➡️ Cache Tokens: Previously processed tokens
This seems to be something supported by the model provider. Something like saving the processing of previously sent input tokens. Like saving a context document to answer questions about it that a user may sent in different API requests.
system prompts.
This is what is submitted to the model apart from the user message.
four gigawatts of compute power
Is energy consumption the best way to measure compute power? What about things like flops per second?
Original models: Cross Encoder Hugging Face organization.
There is one multilingual: https://huggingface.co/cross-encoder/mmarco-mMiniLMv2-L12-H384-v1
https://wikitech.wikimedia.org/wiki/Help:Glossary#General
https://wikitech.wikimedia.org/w/index.php?title=Obsolete:Glossary&action=edit&oldid=2093070
context.append(pt)
The context for the Q&A model is made of the plaintext passages returned.
search_k=-1,
Does this mean search_k hasn't been provided and will default to n_trees * depth as explained in https://github.com/spotify/annoy?
Larger search_k values will take longer to return.
nns[1][i]
This is distances.
result_depth=3):
Number of approximate nearest neighbors.
ANNOY_INDEX.build(ANNOY_NUM_TREES)
More trees gives higher precision when querying.
"set n_trees as large as possible given the amount of memory you can afford,"
return wikitext_to_plaintext(mw.parse(wikitext).get_sections(flat=True)[0])
Does this mean the page's lead will be returned if something fails?
if header == section:
If section not found (e.g., it has been removed or renamed), nothing is returned.
Get the current text for those sections via the Mediawiki APIs
Probably because the section's content may have changed since the last time the dumps and embeddings were processed.
if num_processed % 1000 == 0:
Show progress every 1000 pasagges processed.
para_plaintext.replace('\n', '<NEWLINE>')
Probably to not break the tsv.
IDX_TO_SECTION.append(section_title)
So we may have multiple identical entries for the same section, if we split the section into paragraphs.
para_emb = EMB_MODEL.encode(para_plaintext)
If section exceeds model's max_seq_length they encode its paragraphs separately instead. But they don't check again if these paragraphs exceed the max_seq_length, or whether they are under the min_seq_length threshold.
EMB_MODEL.tokenizer.tokenize(section_plaintext)
The section prefix is prepended to the section plaintext before encoding below.
Shouldn't we add the section prefix here to count tokens?
section_prefix
E.g., Page title. Section title.
section_title
E.g., Help:Page_title#Section_title
.replace('\t', ' ')
Probably to not break tsv file.
flat=True
So sections don't contain their subsections.
using the all-mpnet-base-v2 HuggingFace sentence-transformer model.
They seem to be using all-MiniLM-L6-v2 instead.
after you have trained your BERTopic model.
We could also tweak HDBSCAN parameters before model training.
you can speed up the outlier reduction process for the "embeddings" strategy
I assume by passing the embeddings through the embeddings parameter.
Using c-TF-IDF representations to assign topics
The default method.
HDBSCAN has many parameters
See https://hdbscan.readthedocs.io/en/latest/parameter_selection.html#
BERTopic(representation_model=representation_model)
Can also be used in update_topics function.
on top of the default pipeline
What does it mean "on top"?
documents
topics?
feature for topic distributions
This? https://maartengr.github.io/BERTopic/getting_started/distribution/distribution.html
if an embedding model was supplied to BERTopic using embedding_model.
Is this because we need the embedding model to embed the query?
soft clustering, like HDBSCAN.
Hard clustering assigns a data point (i.e., a document) to a single cluster.
Soft clustering instead assigns a probability to a data point of belonging to multiple clusters.
topics, probs
Stores an array of topic codes and probabilities for each document in the docs array.
embedding_model = "sentence-transformers/all-MiniLM-L6-v2"
Why do I have to manually specify the embedding model? Isn't this defined when I create the model using BERTopic()?
By default, four of the most representative documents will be passed to [DOCUMENTS].
This seems to imply that this LLM-assisted topic refinement step would be supported within the BERTopic library itself.
Volume over time of posts containing our anti-climate keywords
I think they should have contrasted this against the overall climate posts.
official:legajo=https://mapa.educacion.gob.ar/legajo/060118000 overpass:search=https://overpass-turbo.eu/...
