Contribute to Open Source. Search issue labels to find the right project for you!

Update all in Project tool


From @vmeau:

it would be really useful to have a button that says “Update all” that would run since the very beginning. Maybe it’s already doing it behind the scene, but as user it’d be great to know that it’s happening from the very top and not from the latest query. :)

Updated 27/06/2017 01:37

Implement QueryBuilder


The purpose of QueryBuilder class is to convert URI/URL query key-values to Camo-acceptable options. Here’s an example:

  1. send GET request to the endpoint: /companies?match=name>goo&show=name,location,logo
  2. the req value grabbed by express will convert the query to: { match: 'name>goo', show: 'name,location,logo', }
  3. With QueryBuilder’s conversion, we should be able to do following Camo statement: const result = await Company.find({ name: 'goo' }, { populate: ['name', 'location', 'logo'] });

Tasks: - [ ] Create a separate QueryBuilder class with its constructor taking an object as a parameter - [ ] Create a “private” translator function that will convert URL query properties to Camo-acceptable properties - [ ] Create a “private” parser function that will parse URL query values to Camo-acceptable values - [ ] Create a getter instance function for query (e.g. { name: 'goo' }) - [ ] Create a getter instance function for options (e.g. { populate: ['name', 'location', 'logo'] }) - [ ] Create a unit test for QueryBuilder (ensure it covers at least 80% of QueryBulider class) - [ ] Manually test e2e flow to make sure QueryBuilder is working correctly

Query Specs: TBD

Updated 27/06/2017 01:40

Async Routing


I know that the bundle size is a much better situation with hyperapp but I am still interested in the possibility of having an async routing solution, after reading #176 I thought I would open an issue to just hash out what I think would be an interesting implementation and whether it should be it’s own fork of the current router or if I could piece together a PR to add it as a feature to the current router.

This may not go over well with many here since it is webpack specific, as far as I understand currently. But the gist of it is kind of covered in react-loadable and webpack async code splitting .

My first stab at implementing this would be to fork the current router (or whatever repo it ends up in), and make the minimum necessary changes for it to accept a promise as a view, instead of expecting only a function. Might also need a small helper function to do the bit react-loadable takes care of where it displays a loading indicator while the async route is downloaded and then rendered.

Do any of you have thoughts about async routing or concerns about the 20,000 foot outline above? I think if I did it right it could be compatible with require.ensure() as well which would make it more webpack agnostic.

Updated 27/06/2017 01:10

Generate Kinfu texture model with attention (BoundingBox) and Ground frame to fix occluded surface



I’d like to create mesh model of objects used in picking task. An application is 6d annotation to create training data.


Robot looks around objects on table, user annotates bounding box which is static on table in usual case. In kinfu nodelet, it takes service request with bounding box and ground frame id to generate occluded surface.

Updated 27/06/2017 01:06

Estilo en recursos de administración


En los siguientes componentes, el estilo cambiado (cambios al tema soria de dojo) afecta en la administración, y faltan que se carguen iconos

  • [ ] Editor de Plantillas <img width=“943” alt=“captura de pantalla 2017-06-22 a la s 12 37 44” src=“”>

  • [ ] Submenu: icono de eliminar y recargar 2017-06-23_115905

  • [ ] Editor de plantillas 2017-06-23_120447

  • [ ] Listado de Agregar recursos,en el editor de plantillas, icono de carpeta y el hover 2017-06-23_120703

  • [ ] Ventana de “agregar secuencia” y “agregar actividad"del flujo de publicación y esto provoca que no se muestre el list box de las actividades creadas 2017-06-26_170324

  • [ ] Recurso Banner, el hover de la pestaña “Estilo” y “Configuración Avanzada” 2017-06-26_172518

  • [ ] Recurso Poll el hover de la pestaña “Configuración de los resultados”, “Configuración de los resultados” y “Configuración Avanzada” 2017-06-26_173727

  • [ ] Recurso Poll el hover de la pestaña “Configuración del estilo” y “Configuración Avanzada” 2017-06-26_173951

  • [ ] Crear sitio Personalizado y Predeterminado y Exportar/importar Modelo: e la pestaña de “Sitios a Guardar” 2017-06-26_182059

  • [ ] Crear repositorio Predeterminado: e la pestaña de “Repositorios de usuarios” 2017-06-26_182444

Updated 26/06/2017 23:40

Error methods


There should be two kinds of error messages:

  • Internal errors (e.g. when an implementation error occurred)
  • External errors (e.g. an image has not the wanted format)

They can me named as follows: * __error(...), __internalError(...), … * __runtimeError(...), __environmentError(...), …

Updated 26/06/2017 22:48

CSV export Bug


noticed that the history graph got “device selector” abilities – although one minor bug, it doesn’t seem like CSV export does anything while the device selector is selected. Not sure if that was intentional but it would be great to export the pin from all devices into one CSV.

