBrains\PyCharm Community Edition 2024.2.1\plugins\python-ce\helpers\pydev\pydevconsole.py", line 364, in runcode coro = func() File "<input>", line 1, in <module> File "C:\Users\dni83241\Documents\Git\resolution_functions\src\resolution_functions\instrument.py", line 468, in _resolve_model raise InvalidModelError(model_name, self) resolution_functions.instrument.InvalidModelError: "PyChop_fit4" is not a valid model for the MAPS instrument version MAPS. This instrument only supports the following models: ['PyChop_fit'].

WITH from:

Traceback (most recent call last): File "C:\Users\dni83241\Documents\Git\resolution_functions\src\resolution_functions\instrument.py", line 466, in _resolve_model model = self._models[model_name] KeyError: 'PyChop_fit4' The above exception was the direct cause of the following exception: Traceback (most recent call last): File "C:\Users\dni83241\AppData\Local\JetBrains\PyCharm Community Edition 2024.2.1\plugins\python-ce\helpers\pydev\pydevconsole.py", line 364, in runcode coro = func() File "<input>", line 1, in <module> File "C:\Users\dni83241\Documents\Git\resolution_functions\src\resolution_functions\instrument.py", line 468, in _resolve_model raise InvalidModelError(model_name, self) from e resolution_functions.instrument.InvalidModelError: "PyChop_fit4" is not a valid model for the MAPS instrument version MAPS. This instrument only supports the following models: ['PyChop_fit'].

Now that you mention it, they are (almost) equally bad - I would prefer the KeyError not to show up at all (wasn't that the default behaviour?) - but the change in text in from e syntax makes it clearer that it is an intended behaviour, so we can go with that at least.

It's more subtle than I expected! Chaining looks slightly more correct, but not a big deal really.

Seems like the output I was hoping for can be obtained via raise InvalidModelError() from None , which gives:

Traceback (most recent call last): File "C:\Users\dni83241\AppData\Local\JetBrains\PyCharm Community Edition 2024.2.1\plugins\python-ce\helpers\pydev\pydevconsole.py", line 364, in runcode coro = func() File "<input>", line 1, in <module> File "C:\Users\dni83241\Documents\Git\resolution_functions\src\resolution_functions\instrument.py", line 468, in _resolve_model raise InvalidModelError(model_name, self) from None resolution_functions.instrument.InvalidModelError: "invalid" is not a valid model for the ARCS instrument version ARCS. This instrument only supports the following models: ['PyChop_fit'].

any thoughts?

ajjackson · 2025-01-24T15:07:37Z

src/resolution_functions/instrument.py

        """
        A list of all :term:`models<model>` available for this :term:`version` of this
        :term:`instrument`.

+        Includes both all the versions of all models and


now updated

ajjackson · 2025-01-24T15:27:07Z

src/resolution_functions/instrument.py

+    longest = np.array(longest) + padding
+    out = []
+    separator = '|' + '|'.join(['-' * (i + 1) for i in longest]) + '|'
+
+    for i, line in enumerate(contents):
+        new_line = '| '
+        for val, length in zip(line, longest):
+            if val is None:
+                new_line += ' ' * length + '| '
+            else:
+                new_line += val + ' ' * (length - len(val)) + '| '
+
+        if line[0] is not None:
+            if i == 1:
+                out.append(separator.replace('-', '='))
+            else:
+                out.append(separator)
+
+        out.append(new_line)
+
+    out.append(separator)
+
+    return '\n'.join(out)


Are they likely to be re-used? If so then collecting some generic output-printing utils makes sense. This kind of modularity is good for testing.

But if they would have private names and be used in one place, I'm not sure it's so beneficial to split them into a new module.

ajjackson · 2025-01-24T15:30:08Z

src/resolution_functions/instrument.py

+        try:
+            model = self._models[model_name]
+        except KeyError:
+            raise InvalidModelError(model_name, self)


I think it can give a cleaner output than "while handling X, Y was raised", but I'm not super familiar with it. Try raising both and see which output looks better?

ajjackson · 2025-01-24T15:32:34Z

src/resolution_functions/instrument.py

+        for model_name, model_data in self._models.items():
+            length = len(model_name)
+            if length > longest_name:
+                longest_name = length
+
+            if isinstance(model_data, str):
+                contents.append([model_name, model_data, None])
+
+                length = len(model_data)
+                if length > longest_alias:
+                    longest_alias = length
+            else:
+                for i, config_name in enumerate(model_data['configurations']):
+                    if i == 0:
+                        contents.append([model_name, None, config_name])
+                    else:
+                        contents.append([None, None, config_name])
+
+                    length = len(config_name)
+                    if length > longest_config:
+                        longest_config = length


Seems promising. It should be done after the None->str conversion, just in case we decide to replace None with "----------" 😉

ajjackson · 2025-01-24T15:40:34Z

src/resolution_functions/instrument_data/arcs.yaml

-        configurations: &configurations
-            chopper_package:
+        configurations:
+            chopper_package: &configurations


What's the reasoning to move this rather than rename it as well?

laziness, lack of attention to detail ...