Estas tags fueron sugeridas por Claude (el motor AI que se usó para armar esta tarea) y no habrían sido incluidas en las tareas finales.
Duplicados investigados - Casos críticos pre-resueltos
Se refiere a que corrió la búsqueda Overpass previamente. El vínculo habría sido incluido en las tareas para las que hubo posibles duplicados identificados, mientras que en las demás solo se puso verificar manualmente.
Referencias no contaminantes - Sin afectar campo website
Se refiere a que usó official:legajo= en vez de website=
Asesoría General Tutelar de la Nación (AGT).
No es de la Ciudad de Buenos Aires solamente?
En Argentina, el mecanismo nacional nos da cursos para monitorear los lugares de detención.
¿A qué mecanismo se refiere en Argentina? ¿La Procuración Penitenciaria de la Nación?
A list of transit agencies who provide public feeds is available on the GoogleTransitDataFeed project site.
It's redirecting to Mobility Data's GTFS validator.
The fact that Zn plasma levels in the positive effect and absence of effect groups were equal upon completion of conservative treatment most likely indicates the prognostic significance of the initial Zn plasma level in relation to the effectiveness of conservative therapy, but not a decisive role in realizing the positive effect of treatment.
Where do they show that levels were equal between groups upon treatment completion?
Day 90
Suman 100%?
An open, prospective, cohort (and with a unique branch phase IV) study
Sin grupo control!
create a database named my_wiki, a user named wikiuser, and assign permissions for the user on that database.
Just noting that in the config script the default username seems to be root. Maybe use root instead of wikiuser in this example?
Moore, ’92,
Is this the year he graduated? Or what?
entry point files, such as: index.php, api.php and load.php.
Of these examples, index.php and load.php are not covered in the "Entry points" section of the documentation, but rather on the "Files" section: https://doc.wikimedia.org/mediawiki-core/master/php/files.html
Social media
The difference between social networks and social media outlined here does not match the difference explained above, highlighted by the interviewees.
64 bits - Rpi 3, 4, 5
64 bits doesn't seem to be compatible with Rpi 2. I had to install Rpi OS and then install YunoHost on top of it.
pay a monthly subscription for the VPN.
An Oracle instance may be used for that. It offers unlimited inbound and 10 TB / month outbound traffic for free.
3, 4 or 5
2 not supported? Not recommended?
evidence of A.I.-induced homogenization.
on the Mark Zuckerberg-owned Meta AI app, you can see a feed containing content that millions of strangers are generating.
The Meta AI app includes a Discover feed, a place to share and explore how others are using AI.
https://about.fb.com/news/2025/04/introducing-meta-ai-app-new-way-access-ai-assistant/
SparkNotes
Tipo "El rincón del vago"?
for each kilowatt hour of energy a data center consumes, it would need two liters of water for cooling,
But this water is not polluted, is it? Couldn't it be used for something else as it leaves the plant? Closed circuits?
one issue unique to generative AI is the rapid fluctuations in energy use that occur over different phases of the training process, Bashir explains.
But how often do these trainings actually occur?
a generative AI training cluster might consume seven or eight times more energy than a typical computing workload,”
How does this compare to servers used for blockchain technologies?
it will not sell products based on A.G.I. technology.)
So Musk is using a paper published by Microsoft Research suggesting that GPT-4 may show sparks of AGI to complain that OpenAI licensed this to Microsoft.
Google has become nearly unusable for me because of those AI-generated summaries, and I'd hate to see Wikipedia head in the same direction.
I'm not 100% sure but I think the way how Google and Wikipedia would use it is different.
The proposed use here in Wikipedia directly summarizes the content of the article where you are reading the summary.
Whereas in the Google case it may be using knowledge embedded in the model itself, plus summarizing some choice among the search results.
Wikipedia has in some ways become a byword for sober boringness, which is excellent. Let's not insult our readers' intelligence and join the stampede to roll out flashy AI summaries. Which is what these are, although here the word "machine-generated" is used instead
The contributor says it's OK that Wikipedia has "become a byword for sober boringness".