Updated 26/06/2017 22:39

Scheduled posting


Implement a feature that would allow to create wall posts according to a certain interval, such as once an hour (random time in hour) or once a day (random time during the day) or morning, evening, etc.

Updated 26/06/2017 22:09

Remove Units and periodic tables


There are only a few cases where units and periodic tables are needed (IO/glue-code and gbasis/grid respectively), so it is not worth breaking modularity for it.

We need to replace those situations with simple dictionaries of data (atomic-nubmer:covalent-radii for grid, symbol:atomic-number for gbasis). There is a little bit of duplication of data, but it’s a small amount that will never get changed. The convenience of having a periodic table class should be reserved for glue-code level or maybe even user code.

Updated 26/06/2017 21:37

Print messages directly, no logger


In order to remove the Log object, we need to print messages directly to stdout.

This is straightforward except that we will then need to have a way of sorting messages by verbosity. The approach we are taking is to prepend messages with a known string (ie 5:) that contains information about verbosity. Higher numbers mean more verbose messages. Users could then grep messages directly (ie. “I want all messages with prefixes less than 5). Or we could later redirect stdout to a streamIO and parse the numbers ourselves. In either case, this would take place at the level of glue-code or higher.

Updated 27/06/2017 00:51 2 Comments

Bibliography Data in new formats


The bibliography data should be changed so that objects being created assign a biblio list attribute which stores the data for citing the method. As functions are called, they should modify the biblio attribute and append new citation data to it in the form of a string (much like how the current logger’s biblio is implemented).

We also need an inspection object (for use at the glue-code level) to recurse over all the objects in locals() or globals() and collate all the biblio attributes into one and then print out citation data.

Updated 27/06/2017 00:49 1 Comments

[Test] run_tests should support --test


I should be able to run:

run_tests --module {module] --test {one or more test names}

in order to selectively run tests in the command line. I can selectively run them in VS, but this is very slow and I avoid it unless I need to debug into the test.

Currently there is a syntax like so: nosetests azure.cli.command_modules.resource.tests.test_resource.DeploymentTest -c nose.cfg

This only seems to work on Python 3 (See #3883) but more to the point, I should be able to leverage the script and not worry about the implementation or where the test folder is.

Updated 26/06/2017 23:02

[WIP] Weapons


Add weapons

  • [X] Weapon Registry
  • [X] Make some weapons
  • [ ] Badge for buy weapon
  • [ ] Window in game for select weapon
  • [ ] Use weapon
  • [ ] CSS helper for weapon
  • [ ] AI use weapon
  • [ ] Update TODO
Updated 26/06/2017 21:40

Meanfield module refactoring


The meanfield module should be slightly refactored to make picking and choosing easier.

The SCF and abstract Hamiltonian should not have any references to HORTON datatypes. The HF and DFT hamiltonians should be separated into their own files, and can make reference to HORTON datatypes such as gbasis or grid. That way, an external user can choose to take the meanfield module alone and delete the HF and DFT modules if they want to break the dependencies on HORTON.

Updated 26/06/2017 21:13

Feature Request: Dont allow players to join while server is shutting down


It would be kind of like the late-bind setting (where late-bind wont let player connect until after the server is fully loaded).

I have quite a number of plugins on my server, and some of them take a while to shutdown (sometimes a shutdown takes 30 seconds). During the shutdown a number of players reconnect, which not only seems to slow the shutdown process but it also gets them in on the server while some plugins are already disabled such as land protections (which is very bad).

I have tried numerous ways to kick players during a shutdown, but plugin event listeners stop listening during a shutdown process, which makes it impossible to kick on join from a plugin during a shutdown. :(

I propose a paper.yml setting to be added to stop players from joining the server once a shutdown has started.

Updated 26/06/2017 23:30 2 Comments

Restrict Expansion class to Meanfield module


The Expansion object can become a mini-matrix class as well, so it needs to be restricted to situations that truly need it. This is common in the Meanfield module.

However, outside of the meanfield module, we can simply refactor the code to store a numpy array of coefficients, and arrays for energies and occupations.

Updated 26/06/2017 21:09

Strip out Matrix class


One of the barriers to HORTON 3.0 modularity is the presence of classes that permeate the entire project. One of these is the Matrix class.

We need to remove the matrix class and replace its functionality with either helper classes, or with named tuples. The appropriate choices will be case dependent. Each module will need to have the functionality removed and tested.