"the human condition"

The get_model_signature method needs the actual name of the model - which _resolve_model was not returning (only the dict). _resolve_model now returns the new model_name as well. This should have the added benefit of making the logging of the new name easier in the future.

Rastislav Turanyi added 3 commits January 17, 2025 13:52

Update YAML files with new model scheme

ffdb9a0

Initial implementation for Instrument

42fb663

The model properties will not work yet

Add new PyChop version for ARCS and SEQUOIA

24ab899

RastislavTuranyi added bug Something isn't working enhancement New feature or request labels Jan 17, 2025

Use current solution

536950a

RastislavTuranyi commented Jan 17, 2025

View reviewed changes

src/resolution_functions/instrument.py Outdated Show resolved Hide resolved

oerc0122 reviewed Jan 17, 2025

View reviewed changes

src/resolution_functions/instrument.py Outdated Show resolved Hide resolved

Update tests

70c0041

Rastislav Turanyi added 2 commits January 21, 2025 11:47

Add properties for listing subsets of models

6cb53c4

Turn dict properties into methods formatting data into tables

ac4bc43

RastislavTuranyi marked this pull request as ready for review January 22, 2025 11:13

oerc0122 reviewed Jan 22, 2025

View reviewed changes

src/resolution_functions/instrument.py Outdated Show resolved Hide resolved

oerc0122 reviewed Jan 22, 2025

View reviewed changes

src/resolution_functions/instrument.py Outdated Show resolved Hide resolved

oerc0122 reviewed Jan 22, 2025

View reviewed changes

Rastislav Turanyi added 2 commits January 22, 2025 14:29

Change from static to dynamic starting points for longest

11efca7

Remove redundante .keys()

1b3f615

RastislavTuranyi mentioned this pull request Jan 23, 2025

Add How-To Guides #15

Draft

1 task

Revert changes to get_model_signature

00fa259

The previous changes were wrong - they broke the method

ajjackson reviewed Jan 24, 2025

View reviewed changes

Rastislav Turanyi added 2 commits January 24, 2025 15:48

Finished half-finished docstring

dfcbd1a

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add new PyChop version and create model versioning #16

Add new PyChop version and create model versioning #16

RastislavTuranyi commented Jan 17, 2025 •

edited

Loading

ajjackson commented Jan 20, 2025

RastislavTuranyi commented Jan 20, 2025

oerc0122 Jan 22, 2025

RastislavTuranyi Jan 22, 2025

oerc0122 Jan 22, 2025

oerc0122 Jan 22, 2025

oerc0122 Jan 22, 2025

RastislavTuranyi Jan 24, 2025

ajjackson Jan 24, 2025

RastislavTuranyi Jan 24, 2025

ajjackson Jan 24, 2025

RastislavTuranyi Jan 24, 2025

oerc0122 Jan 22, 2025 •

edited

Loading

RastislavTuranyi Jan 22, 2025

ajjackson Jan 24, 2025

RastislavTuranyi Jan 24, 2025

ajjackson Jan 24, 2025

ajjackson Jan 24, 2025 •

edited

Loading

RastislavTuranyi Jan 24, 2025

ajjackson Jan 24, 2025

RastislavTuranyi Jan 24, 2025

ajjackson Jan 24, 2025

RastislavTuranyi Jan 24, 2025

ajjackson Jan 24, 2025

RastislavTuranyi Jan 24, 2025

ajjackson Jan 24, 2025

ajjackson Jan 24, 2025

ajjackson Jan 24, 2025

ajjackson Jan 24, 2025

RastislavTuranyi Jan 24, 2025

ajjackson Jan 24, 2025

Add new PyChop version and create model versioning #16

Are you sure you want to change the base?

Add new PyChop version and create model versioning #16

Conversation

RastislavTuranyi commented Jan 17, 2025 • edited Loading

ajjackson commented Jan 20, 2025

RastislavTuranyi commented Jan 20, 2025

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

oerc0122 Jan 22, 2025 • 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

ajjackson Jan 24, 2025 • 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

RastislavTuranyi commented Jan 17, 2025 •

edited

Loading

oerc0122 Jan 22, 2025 •

edited

Loading

ajjackson Jan 24, 2025 •

edited

Loading