They also take the (in my opinion) correct decision of calling it "machine-generated" as a way to hide what they think should be called "AI-generated" (which is an ambiguous term that doesn't help with discussing the topic).
This strategy brief was published in April 2025
See (now resolved) Phabricator task: https://phabricator.wikimedia.org/T340693
improving the discoverability of information on Wikipedia
How is this supposed to give editors time back?
I used Ollama to pull in Mistral 7B.
Is that what he used both for LLM and embedding models?
A wiki, on the other hand, allows users to simply hit an edit button
Open annotation on a public layer, such as with Hypothesis here, may be one intermediary alternative.
limited function approximator expressiveness. In particular, neural networks are only universal approximators as their size goes to infinity.
As I understand it, models are just an approximation. Unless they are infinitely large.
the availability of cheap AI-approximations might bethought to only increase public knowledge,
They don't seem to include this increase in public knowledge due to cheaper AI-approximations on their model.
java -Dfile.encoding=utf-8 -Xms256M -Xmx1024M -jar tabula.jar
See https://github.com/tabulapdf/tabula/pull/1174
May need to add --add-opens java.base/java.util.zip=ALL-UNNAMED --add-exports java.desktop/sun.java2d=ALL-UNNAMED
decenters Wikimedia tech development from the Global North.
I don't see this explicitly reflected on the specific priority's page.
communities that develop.
What "communities" do they mean?
We extracted excerpts of the interviews that related to those 6 tags and compiled them into a spreadsheet, totaling 85 excerpts
Is this spreadsheet publicly available somewhere?
Mitigate (baseline success) or reduce (strong success) the burden on experienced editors and administrators as a result of the increase in new editors using mobile devices
How?
Through focusing on improving the mobile experience, the expectation is this includes improving administrator satisfaction and workflows.
How is this supposed to happen?
what does the movement most need from us right now to advance the mission?
Who is "us"?
There are many reasons for avoiding those systems: Skill decay especially in abilities like “critical thinking” or “problem solving” comes to mind. The ecological impact of those systems is another aspect that can’t be overlooked if one cares about the world. The way those systems harness and accumulate all data and information they can get without respect or compensation for the creators
Skill decay, environmental impact and lack of compensation to or consent from training data creators are some of the reasons not to use generative AI.
prone to what the tech industry calls “hallucinations.”
What about now, with search-the-web capabilities added to these chatbots, such as ChatGPT and DeepSeek?
You should see the server immediately accept the follow request.
So this Accept activity is not sent in the response of the Follow activity, but rather on a separate delivery to the follower?
You should see your Mastodon account sending a follow request to the actor me on the console where the server is running:
I'm not getting this. Instead, it's complaining that No actor key pairs dispatcher registered.
@e875a03fc2a35b.lhr.life
It knows its domain! https://fedify.dev/manual/federation#how-the-federation-object-recognizes-the-domain-name
inbox: ctx.getInboxUri(identifier), // Inbox URI
This won't work without registering the inbox listener above first.
"/inbox"
This is for the instance's shared inbox.
run the following command to query the actor:
As is, this causes an error to be logged by the logger:
ERR fedify·webfinger·server No actor handle mapper is set; use the WebFinger username 'me' as the actor's internal identifier.
I understand this is related to Fedify's default behavior of assuming the actor's identifier to be the same as the WebFinger username. See https://fedify.dev/manual/actor#actor-identifier-and-webfinger-username
Ubuntu
Apparently, compatible images are Ubuntu Minimal: https://wiki.ubuntu.com/Minimal
localhost.run
At the time of writing this, they have suspended free plans temporarily.
It doesn’t matter if the offending account is on your server or a different one, these measures are contained within your server, which is how servers with different policies can co-exist on the network:
In Mastodon, offending accounts are reported to the server where the reporting account resides. It is this server that decides what to do with the offending account's content. Reports can be forwarded to the offending account's server, but this is optional.
If you don’t opt in, then your public posts are only searchable by hashtag.
See https://blog.joinmastodon.org/2018/07/cage-the-mastodon/ for why this is a design decision.
it is reasonable that the delivering server remove the subscriber from the followers list.
