(untitled)

where do you get scientific journals? private trackers? libgen? I want an rss feed/auto downloader thing so I can keep up to date, does anyone have anything similar? idk if this is the right board

Sci-Hub man. You don't need anything else. And for rss, most journals have rss feeds, and if you're interested on a single topic, sites like PubMed allow you to make custom rss feeds from keywords.

>>33500 thanks dude pubmed just works for biomed though right? if I wanted an RSS feed for compsci(etc) Iwould need to do it per journal? thats fine but it would be handy if there are other sites that allow for keyword searches

>>33500 >Sci-Hub man. You don't need anything else. That may be a great resource. Nonetheless that second sentence is dangerous. Maybe it has all you need (of a certain context), but the control of that is with them, not you. That sentence promotes a monoculture---irregarding whether you meant that, it may be read that way: use this, and everything else is unnecessary, (and anything more will be wasteful, so don't use anything more)---and (externally determined) monocultures are dangerous: what happens if the monoculture goes bad?; you have no control of it: it could go bad.

>>33502 how to locally mirror sci-hub?

>>33502 It was just a cheeky comment, but you're right.

>>33502 a bit of an overreaction. most people know better than to put their eggs in one basket.

libgen, sci-hub, open-access journals(they usually rate-limit downloads), and journals I get access to through my (former) school. Sci-hub has a lot of papers but not nearly all of them.

>>33510 You need to be careful with open access journals, many (but certainly not all) of them accept low quality articles; most are pay-to-play and feature very little peer review. Some sciences are too hard for this to be a major concern (mathematical based, like applied physics, and computer science), but is dangerous for things like psychology, sociology, and medicine.

>>33511 Lol, math has the most cranks of any discipline. inb4 >math is not a science

>>33522 Math is either correct, or incorrect. Not much room for crankery. Nice try though fool

>>33502 I wish websites which REALLY needed a .onion website had a .onion service, for example, scihub

>>33499 Many CS articles are also directly published on arxiv.org

>>33511 I think the difference is not that those sciences are "hard", but rather that they have strong traditions of open access. Math was some of the first academic stuff to be put online for free. >>33525 it is also hard to verify if something is correct because so much of the terminology is highly specialized. For many new math research papers, there are 20-30 people worldwide who are truly qualified to assess it for correctness. This is why the peer review process is so important. OTOH in math you don't get soykaf like P-hacking so that's good, and there is usually a long period between when a groundbreaking result is published and when it is accepted into the mathematical canon.

>>33501 Lobst.er/t/compsci.rss

>>33561 >Math was some of the first academic stuff to be put online for free. Sometimes I forget that this isn't the norm in non-math/cs fields. It's extremely rare to find a math paper that doesn't have a pre-print available on arxiv. I don't know why everyone doesn't do this.

>>33573 *new* math papers are all on arxiv, unfortunately it's still hard to find papers written before ~2000 for free online.

>>33579 I should correct this... it's hard to find them served for free without rate limiting. Project euclid and other projects do have quite a lot of papers for free online.

>>33511 >dangerous for things like psychology, sociology, and medicine. kek these fields have no standards anyways

>>33511 Sadly, but even high quality journals like ACM are not immune to crappy papers. I am currently working in the field of Machine Learning an read some papers about text summarization. The Pegasus-Model Paper claims to be better in this kind of task because of their pre-training-method. But in fact they also altered the model in that kind that it is not comparable to state-of-the-art-BART model, because of a higher layer count. This is crappy science in my eyes. Because it does not prove that their pre-training-method is better than the pre-training-method in pegasus. The results could be better just because of the higher layer count. But hey ... its "better" than ALL other soykaf state-of-the-art-models.

>>33579 Just get in contact with the authors if you do not find a specific paper online. The most authors are likely to help you.

>>33504 >how to locally mirror sci-hub? Do you have a petabyte of storage space to spare?