• 13 Posts
Joined 2Y ago
Cake day: Jun 30, 2020


I consume analytics and identify topics I am interested in via jupyter sometimes i just use ipython if I don’t want to leave the terminal – I need to build more of a frontend but I’ve not got there yet. I mostly read the articles in the terminal. And yup my plan is to find a good db but I am not sure what to use yet.

Yeah, it is odd how much weight is put into the slur filter and how quickly the discussion gets to the slur filter when the code is open source. I wish the discussion was focused past that and on more constructive issues but \ ’ ’ / . I imagine it is more of a vocal minority commenting in the thread on these topics. It will be interesting to see if there is an influx of users.

I do think there is value in the theater of this but I think it is also worth noting that the TST doesn’t have a great track record of winning these fights and has some finical transparency improvements that can be made. I don’t want to discourage people from donating but maybe split your money between them and the ACLU. I’ve by no means fact checked this article but it does seem to raise some valid concerns ( https://queersatanic.tumblr.com/post/661662523787034624/mary-doe-speaks-her-story-the-satanic-temple-a ). I feel like I saw another more common source making the same arguments but I am struggling to find it.

edit: spelling is hard

I think this is probably true for most providers. They could add logs if they were legally required but don’t actively keep them. I think there is way too much stock put in the ‘we don’t log’ comments that are common amongst privacy tools. Most VPN providers can log if they have to and often do log some data for service abuse and load monitoring but quibble over the definition of what ‘we don’t log’ means. I used to work for a VPN provider where we kept statements in our privacy policies about some logging and users ripped us apart despite these comments being truthful + other providers being dishonest ( or at least confusing ); but since so many providers provided false confidence via slamming all over their site that they don’t log the user base buys into these statements as 100% true ( and unchangeable ) and providers that try and provide a realistic view of what can happen get slammed. I am happy to see that proton put the statement up. I would have preferred they had statements up already but just because another provider says they don’t log I wouldn’t trust these statements. For me, I am not too worried if the provider can log some data like ip when they receive a non-avoidable court order ( https://en.wikipedia.org/wiki/United_States_Foreign_Intelligence_Surveillance_Court ) as I generally expect this to be true for all services and my threat model isn’t to avoid three letter agencies. If your threat model requires avoiding three letter agencies then trusting almost any service provider is going to be difficult. Obviously you should be using tor to connect to anything but you would have to assume almost everything with a server is either compromised or can be given certain court orders. Using services like briar seem like your best bet ( https://briarproject.org/ ).

I am not happy with it yet but that is because I want it to be perfect and it never will be but I do find that I engage with content at a larger scale and more varied than I do when I go to a single source. I am using the nltk features from newspaper for key word extraction + the trending sources to monitor a few hundred sources. Currently I store all the meta data + links ( urls ) + wikipedia links in a pandas dataframe ( which is becoming a problem ) and visualize trends and data about news in a jupyter notebook. For the enhanced summaries + named entity extraction I am using spacy (https://spacy.io/) from there I use SPARQL ( https://en.wikipedia.org/wiki/SPARQL ) to query dbpedia (https://en.wikipedia.org/wiki/DBpedia) to augment entity knowledge ( ex: adding data about the size , industry of a company or summary explanations of scientific concepts, etc ). The named entity matching and augmentation is the portion that needs the most work. Newspaper has some nice caching features so I query all sources everyday but only pull in new articles.

I might play around with moving portions of the data into a graph db and some better ways to query based on concepts. Right now I just write python code to query the pandas DB based on different parameters.

Are you happy with your solution ? Can you share a bit more about your pipeline?

On my phone i use feeder ( android, not sure if it is on ios ). On my computer I use newspaper3k ( https://newspaper.readthedocs.io/en/latest/ ) – I built out some additional summary tools and nltk tools that allow me to find article on similar topic from sources with different bias + some named entity extraction that easily joins into dbpedia. I intend to contribute the additional features I’ve added but haven’t done so yet as the code is rough.

