beat issueshttps://gitlab.idiap.ch/groups/beat/-/issues2015-05-13T15:28:25Zhttps://gitlab.idiap.ch/beat/beat.web/-/issues/145Unify toolchain creation page so all happens in a single form2015-05-13T15:28:25ZManuel GüntherUnify toolchain creation page so all happens in a single formWhen I try to create a new version of a toolchain, such as:
https://www.beat-eu.org/platform/toolchains/tutorial/full_lbphs/2
I can only see and edit the description of the toolchain, but the toolchain itself is not shown, and I cannot...When I try to create a new version of a toolchain, such as:
https://www.beat-eu.org/platform/toolchains/tutorial/full_lbphs/2
I can only see and edit the description of the toolchain, but the toolchain itself is not shown, and I cannot edit it.
The same applies, when a new toolchain is created.API Documentationhttps://gitlab.idiap.ch/beat/beat.web/-/issues/146The LaTeX interpreter in .rst files is quite poor2015-05-07T06:34:02ZManuel GüntherThe LaTeX interpreter in .rst files is quite poorFor some reason, the interpreter for LaTeX formulas in the documentation is rather poor.
When I compare the output of ``rst2html`` (generated on the Idiap machines; attached) with the web output:
https://www.beat-eu.org/platform/algori...For some reason, the interpreter for LaTeX formulas in the documentation is rather poor.
When I compare the output of ``rst2html`` (generated on the Idiap machines; attached) with the web output:
https://www.beat-eu.org/platform/algorithms/siebenkopf/GaborGraphSimilarity/12
I can see a clear difference, with a clear preference to ``rst2html``, at least in my browser (iceweasel).
This issue does not need to be solved immediately, but in the long term, formulas like the one in the link above don't really look nice.
[GaborGraphSimilarity.html](https://gitlab.idiap.ch/biometric/beat.web/uploads/b8d8d8568afd8f11af1b01acd1d925e4/GaborGraphSimilarity.html)
https://gitlab.idiap.ch/beat/beat.web/-/issues/147When documenting data formats, the location of the .rst file is shown2015-05-06T15:14:33ZManuel GüntherWhen documenting data formats, the location of the .rst file is shownI recently added documentation for a data format:
https://www.beat-eu.org/platform/dataformats/siebenkopf/jet/1
After saving the long description, the location of the .rst file is shown, instead of its (parsed) content.
Note: When...I recently added documentation for a data format:
https://www.beat-eu.org/platform/dataformats/siebenkopf/jet/1
After saving the long description, the location of the .rst file is shown, instead of its (parsed) content.
Note: When I click on ``edit``, the correct .rst file is opened again.Tutorial for the FG2015 ConferenceLaurent EL SHAFEYLaurent EL SHAFEYhttps://gitlab.idiap.ch/beat/beat.web/-/issues/148Cannot return plot output in an analyzer2015-05-06T15:07:33ZManuel GüntherCannot return plot output in an analyzerWhen creating a new version of an analyzer as:
https://www.beat-eu.org/platform/algorithms/siebenkopf/ROC/14/
I cannot select the output used for plotting, but only ``float32``, ``int32``, ``bool`` and ``string`` can be selected.When creating a new version of an analyzer as:
https://www.beat-eu.org/platform/algorithms/siebenkopf/ROC/14/
I cannot select the output used for plotting, but only ``float32``, ``int32``, ``bool`` and ``string`` can be selected.Tutorial for the FG2015 ConferencePhilip ABBETPhilip ABBEThttps://gitlab.idiap.ch/beat/beat.web/-/issues/150Doc strings on search save are required, but not displayed2015-05-07T07:40:08ZAndré AnjosDoc strings on search save are required, but not displayedWe have an issue with documentation strings on searches. Here is what happens:
1. The user must provide a short description **and** a long description while saving
2. At the user micro-site, these are not displayed
This is what sh...We have an issue with documentation strings on searches. Here is what happens:
1. The user must provide a short description **and** a long description while saving
2. At the user micro-site, these are not displayed
This is what should happen:
1. The user *may* provide a short description and/or a long description.
2. The short description should be displayed at the user micro-site like for other objects.Tutorial for the FG2015 ConferenceFlavio TARSETTIFlavio TARSETTIhttps://gitlab.idiap.ch/beat/beat.web/-/issues/151Can request attestation for experiment that I don't own2015-05-07T07:22:55ZAndré AnjosCan request attestation for experiment that I don't ownThe button to request an attestation still shows up even if the experiment is not yours.