Updated 26/06/2017 21:07

Adds heartbeats


:cl: Tacolizard add: NanoTrasen has now outfitted their employees with Extra-Loud™ Genetically Modified Hearts! Now you can hear your heart about to explode when you take meth, or hear it slowly coming to a stop as you bleed out in critical condition! /:cl:

ok yeah the code is kinda janky, but I’m waiting for oranges to hopefully do a sound manager rewrite.

both of the new sounds are licensed under Creative Commons 0, so don’t worry about it.

Updated 27/06/2017 00:54 24 Comments

Errors when converting


Raise errors, when something is not correct during parsing. Some situations where this might be useful:

Throwing error and exit program: * duplicates in class/field names * missing properties/name object * incomplete property object (e.g. collection is missing)

Printing warning but still go on: * unlikely property (e.g. empty name) * additional and therefore unused stuff

Also put the line into the error-message

Updated 26/06/2017 20:17

`hub fork` fails if fork (and remote) already exists


While attempting to replicate the issue I thought I was seeing for, I uncovered a teeny quirk with how hub fork behaves for existing forks.

  • If fork does not exist: hub fork creates the fork on github, adds the remote for the new fork, and fetches. Returns 0 status.

  • If fork exists but remote doesn’t: hub fork adds the remote for the fork and fetches. Returns 0 status.

  • If fork exists and remote exists: hub fork errors out with fatal: remote jasonkarns already exists

Given that forking an already-forked repo safely adds the fork as a remote, I’d kinda expect the last scenario to just trigger a fetch. That is, I’d expect all three scenarios above to result in a local clone with:

  1. named remote for the fork
  2. remote recently fetched
  3. successful return code

Right now, only the first two scenarios do this. (The third just needs to do a fetch and exit cleanly.)

I have a hard time describing the second scenario above (fork exists, but remote doesn’t), but it’s like a partially-idempotent command. If the third scenario were implemented as requested, the hub fork command would be fully idempotent.

Updated 26/06/2017 21:08 1 Comments

Toml Jinja filter


Description of Issue/Question

Lots of software, the TICK stack in particular, uses Toml for configuration files. Writing pillar data in Toml is quite difficult in Salt and as far as I’m aware requires writing your own Toml encoder in Jinja using macros. It would be great if there was a Toml filter, like we have JSON and Yaml filters, so we could do:

{{ config | toml }}

to write a Toml configuration file.

Updated 26/06/2017 19:42 1 Comments

Auto-reconnect colyseus.js to server


In my development environment, any time I save/restart the server (I’m using nodemon), it reboots. The client will disconnect and then I have to refresh the page to reconnect. In a live game, this means that if I push an update to the server every player will disconnect and not attempt to reconnect automatically.

Automatic reconnection (with exponential backoff configurable, as primus has) would help this a lot.

Updated 26/06/2017 19:10 1 Comments

Develop full test automation


Should be valid and consistent in test scenarios including:

Circular vs. elliptical vs. parabolic vs. hyperbolic orbits need to return accurate data Transferring to a new parent body frame needs to return consistent position vectors Orbital element calculation needs to be accurate when tested against several different sets of orbital data Check accuracy in high vs. mid vs. low inclination Check accuracy in high vs. mid vs. low eccentricity Longitude of ascending node must be consistent over time Predicted vs. measured true anomaly should be the same for a given amount of time after a known point Angular orbital elements shouldn’t go past 360 degrees

Updated 26/06/2017 17:11

Optimized launch profiles


Using air density profiles for Kerbin, there should be a way to find the most fuel-efficient trajectory for launch. This could definitely be done numerically, or by the calculus of variations. Other options are possible.

Updated 26/06/2017 16:50

Overriding basePath


Could add support for overriding of basePath in specification when loading from SwaggerUIBundle please?

This could fix the issue with multiple basePaths requested in here

At least for our case since we are using hapi views and then we can modify html context to render. So then you can use this context to modify basePath property when instantiating SwaggerUIBundle.

Updated 26/06/2017 20:58 1 Comments

Add aggregation operators added in MongoDB 3.4


This adds almost all new aggregation operators introduced in MongoDB 3.4 to the aggregation builder. Still missing: * [ ] $collStats: I’m not sure what a smart API would be, but most likely I’d make the signature something like collStats(bool $latencyStats, bool $latencyHistograms, bool $histograms) and construct the expression from the three bools given. * [ ] $switch: this would require using different names for case, then and default as those are reserved words. Only when using PHP 7 can we use these words as method names. One option would be to add other methods and mark them as deprecated from the outset.

Updated 26/06/2017 18:20 2 Comments

Fork me on GitHub