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

Need a cleaner way to specify parameter file paths

UWB-Biocomputing/BrainGrid

Right now, we specify the main parameter file name on the command line. The neuron lists have relative path names inside that file. Those paths are relative to the working directory where BG is run, not where the parameter file is. We should come up with a way to make this more obvious, or we should do something like add a command line argument that specifies the parameter file directory and have the NList file paths be relative to that (and eliminate the need to provide the path in the command line argument for the main parameter file).

<!— @huboard:{“order”:12.983111824087368,“milestone_order”:12.962356491611725} –>

Updated 13/10/2017 15:39

Update tutorial to templates and styles paths compatible with webpack

NativeScript/docs

From @felipeferri on October 12, 2017 12:44

I was using the excellent Groceries tutorial as a reference for my first Nativescript app, using Angular and Typescript.

My app worked ok on the simulators and devices, but when I ran webpack I started to get lots of error messages of templates being looked for in weird places, such as:

../pages/splash/pages/splash/splash.html

This was happening because I was using, as the tutorial suggests, the following format for setting the templates paths on the @Component decorators:

pages/splash/splash.html

This works then running the app, but it seems that webpack resolves this relative path over the current working directory. So I found out that the correct way of setting the templates and styles paths is using:

@Component({
    selector: "splashview",
    moduleId: module.id,
    templateUrl: "./splash.html",
    styleUrls: ["./splash-common.css"]
})

And if you check the Groceries code on github the paths are set like this.

Some answers I found on Stackoverflow (see here for example) suggest it is not necessary anymore to use the moduleId line, however I tested it and noticed it breaks the app if you omit it.

I propose the tutorials are corrected to use this format for the path, or at least the documentation on webpack is updated to include this.

Copied from original issue: NativeScript/nativescript-angular#1038

Updated 13/10/2017 15:28 1 Comments

Hierarchical packs

stohrendorf/confckurator

Variable values are specified in each pack. As several packs can often share the same variable values, it makes sense to implement some sort of “meta pack system”, so that you can define a pack with some default values for some variables, and other packs may specify that pack as a base and complete the variable values and (optionally) override base values.

Updated 12/10/2017 16:39 1 Comments

SARL Acronym meaning

sarl/sarl

I have more and more people asking me that is the meaning of “SARL”.

I know we have not yet decided of this meaning. I would like to propose a contest on Facebook and Google Group for giving the opportunity to the community to propose a meaning.

To be accepted, the proposed meaning should be put as a comment of this issue, in order to keep track of the initial proposer.

The core developers of SARL (@gallandarakhneorg, @ngaud , @srodriguez) will decide of the best proposal.

You could also vote for a proposal by adding a :+1: or :-1: or :heart: (multiple votes are accepted)

Deadline: Release of Version 1.0.0.

<!— @huboard:{“order”:2.5896349805435154e-54,“milestone_order”:6.807904890217455e-59} –>

Updated 16/10/2017 13:40 16 Comments

More useful execution log output

daisy/pipeline-mod-braille

Moved from https://github.com/snaekobbi/issues/issues/29:

This is some information that might be useful for braille experts to have: - [ ] What happens during volume breaking - [ ] Consequences due to hyphenation and long words either not fitting on a line, causing orphan/widows, etc. - [ ] Characters that cannot be translated to braille due to them not being present in the braille table. - [ ] What happens to tables, how are they laid out, when do they go from being rendered in one way as opposed to another way - [ ] The location in the input file for all these messages - [ ] In addition to being output in the execution log, summarize relevant issues and warnings in a job report, including the number of times each type occured - [ ] How far into the book the conversion is (page number or anything else that can be counted, maybe say for each 100th paragraph something like “Processed {N*100} paragraphs…”) - [ ] What part of the conversion is currently being performed (translation, layout, volume breaking, etc.) - [ ] Remove or hide warnings and errors that are expected, does not add any useful information, and is confusing for end-users

Updated 09/10/2017 10:34

Link the cast operator to X and the "toX" function.

sarl/sarl

In SARL cast is done with the following operator: sarl value as Type Consider: value as String. It may be interesting to generate the following Java code: java value.toString()

This principle may be extended as follows: * Consider the expression v as T. * If the type of v is instance of T, generates: (T) v. * If T is a primitive type and the type of v declares a method with the prototype: def toPrimitiveT() : T. * If the type of v declares a method with the prototype: def <ST extends T> ST toST() (multiple choices will generates an error). * Otherwise the standard “Cannot cast x to T”.