Servers may remove an actor from another actor's followers list if the follower actor remains unreachable for a long period of time.
In the case of receiving an Accept referencing this Follow as the object, the server SHOULD add the actor to the object actor's Followers Collection.
Does this mean receiving and Accept from the followed user in a client-to-server interaction?
Otherwise, if the server receiving the Accept refers to the follower's server, shouldn't the Following collection be updated instead?
retweets were shown to be reliable indicators of ideological alignment.
What about quote retweets? Aren't they used to express disagreement too?
(with authorization of the submitting user)
What does this mean? Is that the user posting the activity? But didn't them post it to their outbox first?
If they understand that forcing you to enshittify the service will send all your users packing and leave them with nothing, they will very likely not force you to wreck your service.
What about them forcing you to put the switching costs on users? That is, what about them forcing you to remove the fire exits?
frequently designed to be antiviral
How is that?
organic food movement
But I think this movement is not as strong in Argentina as it is in Europe.
Italian way
May work in Argentina too!
Mastodon could implement direct message end-to-end encryption in the future for its clients.
I guess XMPP could be used for this?
todas usan un protocolo llamado ActivityPub,
Qué hay de otros protocolos como el protocolo AT?
there are a lot of people who have shared how to fetch that)
This seems to be one of these such tools: https://github.com/YuriCosta/WhatsApp-GD-Extractor-Multithread
So far, we have created 2,200+ new articles on Wikipedia, which generate 8m views per year.
The articles in this category are: - English Wikipedia articles only. Do they support writing in other languages too? - Translated articles (or that's what the category says).
Alberto Brandolini formuló el “principio de asimetría del disparate” como “la cantidad de energía necesaria para refutar tonterías es un orden de magnitud mayor que la necesaria para generarlas”.
Para Randi, y para casi todos los otros magos profesionales, lo fascinante es que quienes los miren traten, infructuosamente, de darse cuenta de cómo están logrando
¿Es realmente eso lo que fascina a los magos? ¿Que el público intente averiguar, sin éxito, cómo hacen lo que hacen? ¿O más bien disfrutan ver el asombro de la gente?
Si perdiéramos irreversiblemente el método de la ciencia, estaríamos condenados a que nuestro conocimiento se detuviera donde está, y a ser incapaces de entender el mundo.
Parece demasiado cientificista. ¿Siempre hubo "ciencia"? ¿Cómo conocíamos el mundo antes, si no?
¿Qué buscaríamos como dato confiable? Si hay investigaciones científicas
El sesgo cientificista de nuestro conocimiento. ¿Hay otras formas de conocer?
anomalous structure (Shenoy et al., 2022).
Can't tell what they mean by "anomalous structure" in the paper they cite.
The specific sources that underpin an article can also determine whose point of view is represented
bias
trained on a massive dataset of Wikipedia references,
hence the importance of the Shared Citations project: https://meta.wikimedia.org/wiki/WikiCite/Shared_Citations
checking if references properly support the claims to which they are connected.
how do they determine what claims references are connected to?
it still wouldn’t be able to correctly match a citation to a specific claim made in a specific body of work.
I'm not sure about this. What about what Google's NotebookLM is doing? Or experiences like STORM to create Wikipedia-like articles?
Lapolaridad
Usan el valor compound que devuelve la función polarity_scores de la clase SentimentIntensityAnalyzer de la librería nltk: https://www.nltk.org/api/nltk.sentiment.vader.html#nltk.sentiment.vader.SentimentIntensityAnalyzer.polarity_scores
sedesprendeenunodeellosunatendenciaasermoderadamentesubjetivo,yaqueincluyeinformaciónmeramentedescriptivaquepodemoscatalogarcomodemayorneutralidad.