Reproduce:
Following this link (while you're not user "siebenkopf"): https://www.beat-eu.org/platform/experiments/siebenkopf/siebenkopf/FaceRe...The button to request an attestation still shows up even if the experiment is not yours.
Reproduce:
Following this link (while you're not user "siebenkopf"): https://www.beat-eu.org/platform/experiments/siebenkopf/siebenkopf/FaceRec-WithOut-Training/2/Banca_P-PhaseDiff/Tutorial for the FG2015 ConferenceSamuel GAISTSamuel GAISThttps://gitlab.idiap.ch/beat/beat.web/-/issues/152(Complex) result types when forking an analyzer are wrong2015-05-07T08:20:09ZLaurent EL SHAFEY(Complex) result types when forking an analyzer are wrongWhen using the algorithm/analyzer editor, the result types which are complex (e.g. plot/scatter/1) are incorrect.
To reproduce the problem:
1. Fork the analyzer [tutorial/postperf/1/](https://www.beat-eu.org/platform/algorithms/fork/...When using the algorithm/analyzer editor, the result types which are complex (e.g. plot/scatter/1) are incorrect.
To reproduce the problem:
1. Fork the analyzer [tutorial/postperf/1/](https://www.beat-eu.org/platform/algorithms/fork/tutorial/postperf/1/) on the production platform
2. Looking at the results, the type for the `roc` field is set to `float32`, while it is supposed to be `plot/scatter/1`.
In addition, the menulist for the types shows several entries `object Object` in addition to the basic ones (such as `float32`), but no entries such as `plot/scatter/1`https://gitlab.idiap.ch/beat/beat.web/-/issues/153Long delay after clicking the 'Go' button2015-05-29T14:43:59ZManuel GüntherLong delay after clicking the 'Go' buttonWhen starting an experiment, sometimes it takes quite some time until something happens (e.g., the execution of the experiment is displayed).
Particularly, when the experiment status is "Scheduled", the experiment submission webpage doe...When starting an experiment, sometimes it takes quite some time until something happens (e.g., the execution of the experiment is displayed).
Particularly, when the experiment status is "Scheduled", the experiment submission webpage doesn't change, until the experiment actually started to execute.
My problem is that there is absolutely no response, whether the submission was successful or not.
In my point of view, there are three options to display the successive submission:
1. Show the experiment execution page immediately after successful submission
2. Change the "Go" button into something else, for example a "Stop" button
3. Open a message box, saying: "Experiment was scheduled, waiting for it to start" with two buttons "Cancel experiment" and "Show experiment", which closes automatically after the experiment was startedAPI DocumentationLaurent EL SHAFEYLaurent EL SHAFEYhttps://gitlab.idiap.ch/beat/beat.web/-/issues/154Edit button in non-editable algorithm2015-05-11T14:53:00ZManuel GüntherEdit button in non-editable algorithmI have an algorithm that belongs to a failed experiment:
https://www.beat-eu.org/platform/algorithms/siebenkopf/ROC/
The experiment was not deleted yet, and I should not be able to edit the algorithm.
However, there is an ``Edit`` b...I have an algorithm that belongs to a failed experiment:
https://www.beat-eu.org/platform/algorithms/siebenkopf/ROC/
The experiment was not deleted yet, and I should not be able to edit the algorithm.
However, there is an ``Edit`` button, which I can click and which leads me to the algorithm editing page.
