I find representing political opinion across a single left/right dimension completely useless in representing opinion and frustratingly simplistic.
In fact I'm on record as calling the 'political compass' 'somewhat valid' because showing human political opinions across two dimensions is superior to a single dimension of left vs right. But, in truth, without yet more dimensions it remains overly reductionist.
What do people use nowadays as #RSSReaders? (Does the concept even still exist?)
It’s to keep track of relevant #ResearchPapers of interest as my main source used to be… Twitter and ResearchGate, which I’ll both be leaving before 2024
@elduvelle a lot of journals have shut down their RSS feeds. I think your best bet for papers might be something like openalex, dimensions or (unfortunately) scopus, but i bet the librarians know better
@elduvelle same reason that all the "enhanced PDF readers" exist despite everyone hating them and why all the paper metadata is technically uncopyrightable but bound by proprietary databases (and why openalex is necessary in the first place): gotta drive people onto the surveillance mediums. The papers that you read, how long you read them, what you click on, what you search for, etc. is extremely valuable "research intelligence" data that is actively and presently repackaged and sold as products like SciVal for eg. recommending maximally fundable topics to researchers and funding agencies, or to pharmaceutical agencies to poach the competition, etc. Semantic Scholar is "AI" in the sense that any natural language parsing algorithm is "AI," as far as I know it isn't part of the surveillance apparatus directly, although AI2 seems passively headed in that direction.
So #ResearchGate just made an agreement with... the predatory publisher #MDPI
I know not everyone liked it but I was actually finding ResearchGate quite useful. Now this is completely trashing the little reputation they had left in the toilet. MDPI might have a few good journals but it is overall famously known for its predatory practices like not listening to the reviewers and not having enough time for proper review between submission and publication.
This is a really disappointing move from ResearchGate that's really not in their interest. I guess they got a big check for it? Makes you wonder what other unscientific content are they getting paid to promote.
I'm going to write to them (e.g. press@researchgate.net) to ask that they cut all ties with MDPI and any other predatory publishers. I hope that many users will complain too. If it has no effect, I'll just have to close my account and never go there again...
"MDPI’s commitment to delivering a high-quality service for our authors" :rofl:
I guess you could say that not having proper peer-review is a service to the authors in a way?? 🤔
Apparently #MDPI doesn't hesitate to pressurize your employer if you criticize them:
"Beall's list was shut down in 2017. Beall later wrote that he had been pressured to shut down the list by his employer University of Colorado Denver and various publishers, specifically mentioning MDPI as a publisher that had "tried to be as annoying as possible to the university so that the officials would get so tired of the emails that they would silence me just to make them stop.""
Anyone knows of a #Mastodon app that allows you to do drafts one or more #threads ? (Not facebook - actual threads) @feditips ? 🙏
Edit: to be more specific: I’d like to be able to create a thread in the app, so a sequence of posts (not just 1 post) that can each have text, images etc. and are in a specific order, and if need be save it and come back to update it later, and eventually post it.
Also, I am using iOS and desktop (Windows). If you check the comments you’ll see that Fedilab seems to allow this but for Android.
Yeah, as someone pointed out these new names like "Threads", "Meta", "X" etc are guaranteed to cause confusion and they're impossible to filter out because of this.
projects that you’ve already almost completed (hiding the fact that they’re almost completed, so that you can use the funding for completely new projects), or
new projects that you haven’t really started yet (for which you might just have some preliminary data)?
Do you think 1) is unethical? Do you think 1) is necessary? Do you think 1) has the highest chances of funding? Please comment :)
There is a third option - write about almost completed projects, but don't hid the fact they are almost completed. Some funders are well known for only funding projects that are almost complete - the joke being that they fund the last figure.
Now the onus is on removing from evaluation criteria the "it's a <glamour journal> paper" shortcut, and disqualify anyone using that from membership in search committes or grant/dept./unit evaluation panels.