Pero si subjetividad y objetividad se derivan de polaridad, y ésta se refiere exclusivamente a el "sentiment strength", nada se puede decir sobre la descriptividad de la información incluida o la neutralidad de la fuente.
fueimportanteelegiryevaluarlasherramientasconconjuntosdedatosdepruebaparaasegurarsueficacia.
cuál sería esta evaluación?
polaridad,objetividad,subjetividad.
es una sola medida! las otras dos derivan trivialmente de la primera
objetividad
según el código fuente, es 1 - abs(polaridad)
lasubjetividad
según el código fuente, es 1 - objetividad
6publicaciones
no eran 6 por medio?
cuatromediosperiodísticos
son cinco
Véase
Se muestran 6 artículos (dos de Página 12 y uno por cada uno de los otros 4 medios). Cuáles son los otros artículos?
analyze_sentiment_nltk:tomauntextocomoentradaydevuelvelapolaridad,subjetividadyobjetividaddeltextoutilizandoelanalizadordesentimientosdeNLTK;
código fuente en https://github.com/gusunavarro/koluel/blob/master/pnl.py
lasaplicacionesnonecesitancomputadoraspropias,porlocualesunagranventaja.
Hasta ahora se habló de Streamlit como de una librería. Pero ahora parece hablar de Streamlit como un servicio.
De cualquier modo, por qué es una ventaja? A cambio de qué ofrece Streamlit este servicio?
EsteartículoexploralaherramientaStreamlit,
El uso de esta herramienta no parece relevante para el tema el artículo. Tratándose de un artículo en el que se mide polaridad, subjetividad y objetividad como aspectos del "sentimiento", habría que destacar las librerías que se usaron para tal fin, en todo caso (que no queda claro si fue TextBlob o NLTK).
Streamlitparalacreacióndelainterfazdeusuario,NLTK4paraelanálisisdesentimientos,
no dice que usaron textblob para análisis de sentimientos?
suelenmutarpermanentemente,
esto se puede resolver indicando la versión específica utilizada
bibliotecaTextBlob3queofreceunainterfazsencillapararealizaranálisisdesentimientosentexto.
biblioteca para Python
Demodoquedesdesulanzamientoen2019,hacrecidorápidamenteyactualmentemuestramásde10millonesdevistasdeaplicacionesymásde20milestrellasdeGitHub.
No es claro. De qué está hablando? Habla de Streamlit y de polaridad de sentimientos, pero no parece haber establecido un vínculo entre ambas.
Qué son las "vistas de aplicaciones"?
create a locale folder in your plugin root with subfolders for each locale,
So which locales are supported?
[signed-by=/usr/local/share/keyrings/josm-apt.gpg]
In Linux Mint 22, I needed to add arch=amd64 between the square brackets as well. Otherwise I got error repository 'https://josm.openstreetmap.de/apt noble InRelease' doesn't support architecture 'i386'.
Фон Альвенслебен
I think he may be Richard Wendler instead.
Pero César –precisa el historiador riocuartense– había nacido en el pueblo de Calamocha, en España (en Aragón)
No encuentro fuentes que confirmen esto. Por el contrario, solo encuentro que habría sido portugués.
dropping the idea of co-creating one or more natural language generation solutions with the community, suggesting to select either an existing natural language generation system, or to adopt the one that one of the fellows initiated during the fellowship.
they mean the one proposed by Ariel Gutman? https://meta.wikimedia.org/wiki/Abstract_Wikipedia/Natural_language_generation_system_architecture_proposal
comunidades de proyectos de Wikimedia
Se refiere a comunidades de "Proyectos Wikimedia" como se define en el glosario (Wikipedia en español, Wikidata, etc).
Se creará un documento similar a un Memorando de Entendimiento o Acuerdo de Nivel de Servicio entre la Fundación Wikimedia y el Consejo Global
Es como si la Fundación Wikimedia se limitara a ofrece servicios de gestión a la comunidad Wikimedia, representada por el Consejo Global.
una distribución equitativa de los recursos, como la establecida por el Consejo Global en consulta con las distintas organizaciones interesadas.
No entiendo si se refiere a la distribución que establezca el Consejo Global, o los recursos que establezca el Consejo Global.
reconocidas por la Junta Directiva de la Fundación Wikimedia.
aconsejada por el Affiliations Committee (AffCom)?
la aplicación y aplicación
"la implementación y aplicación" o "implementando y haciendo cumplir"
Equidad
Equidad en el conocimiento según Wikipedia en inglés (Knowledge equity):
Knowledge equity is a social science concept referring to social change concerning expanding what is valued as knowledge and how communities may have been excluded from this discourse
No habría que definirlo en la carta del movimiento?