<!— @huboard:{“order”:2.5898939440415696e-54,“milestone_order”:6.808585680706476e-59} –>

Updated 08/10/2017 07:51

Feature/25

BobocodeCourse/gojohnny-project

25 added UserController in web/controller with validation DTO, added postgres driver, added theamleaf for transfering DTO throught context.

<!— @huboard:{“order”:85.50425,“milestone_order”:76,“custom_state”:“archived”} –>

Updated 16/10/2017 17:37

child deletion security.

Picolab/wrangler

when a child pico is deleted there is a long chain of events that end with a scheduled event. this scheduled event is what removes the actual pico. It is possible for a malicious event to by pass the event chain and remove a single pico without thought about the child’s children picos. This allows a rogue grand child to exist that is un-reachable by wrangler in traditional ways. the classic engine used special event domains that only “root” process could raise to prevent this. An alternate solution could be creating and storing a uuid in an entity variable, sending the uuid as an attribute in the scheduled event. Then in the rule which is removing the pico, checking that the provided uuid is a valid stored uuid. This would increase the difficulty of creating rogue picos.

<!— @huboard:{“order”:0.15621875468687507,“milestone_order”:1.8585520386272383e-09} –>

Updated 07/10/2017 16:18 1 Comments

wrangler function return type?

Picolab/wrangler

wrangler functions return a map with the expected return item as the key. for example {channels:[<channel1>,…]}. The main reason for this was to allow an error/status object to be returned with in the map. The error object did not make sense any where in wrangler and has been deprecated. Now the map keyed to the results of the function causes a lot of over head in krl which could be avoided if we simply return the results. Example wrangler:channel("ted"){"channel"}{"id"} vs wrangler:channel("ted"){"id"} . I think we should abandon the return map and return just the results with the promise each function will only return a certain type. This is an api changing suggestion and would break any current code using wrangler.

<!— @huboard:{“order”:0.15623437656234376,“milestone_order”:1.8592955709632457e-09} –>

Updated 07/10/2017 03:16

Flights Grid not Populating Hit Counts

US-CBP/GTAS

Hit counts are represented for the passenger grid.

Flight Grid: uploads/f6de8fbf-8a10-46c4-9bef-89c880beaf82/Screen Shot 2017-10-06 at 12.24.00 PM.png

Passenger Grid: uploads/59154245-e015-432b-848f-0bc89800b800/Screen Shot 2017-10-06 at 12.25.14 PM.png

No errors are thrown, basically the frontend gets a 0 for rulehitcount and watchlistcount. I think when it gets the flight by criteria it pulls the flight model from the Flights Table where the rulehitcount and watchlistcount are 0. uploads/d3910c78-80e5-4e4e-8d15-5f1c2559d962/Screen Shot 2017-10-06 at 12.28.23 PM.png

However, the actual values are in the hit summary table: uploads/2d0b7326-c045-426e-9bfc-4ab094aa6c89/Screen Shot 2017-10-06 at 12.30.12 PM.png

<!— @huboard:{“order”:4.439949763243843e-09,“milestone_order”:7.113871575813207e-08} –>

Updated 06/10/2017 19:22 1 Comments

KNP menu enabled true not working

avanzu/AdminThemeBundle

I’m having a problem with knp_menu. My config is avanzu_admin_theme: knp_menu: # knp menu integration enable : true

but when I put {{ dump(avanzu_admin_context.knp_menu) }} on default-layout.html.twig it dumps

array:3 [▼
  "enable" => false
  "main_menu" => "AvanzuAdminThemeBundle:MenuBuilder:createMainMenu"
  "breadcrumb_menu" => false
]

So, knp is never activated.

Also, (maybe this should go on another thread) if I change

enable

default option at ContexHelper.php to true I get

Unknown “knp_menu_render” function at

knp-menu.html.twig at line 2 (knp_menu_render(avanzu_admin_context.knp_menu.main_menu, {)

Maybe I need to install knp bundle?