Nevertheless, when I try to ``Save`` the algorithm, I get the error message:
```
Algorithm not created, please fix the errors below
{"detail":"The algorithm isn't modifiable anymore (either shared with someone else, or needed by an attestation)"}
```
So, it seems that the algorithm knows that it is not modifiable, but still the ``Edit`` button exists.API DocumentationPhilip ABBETPhilip ABBEThttps://gitlab.idiap.ch/beat/beat.web/-/issues/156Saved search is not displayed2015-05-07T10:08:10ZAndré AnjosSaved search is not displayedIf I click (from the user micro-site, Databases panel), on this query for the AT&T database:
https://www.beat-eu.org/platform/search/?query=db:atnt%20type:results
And I save this search, with its default name, and then I go back to...If I click (from the user micro-site, Databases panel), on this query for the AT&T database:
https://www.beat-eu.org/platform/search/?query=db:atnt%20type:results
And I save this search, with its default name, and then I go back to the micro-site, "Searches" panel and click on the saved search, I cannot see it properly.Tutorial for the FG2015 ConferenceFlavio TARSETTIFlavio TARSETTIhttps://gitlab.idiap.ch/beat/beat.web/-/issues/157Team creation and editor could see a face lift2015-05-22T15:16:05ZAndré AnjosTeam creation and editor could see a face liftApparently we're using Django's form widgets to get that going, but it certainly does not look as nice as the other editors we have.
It would be good to see it face-lifted after the tutorial.Apparently we're using Django's form widgets to get that going, but it certainly does not look as nice as the other editors we have.
It would be good to see it face-lifted after the tutorial.API Documentationhttps://gitlab.idiap.ch/beat/beat.web/-/issues/158The sentence "Public for specific users..." is not sensical2015-05-07T08:51:20ZAndré AnjosThe sentence "Public for specific users..." is not sensicalMaybe we could modify it to "Shared with specific users..." instead.Maybe we could modify it to "Shared with specific users..." instead.Tutorial for the FG2015 ConferenceSamuel GAISTSamuel GAISThttps://gitlab.idiap.ch/beat/beat.web/-/issues/159Type casting between int64 and uint64 not possible2015-05-09T08:06:40ZManuel GüntherType casting between int64 and uint64 not possibleI recently stumbled upon an error message, which I don't quite agree with. I wanted to write the ``number_of_positives`` to the plot/isoroc format. Since I am using python, I simply used the ``len(...)`` function. However, I got the erro...I recently stumbled upon an error message, which I don't quite agree with. I wanted to write the ``number_of_positives`` to the plot/isoroc format. Since I am using python, I simply used the ``len(...)`` function. However, I got the error message:
```
TypeError: cannot safely cast attribute `number_of_negatives' on dataformat `plot/isoroc/1.data_type' with type `int64' to `uint64' without precision loss
```
which looks weired to me. In C++, the cast from ``int`` to ``unsigned`` **always** succeeds, so I don't understand, why this is not the case in BEAT.
*(In fact, is there any automatic cast that would succeed in BEAT?)*
So, what do you think about that? Should we allow the automatic conversion from ``signed`` to ``unsigned``?API Documentationhttps://gitlab.idiap.ch/beat/beat.web/-/issues/160Attestation of a failed experiment?2015-05-08T08:03:27ZManuel GüntherAttestation of a failed experiment?I have a failed experiment, which was caused by a programming issue. However, the platform allows me to ask for an attestation for that experiment (at least I can see the ``Request an attestation`` button.
Does that make any sense, or c...I have a failed experiment, which was caused by a programming issue. However, the platform allows me to ask for an attestation for that experiment (at least I can see the ``Request an attestation`` button.
Does that make any sense, or can we disable or remove the button in case of failed experiments?https://gitlab.idiap.ch/beat/beat.web/-/issues/161Creating an experiment from the toolchain page2015-05-11T15:02:35ZLaurent EL SHAFEYCreating an experiment from the toolchain pageCurrently, we cannot create an experiment from the toolchain page.
In contrast, to create an experiment, the first thing to do is to select a toolchain.
Since the workflow is to first create a toolchain, and next to configure an expe...Currently, we cannot create an experiment from the toolchain page.