so in a nutshell, they forked Lemmy to something called “Lenny” because they wanted to use the n word? I have no words…

this feels like a straw-man representation of what is being said. I am not saying I agree with everything or even disagree with everything that is being said on HN but I think it is healthy to have code & policies critiqued from both inside and outside Lemmy. We don’t have to agree with them and as the devs have pointed out it is trivial to fork and modify the lemmy code. Writing off the entirety of what is being said doesn’t seem healthy for lemmy in the long run. I think it is useful for us as a community to listen to outside voices in good faith ( again we don’t have to make changes or think they are correct ). For instance , some of the comments about making it easier to add specific slurs to the slur filter seem valid to me and to the devs ( given they they merged in a PR to do so ). I guess I am responding to more than just this comment. I hope Lemmy doesn’t turn into an us v. them community where we write off the them so easily.

initially, just to support tools that federate or intend to federate + align with product design principals I can get behind. Now , I find the content to be more relevant to my interest , aka the signal to noise ratio is better.

the pixel phones have fantastic hardware security which is why graphene + calyx support them.

Here is graphene’s explanation for why certain devices are supported and why some aren’t:

That being said, I love the intention & ethics of the fairphone so I guess it depends on how far down the rabbit hole you want to go and where your relative values sit ( ethically sourced / repairability v. hardened security ; hopefully they will cont. to converge. ).

In terms of giving money to google, you can always go the refurb route which depending on who you buy it from may or may not directly support google. But, it is a good point. Even if you buy refurb you are at least indirectly supporting google.


the pixel phones have fantastic hardware security which is why graphene + calyx support them.

If you want to have a privacy respecting phone, e/OS might keep more functionality thanks to microG.

calyx os, which is a fork of graphene, also optionally offers micro G upon setup. I’ve used both calyx & graphene – I needed certain services for my work so I ended up on Calyx with MicroG turned on, which is a privacy sacrifice but so far has been rock solid for me and allowed me to use the services I needed. This project looks interesting and for people who aren’t comfortable flashing their own device it may be a good option; however, I am not sure that people who can’t/don’t want to flash a device will end up being happy with graphene. You can grab refurbed pixel 4a for ( 300 USD ) and the calyx/graphene flashing process is pretty straightforward.

Here are a few resources for those that might be interested:

If I were going to make some recommendations it would be to start off with CalyxOS on a refurbed device ( save some money ), if you expect to need notifications + other google services then turn micro g on. If you can get away without notifications then graphene is probably a good option for you.

No, its has been working fine for me on several devices all day.

Elements of Statistical Learning: data mining, inference, and prediction. 2nd Edition.
One of my favorite books. It is available for free online. I keep a hard copy on my desk all the time and have gifted multiple copies to people. I highly recommend giving it a read.

Here is an interesting blog talking about some of the limitations of federated learning which highlights some of the ‘over promising’ that happens when companies publically speak about the topic: https://www.comparitech.com/blog/information-security/federated-learning/

I've recently implemented something similar to this and found the use case and problem very promising. It does seem to be a good step towards privacy & security but given model investigation techniques it probably isn't as private as it might sound at first glance -- particularly on tabular or ill-prepared datasets. Either way, it is an interesting paradigm to adopt and definitely enhances privacy.

As for koreader: I installed it on my jailbroken older Kindle, but it really breaks easy use via USB cable and desktop Calibre, which is why I switched back to the default Kindle UI. Not sure if that is a limitation of the Kindle version, but it does not seem to be made with basic use-cases in mind but rather seems to cater to some sort of power-users.

I know that it is more fragile on the Kindle in terms of newer versions of the kindle firmware breaking KOreader but I am not sure if that would have impacted you. I had it on an older kindle that when the firmware updated I couldn’t get KOreader to work again. A while back I actually looked into building my own Linux, e-ink reader using a raspberry pi and an e-ink screen but I couldn’t find any e-ink screens with a respectable pixel density for less than 500$.

Currently I use a Kobo with with KoReader ( https://koreader.rocks/ ) on it – this gives me an ssh server on the device and some integrations with calibre. Calibre lets me scrape news sites via rss feeds and destinate articles to my Kobo like a magazine. Its is pretty sweet but took a moderate amount of work to get setup and occasionally has some problems. Being able to have a true linux e-ink device would be a game changer for me… I read on my desktop using foliate ( https://johnfactotum.github.io/foliate/ ) occasionally because I like to take notes as I read but being able to use foliate on an e-ink device sounds great. The price is a bit more than I would like but also Pine seems to price things pretty fairly so I would give them the benefit of the double that it is priced fairly but I could be wrong.