Updated 05/10/2017 12:23 1 Comments

[ERROR] general#search_redirect (ActionController::BadRequest) "Invalid query parameters: inval...

mysociety/alaveteli

This is an odd one, looks like the request was: REQUEST_URI : /search?query=%s

An ActionController::BadRequest occurred in general#search_redirect:

  Invalid query parameters: invalid %-encoding (%s)
  lib/strip_empty_sessions.rb:13:in `call'


-------------------------------
Request:
-------------------------------

  ERROR: Failed to generate exception summary:

  ActionView::Template::Error: Invalid query parameters: invalid %-encoding (%s)

  [FILTERED]/www.whatdotheyknow.com/alaveteli-2017-09-27T11-56-09/vendor/bundle/ruby/2.3.0/gems/rack-1.6.8/lib/rack/utils.rb:127:in `rescue in parse_nested_query'
  [FILTERED]/www.whatdotheyknow.com/alaveteli-2017-09-27T11-56-09/vendor/bundle/ruby/2.3.0/gems/rack-1.6.8/lib/rack/utils.rb:117:in `parse_nested_query'
  [FILTERED]/www.whatdotheyknow.com/alaveteli-2017-09-27T11-56-09/vendor/bundle/ruby/2.3.0/gems/rack-1.6.8/lib/rack/request.rb:371:in `parse_query'
  [FILTERED]/www.whatdotheyknow.com/alaveteli-2017-09-27T11-56-09/vendor/bundle/ruby/2.3.0/gems/actionpack-4.2.9/lib/action_dispatch/http/request.rb:339:in `parse_query'
  [FILTERED]/www.whatdotheyknow.com/alaveteli-2017-09-27T11-56-09/vendor/bundle/ruby/2.3.0/gems/rack-1.6.8/lib/rack/request.rb:191:in `GET'
  ...

-------------------------------
Backtrace:
-------------------------------

  lib/strip_empty_sessions.rb:13:in `call'
Updated 09/10/2017 14:48

Equipment page

kkragenbrink/roll20-writh-custom

A secondary sheet page which contains equipment slots; possibly including where the equipment is worn.

<!— @huboard:{“milestone_order”:0.9969049545486349,“order”:0.9965062922373764,“custom_state”:“blocked”} –>

Updated 15/10/2017 02:19 1 Comments

Need to refactor AllSynapsesDeviceFuncs_d.cu

UWB-Biocomputing/BrainGrid

Issues #173 and #195 are connected to this. It appears there are common device-side functions that are shared among different synapse classes; these functions have assumptions that aren’t always valid. Need to look at the requirements for dynamic spiking synapses and for Izhikevich synapses and come up with a more general solution, or refactor this code so they each have their own functions.

Updated 02/10/2017 20:28

add blog not found page

jens-meiss/blog

if no blog was found under the url “/blog/<blog-path>” a error page should show the issue e.g. “Sorry, no blog found with path /blog/<blog-path>. Maybe check the popular blogs instead. ”

<!— @huboard:{“order”:6.987411962024189,“milestone_order”:6.986014689226197} –>

Updated 30/09/2017 20:49

Honor HTML links in popups

Esri/AuGeo

This is an enhancement request to support HTML links in the popup description. The use case is: I want to use AuGeo to find an asset, and then launch Survey123 to log an activity or inspection for that asset. Ideally, I would like to use Survey123’s custom URL Scheme as described here: https://community.esri.com/groups/survey123/blog/2016/07/30/understanding-survey123s-custom-url-scheme

Stakeholder: Clive Reece (Esri). Dakota DOT.

Updated 29/09/2017 23:02

Stronger warnings for those replying to user-user messages

mysociety/alaveteli

The current warning text users receiving a user-user message are sent is:

---------------------------------------------------------------------
[User] has used WhatDoTheyKnow to send you the message below.
Your details, including your email address, have not been given to anyone.
If you reply to this message it will go directly to [User], who will
learn your email address. Only reply if that is okay.
---------------------------------------------------------------------

a WhatDoTheyKnow user has suggested we add to that saying that your name, and IP address may be revealed too, if you reply by email. (If people have a real name set in their email software).