As it stands, there are many small non-Western Wikipedias, and although they may not be as big or have as many citations as English Wikipedia, their perspective reflects the community that they serve.
I don't want to be naive, but it may be good to remember here that local communities will still have the choice whether to introduce the Abstract Wikipedia content in their local language editions, or not. That is, nobody will force anyone to use Abstract Wikipedia content in some way or another.
But I understand this thought may be naive.
Proponents argue that this project will allow marginalized editors to make changes that will propagate throughout all Wikipedias, not just their language.
So probably it will be the case that these marginalized editors will be able to share their view of the world in topics where these views are not challenged by other more powerful views. For example, topics relevant to this marginalized communities, and unknown to the other communities.
The reloading (rebuilding) of the graph from the Wikidata dumps takes between 1 and 2 months, sometimes more.
When does the graph need to be rebuilt from Wikidata dumps? When servers crash and content in RAM memory is lost?
requests for linked data in Turtle format
Are these and other linked-data format requests created live?
type owl:DatatypeProperty (linking to data literals). Most of these properties are used to encode different uses of over 4400 Wikidata properties, which simplifies data processing and filtering
different uses of Wikidata properties
Do they mean the same property used in different contexts, such as in statement o reference auxiliary nodes?
real-time linked data and weekly RDF dumps
What is "real-time linked data"?
two formats: a complete dump of all triples in Turtle, and a smaller dump of only the simplified triples for wdt: properties in N-Triples.
Turtle and N-Triples are data formats for storing RDF data.
Figure 7: Number of pages on Wikidata:
What do the inflection points mean? * ~March 2013 * ~July 2017: Freebase migration? Freebase was closed in 2016... Or lexicographical data from early 2018? Points in x-axis represent January or July? * ~July 2019 * ~April 2020
locations of movie narratives (above),
Some points are actually multiple points at the same place that show when you click on them. For example, many movies are narratively located in Argentina, but many of them show as a point in the middle of the country.
As a result, it may seem that not so many movies are narratively located in one place, but actually it is because the place has not been specified down to the locality or province.
Ideally, these "many-points" points may be bigger the more points they represent.
Twitter [23])
Nowhere in the cited paper does it say that Twitter uses Wikidata in any way. Instead it says that the researchers of the cited paper used Wikidata for their research purposes, about Twitter.
>1.44 billion
According to Wikipedia's article about Google's Knowledge Graph, by 2020 it had 500 billion facts (statements), that is roughly 350 times the number of statements (although without annotations/qualifiers, I assume), and 5 billion items, that is roughly 50 times the number of items.
Of course the problems with the size of Wikidata is not only technical (e.g., SPARQL queries times) but also social: how can we maintain this amount of data?
answered "roughly one-third" of the 100 billion monthly searches
I wonder how Google's Knowledge Graph role may be changing in this era of generative AI's.
this deletion within minutes did not at all rely on examining "evidence of Dr. Strickland’s professional endeavors" – rather, it was done based on the "Unambiguous copyright infringement" speedy deletion criterion,
But this seems to fail to acknowledge that in the The Signpost's article linked to above it says that once it was rejected because of copyright infringement, but another it was because of insufficient sources.
bluebell
Blubell is a plain-text language, similar to Markdown, which simplifies the authoring of Akoma Ntoso documents.
Users of our plugin said they still visited Wikipedia directly and reported that when they knew the information they were getting from ChatGPT was coming from Wikipedia, they tended to trust it more.
Did they control for the fact that users of the plugin are probably commited users of Wikipedia already? That is, how representative of the general population of ChatGPT users is the sample used in this survey?
python manage.py runserver
By default this runs the lawlibrary site.
To run another site (e.g., open_by_laws, change the os.environ.setdefault("DJANGO_SETTINGS_MODULE", "lawlibrary.settings") line in manage.py.
elasticsearch
I only found in the dokku deployment documentation how to configure the search index:
python manage.py search_index --create
Currently, auto always selects hocr.