In contrast, to create an experiment, the first thing to do is to select a toolchain.
Since the workflow is to first create a toolchain, and next to configure an experiment, it would be nice to add a button such as 'create an experiment' from the toolchain page. This would prevent to have to click several times before reaching the desired page.API Documentationhttps://gitlab.idiap.ch/beat/beat.web/-/issues/163Dataformat validation is still unprotected2015-05-19T12:33:19ZAndré AnjosDataformat validation is still unprotectedFollowing an analysis from @laurent.el-shafey, the data format errors you saw passing by are related to a lack of protection (missing try/except clauses) during their validation.
Here is one of the backtraces:
```text
File "/remot...Following an analysis from @laurent.el-shafey, the data format errors you saw passing by are related to a lack of protection (missing try/except clauses) during their validation.
Here is one of the backtraces:
```text
File "/remote/sw/beat.web/src/beat.web/beat/web/common/serializers.py", line 203, in get_declaration
return obj.json()
File "/remote/sw/beat.web/src/beat.web/beat/web/common/models.py", line 298, in json
return simplejson.loads(self.load_contents('file'),
File "/remote/sw/beat.web/src/beat.web/beat/web/common/models.py", line 272, in load_contents
f.open()
File "/remote/sw/beat.env.deploy/usr/lib/python2.7/site-packages/django/db/models/fields/files.py", line 79, in open
self.file.open(mode)
File "/remote/sw/beat.env.deploy/usr/lib/python2.7/site-packages/django/db/models/fields/files.py", line 49, in _get_file
self._file = self.storage.open(self.name, 'rb')
File "/remote/sw/beat.env.deploy/usr/lib/python2.7/site-packages/django/core/files/storage.py", line 35, in open
return self._open(name, mode)
File "/remote/sw/beat.env.deploy/usr/lib/python2.7/site-packages/django/core/files/storage.py", line 172, in _open
return File(open(self.path(name), mode))
IOError: [Errno 2] No such file or directory: u'/remote/prefix/dataformats/40/07/robert/FaceBoundingBox/1.json'
```API Documentationhttps://gitlab.idiap.ch/beat/beat.web/-/issues/164Package documentation2015-05-29T14:43:01ZAndré AnjosPackage documentationPackage documentation needs to be created for all of the public API.Package documentation needs to be created for all of the public API.API DocumentationLaurent EL SHAFEYLaurent EL SHAFEYhttps://gitlab.idiap.ch/beat/beat.web/-/issues/165`dataset-name` in search causes 500 internal error2015-05-12T16:11:21ZLaurent EL SHAFEY`dataset-name` in search causes 500 internal errorWhen selecting the field `dataset-name` in a search, pressing the `apply` button will cause a 500 internal error.When selecting the field `dataset-name` in a search, pressing the `apply` button will cause a 500 internal error.BEAT TourLaurent EL SHAFEYLaurent EL SHAFEYhttps://gitlab.idiap.ch/beat/beat.web/-/issues/166Dataformat validation on beat.web is incomplete2015-05-12T16:18:35ZLaurent EL SHAFEYDataformat validation on beat.web is incompleteWhen creating the following dataformat on beat.web,
```
{
"bla": "uint32",
"blo": "uint24"
}
```
a 500 internal error is generated instead of a proper error message.When creating the following dataformat on beat.web,
```
{
"bla": "uint32",
"blo": "uint24"
}
```
a 500 internal error is generated instead of a proper error message.BEAT Tourhttps://gitlab.idiap.ch/beat/beat.web/-/issues/167beat.web.toolchain.utils imports non-existing function 'experiment_to_json'2015-05-12T15:23:27ZLaurent EL SHAFEYbeat.web.toolchain.utils imports non-existing function 'experiment_to_json'beat.web.toolchain.utils currently tries to import the non-existing function 'experiment_to_json' from beat.web.experiments.utils.beat.web.toolchain.utils currently tries to import the non-existing function 'experiment_to_json' from beat.web.experiments.utils.API Documentation