This strikes me as a bit excessive. How far should we go - should we explicitly warn people that anything they have in an email footer will be sent to people they send emails to?

The user who wrote to us was trying to stay anonymous; but had emails from a pseudonymous email account forwarding to their real account, and they sent a reply from their real account. This doesn’t feel to me like a problem with Alaveteli / WhatDoTheyKnow.

I don’t think there’s a need for action at the moment but we could note any similar suggestions / reports on this issue thread.

Updated 09/10/2017 14:49

Contests

theguy404/DuBot

have the ability to pick random people from a specific room or the entire server or people who say a specific phrase to the bot, and be able to !Help or something the bot for with a command to give them contest rules and information perhaps a countdown timer as well to let them know when the contest starts/ends or is on the go.

<!— @huboard:{“order”:1.9998000199980002,“milestone_order”:1,“custom_state”:“”} –>

Updated 29/09/2017 08:47

Missing data for one or more meters causes confusion

BathHacked/energy-sparks

When data is missing for 1 or more meters, but not all, this causes confusion. In the example below, 1 of the school’s 4 meters was missing data for Monday 25th August, leading the school to think that they had used electricity on that day. This is confusing, especially if the missing data then appears on the next data update. Can we flag the missing data in some way?

image

<!— @huboard:{“order”:233.4375,“milestone_order”:301,“custom_state”:“”} –>

Updated 08/10/2017 13:11

Add graphs showing frequencies wait times for responses, overall, for each body and for classes of body

mysociety/alaveteli

Related to the proposal at #4254 to show number of days a request has been awaiting a response on request pages perhaps what would really be interesting are graphs showing the frequency of responses received on each number of days after a request; perhaps overall and for each body, or tagged set of bodies.

This could be a feature under /statistics or maybe something linked from each public body page, or public body tag search.

Data could also be provided in words; eg. % of requests responded to on the 19th or 20th day, % of requests responded to within a day etc.

Users would be able to see if their request is being dealt with in an atypical way. Those commenting on the operation of FOI law (either overall, or in particular bodies) would have more information on which to base their comments.

Updated 09/10/2017 11:53

Mention Stackoverflow tag in README.md

Ardulink/Ardulink-2

We should offer a way for Ardulink users to discuss things. Stackoverflow provides such a functionality. There is at least one question with tag “ardulink” and I did subscribe on that tag.

If you agree you should add a note in README.md like “We are listening on #ardulink at stackoverflow”. If you prefer some other platforms please refer them or just close the issue if you think there is no need.

Updated 27/09/2017 07:05 2 Comments

Szeretnék részt venni ... link az adatom.hu nyitó oldalra

edemo/PDOauth

ki lett javítva a kérdőív ( kb egy hónapja) jó lenne ha ki tudna kerülni a link a második bekezdés után a nyitó oldalra: https://docs.google.com/forms/d/e/1FAIpQLSdVunlr1oXUWG1Id3mijlIaO3xuqR1qpuJdpfLHRk33uiAgaQ/viewform

<!— @huboard:{“order”:1.8595547470503598e-47,“milestone_order”:1.542860958075046e-121} –>

Updated 23/09/2017 08:44

Is the `DeadEvent` event be put back into the SRE or the SDK?

sarl/sarl

The DeadEvent was introduced into Version 0.1 in order to notify the agent that an event was not handled by the agent (a kind of “default” case). This feature was commented out because it is not documented into the SARL SDK and it causes lower performances.

The question is: should we re-introduce it, or an equivalent mechanism for noifying that an agent has received an event but has not defined an event handler for consuming it?

<!— @huboard:{“order”:2.59067098992419e-54,“milestone_order”:6.816760478688097e-59} –>

Updated 22/09/2017 15:40

Para: add SH-101 style sequence programming

perkowitz/issho

In ParaModule, create a step-programming mode where you can enter in notes from the keyboard (or from a connected midi controller?) along with ties and rests, like the SH-101 or JX-3p sequencer.

  • Hold down step mode button (or press once to start, once to stop?)
  • Enter in notes (hold multiple for chord)
  • Velocity from pad velocity
  • Use spare pads in keyboard for rest & tie
  • Current sequence keeps running until done
Updated 22/09/2017 02:06

Fork me on GitHub