In OCRMyPDF v15.4.4, this doesn't seem the case. When using the --pdf-renderer hocr option explicitly:
* I get a .hocr file in the temporary files folder;
* I avoid segmentation errors when reading the file with Mozilla's PDF.js;
* When text is selected using a PDF reader, I can see the OCRed text (instead of blank characters).
Some important files include:
What about an hOCR file?
Glyphless font
Probably an invisible font for invisible text overlay. Is it possible to use a non-glyphless font, yet invisible (i.e., it would only show when highlighting the text)?
PDFs have no concept of words and lines internally
Therefore, is it up to the PDF reader to determine what is a "word", a "line", a blank space, etc?
un estudio sobre "malos viajes" realizado en la Universidad Johns Hopkins han demostrado que el 84% de los participantes, cuando les entrevistaron 6 meses después de la experiencia, dijeron que se habían beneficiado del "mal viaje", aunque fuera una de las experiencias más difíciles de su vida.
Jesse, R., & Griffiths, R. R. (2014). Psilocybin research at Johns Hopkins: A 2014 report. Seeking the sacred with psychoactive substances: Chemical paths to spirituality and to god, 2, 29-43.
somoscopyleft“avant la lettre”,simplemente porque de no existir la práctica generalizada de la copiano autorizada y el compartir entre pares, la mayoría de nosotros no ac-cederíamos a los bienes culturales: no podríamos pagar la licencia delsoftware que usamos ni los libros que leemos ni la música que escucha-mos ni las películas que vemos.
Pero esto ha empezado a cambiar con la introducción de plataformas como Netflix, Spotify, Steam, etc
algunos autores viven delcopyright: al conocer quiénesy por qué obras, seguramente, dudará de si eso es lo que quiere para suvida
Quiénes y por qué obras??
Communication skills for collaborative leadership
Let's Connect clinics require registration via Eventbrite. See https://www.eventbrite.com/e/lets-connectleadership-development-working-group-x-wikimania-2023-tickets-672597927507
For political leanings, the Facebook Audience API[supp 2] provides five levels: Very Conservative, Conservative, Moderate, Liberal, Very Liberal.
How does Facebook estimate this leaning? What does Facebook say about it? Is it based on a US perspective?
Cerro Pelado Lago Club
Al día de hoy: Acampar: $4500 noche por vehiculo Pasar el día: 2000 por vehículo no hay alquiler kayak no hya bar
pero se puede alquilar kayak o ir al bar de dos predios alrededor (sin pagar el día en ellos)
gratuitamente
Gratuidad de las donaciones. Gratuidad implica ausencia de contraprestación.
medios de pago equivalentes
El texto es de la resolución general 4444 de AFIP. Menciona "medios electrónicos autorizados".
Acá una lista de medios de pago electrónicos según el BCRA: https://www.bcra.gob.ar/mediospago/politica_pagos.asp
http://si000031.ferozo.com/PDF/Acoso_escolar_2014-1.pdf
The problem
In addition, JavaScript implementations jsonpath and jsonpath-plus do not implement script evaluation safely.
?access_token=ACCESS-TOKEN"
Because it uses https, it should be safe to pass the token as URL query string.
Headers can bee used too: https://stackoverflow.com/questions/29876906/is-it-possible-to-pass-facebook-graph-api-access-token-through-request-header
An app access token
It may also be used {app-id}|{app-secret} as stated in the app access token documentation.
The easiest way to get permissions from app users is to implement Facebook Login.
So it's not the only way?
It should, at the very least, tag the tile as "done", and not expose it as a playable tile again.
that would be if decision was "yes" or "no", I guess.
Does the game provider know who is playing?
key-value-pairs to pass to the Wikidata API
parece ser bastante libre
yes=1,
cuando hay varias opciones "yes" (como en los juegos de juanman), cuál se elige con el shortcut?
low=1
at the root? { tiles: [...], low: 1 }??
add it permanently,
what does this involve? what are the steps
callback parameter
query parameter?
INVALID_EXPRESSION_ERR If the expression is not legal according to the rules of the XPathEvaluator, an XPathException of type INVALID_EXPRESSION_ERR is raised.