Hey! I just found and started listening to this podcast. I am about three episodes in and I am really enjoying it. I would definetly recommend it to others. I enjoy how structured each episode is and the overall podcast.

We are not the Buddha’s close disciples and the implications of this upon reading Buddhist literature. I am not sure how I feel about this article. I am sharing it because it was thought provoking, I hope others find it so as well.

Six ways to deal with life
I found this read to be very enjoyable & useful. I think that you could probably just read the first point. All the subsequent points seem like variations on the first ( and last ). But, either way a good read.

I was recently gifted “The Book” by Alan watts. I will probably try and give it a read next week, seems pretty light.

dogen zenji brought zen to Japan in the 13 century and is generally considered on of the founding figures of soto zen. His philosophy was very focused on minimalism in terms of spiritual practice. I am not a theist centric individual so his text have always appealed to me. he tends to be more focused of experiencal practice with a heavy emphasis on zazen. particularly the act of just sitting without expectations, a form of meditation called Shikantaza.

I really enjoy dogen but I don’t think his text are super approachable without a lot of context. Hardcore zen is a fun read and a pretty good intro into zen and dogen if you’re interested.

What are you reading?
Hi , I want to get this community going!! So I am hoping to spark some constructive discourse over books. Everyone loves them :) I would like to hear what you all are reading. Here is my active list ( it is pretty Dogen heavy right now ): 1. The Mountains and Water Sutra: A Practioner's Guide to Dogen's Sansuikyo -- I am not very far into this book yet. I am finding it more accessible than reading some of Dogen's works directly. 2. Realizing Genjokoan: The Keys to Dogen's Shobogenzo -- I recently read the Shobogenzo. I enjoyed the read but felt like I didn't grok a lot of the subtler / contextual ideas. This book is illuminating. I would suggest reading it first and then the Shobogenzo , if you're inclined. 3. The Once and Future Witches: Completely unrelated to Buddhism but a fun read :) Anyways, these are the books I am reading. I am a Soto Zen student , hints the Dogen focus. I would love to hear other Zen book suggestions but I am also very interested in non-Soto Zen books/recommendations :) - -

I use micro + jupyter labs. I’ve gone through a lot of IDEs over the years but found this exp. to be the simplest. I do most of my editing in micro and most of my test / exp. / POCs in notebooks. I am a data scientist so my coding tends to be less on the CS side and more on the statistical POC side which is probably why I don’t need a heavier IDE than micro.

calyx & graphene are great options. If you want some playstore functionality Calyx is the one to go with if you are fine without google services graphene is more secure but also only runs on a few phones. I have graphene on a pixel 4a – i use the aurora store to grab apks that aren’t on f-droid; they have some nice functionality where you can filter for apps that are in the store but don’t require google services – a lot more work without google than you would think. But, it really depends on how far you want to get away from google. I recommend checking graphene out if you have a device that supports it.

What sources do you use for news?
I always struggle with what sources I should be reading for news (particularly political news). I don't want to read only sources that I agree with, but I also struggle finding news sources that aren't sensationalist and put forward varied view points. Here are a few of the places I frequent (criticism of these sources or other recommendations are welcome.) I don't think my political news sources are well varied so any recommendations there would be great as well. - hackernews - arstechnica - the economist - axios - MIT News - Wired

Emma Ruth Rundle & Thou - The Valley (Official Audio)
From 'May Our Chambers Be Full,' a collaboration between Emma Ruth Rundle & Thou. I find the pure energy and power of Thou mixed with the hauntingly beautiful voice of Emma Ruth Rundle to be absolutely amazing.

Purism– Librem 14
A new purism laptop. Looks like some solid upgrade: 4k , more ram, better screen / bevel ratio, build quality increase, and processor upgrade + all the previous privacy features.

Some instrumental drone metal. Earth is a great band that not enough people know about. They definitely aren't for everyone, but those who appreciate them tend to love them. Hope you enjoy.

1. America doesn't have the coronavirus under control. 2. Outbreaks in various parts of America put "the entire country at risk." 3. He [Fauci] wouldn't "be surprised" if we hit 100K new cases a day, up from 40K.