How could this error occur? Shouldn't the XPathExpression, whose evaluate method we are running, have been created via the XPathEvaluator's createExpression method? Shouldn't the error have been raised there instead?
Is this related to the bug described here?
Use the node.exe --preserve-symlinks switch in your launch configuration runtimeArgs attribute.
Note that this won't work if the debugger is launched via npm:
// .vscode/launch.json
{
"request": "launch",
"runtimeExecutable": "npm"
}
This is because the --preserve-synlinks argument would be passed to npm instead of to node.
In such cases, the env attribute may be used, setting NODE_PRESERVE_SYMLINKS to 1, but note that this won't work if node is not the first command in the npm script being run (e.g., do something && node).
"outDir": "./dist/",
Won't this be ignored because we are using webpack?
instantiate a new instance of the plugin in the plugins property and make sure that we set hotOnly to true in devServer.
Since webpack-dev-server v4, HMR is enabled by default. It automatically applies webpack.HotModuleReplacementPlugin which is required to enable HMR.
The following line lets webpack know we’re working in development mode — This saves us from having to add a mode flag when we run the development server.
But won't this set the development mode even when I'm building for production?
losusuarios no realizan directamente una donación al proyecto y lo hacen a la entidadresponsable de su desarrollo, que puede aplicar esa aportación para Wikinoticias ocualquier otro de sus proyectos, sin que el usuario tenga control sobre ello.
lo de que el usuario no tiene control es cuestionable, ya que puede participar de discusiones en torno a qué proyectos financiar, presentar proyectos, votar miembros del concejo, etc
sin embargo, al permitir lapropia filosofía de la plataforma el anonimato en los usuarios, no siempre se encuentrainformación relevante en estos perfiles
Creo que es relevante destacar aca que se pueden ver todas las ediciones que un usuario hizo en el sitio.
la libertad de los usuarios para ejecutar, co-piar, distribuir, estudiar y mejorar el software informático (Stallman, 2004).
No es esto el software libre en vez del codigo abierto?
Supported types are:
Some types missing from the documentation: * files (see https://wikidata-game.toolforge.org/distributed/#game=10) * url * html * map
Example JSONP:
Is there an additional options property defining game types?
"options":[{"name":"Entry type","key":"type","values":{"dog":"Dog","bridge":"Bridge","tree":"Tree","hieroglyph":"Hieroglyph","musical instrument":"Musical instrument","mountain":"Mountain"}}]
action=tiles&num=1&lang=en&callback=xyz
there seems to be a "type" parameter specifying the game version requested
Does not work on sites that enforce Content Security Policy (CSP),
See this workaround using HeaderEditor extension.
stories may also be subject to review from time to time in case changes about the topic come to be surfaced.
But because may stories may be made from a single article, and users can change the text they "imported" from Wikipedia, it would be cumbersome for editors changing some text in the article to make sure whether they have to change it in a story.
Couldn't the text become linked to the Wikipedia article somehow?
Citoid service, API and extension enhancements
Wikipedia integration
API and web proxy
now known as "server"
front-end
aka. "editor"
export them directly from your store setup file such as app/store.ts and import them directly into other files.
Where would I import them into? I guess I may need to import them into my slice files. But these export reducers which are imported into the store file, creating a dependency loop. Right?
Edit: See below; they say it's a safe circular import.
If the selector returns a different value than last time, useSelector will make sure our component re-renders with the new value.
How is this difference calculated? If the value is an array, or an object, given that state is updated immutably, won't it always be different?
in a real Redux app, we're not allowed to import the store into other files, especially in our React components, because it makes that code harder to test and reuse.
I don't understand this, because we could still call react-redux's useDispatch to get the store's dispatcher, couldn't we?
Concurrent React and Suspense are coming and raise a whole bunch of questions. Will current state management solutions work the same way as before?
This has probably been solved in React-Redux v8
you have to write whatever state handling logic you need on top of that, in order to define the value that gets passed into a context provider. Typically that's just done through React component state,
See here
for the entire component tree
Wouldn't that be of the components that have subscribed to the Redux store?
mapState or useSelector).
Both functions of React Redux.