And to further drive the stake in by raising the issue with high-retraction rate journals, among which many glamour journals. Publishing in a glamour journal should sound suspicious, should require special dispensation to be included among evaluation criteria for recruitment/promotion/awards.
So is this the deal with the new @eLife system:
1- you send your preprint
2- it is reviewed (if they so choose)
3- you eventually upload a final “version of record”
OR
4- you send the reviewed preprint (+reviews) to another journal for publication
Has anyone tried that step 4? Do the “other journals” accept to publish something that’s already been reviewed by eLife?
Of course, you could do 3 instead but does having an eLife paper under the new system “count” for your CV?
100% we need more synthesis that is both scientist-facing and public-facing. i'm excited for the transmitter to fill this niche (& thank you for the kind words @NicoleCRust -- this is exactly what i am trying to do!)
[another org i was writing for did a very cool thing in this arena with themed "reports", see here for example https://massivesci.com/reports/drugs/]
@elduvelle I mean, what is there that Windows does, that Linux can't as far as just features go?
The main reason why Linux has been struggling in the desktop space, is simply that there are many companies that don't care to support Linux, and as such you just get a plain more painful experience trying to do the same things.
It's not really Linux's fault, but rather that of a self perpetuating monopoly. If enough people actually make the jump eventually, this can look VERY different.
@elduvelle@jonny I kind of feel the same. Basically, everything takes longer with Linux, sometimes much longer, and unfortunately these days that's time I just don't have. This is a shame for me because I like the philosophy and spirit of Linux much more, and I appreciate how much more hackable and composable everything is. It's in my plan to get round to installing WSL and trying to use more Linux stuff when I can, but it hasn't happened yet and it's been on the list for several years at this point.
Grant-writing question - isn’t it weird / stressful to put all your ideas out there, for a 10% chance to be funded, while at the same time taking the risk that other researchers (intentionally or not) might use these ideas for their own benefit?
(Of course this wouldn’t be a problem if we weren’t all competing with each other.. but we currently are) #Academia#Research#Science#Grants
@elduvelle There is this concept in Taoism of letting go of owning your ideas in order to give them immortality. I really quite like this way of thinking about it.
@elduvelle Understandable concern! A few positive points, admittedly from NIH review perspective:
NIH study sections need to cover a wide range of expertise. Few on the panel work directly on the same topic as you. That greatly reduces the people who might run with your idea.
Reviewers see you have fleshed out the methodology and setup and that you already have pilot data. They know you are way ahead!
I would guess it is super uncommon for someone to feel they want to steal an idea.
Question for all tenure-track or tenured #Academics of any field:
Do you have to teach (at least some) in your current position?
Would you prefer not to do any teaching?
@elduvelle There have been several attempts at unionization, but there are way too many senior faculty whose entire careers are built on flattery of the system as it is. It would have helped, but only if there had been the kind of faculty who would work in solidarity with one another.
@elduvelle Let people do what they're good at: research, teaching, or both. It makes little sense to force people to do work that they're not good at or not inclined to do, yet universities do it all the time.
I am back using #Linux#Ubuntu and am already (figuratively) crying after 10 min because I simply want to install something but clicking on the downloaded install file doesn’t install it, nope, it just tries to open it as a text file, and I just wonder WHY do they have to make everything SO DIFFICULT
-] Tasks “manager”: tried Obsidian, One Note, Google Keep, Notebook… I can’t seem to find The One that works for me (and syncs with my laptop & phone). Taking suggestions!
-] Something to read and annotate pdfs: I got a Remarkable2 tablet, and then an Onyx Boox Note Air 2, and they’re just too slow and do not easily sync with my laptop and Zotero. For the most important papers I have to say I might prefer paper.. Also taking suggestions for this.
Edit: trying to make the bullet points not disappear in #IceCubes is still hard
for a task manager/todo list, I've abandoned digital and gone to a paper notebook. syncing issues disappeared:)
I found the #BulletJournal system very useful in providing a flexible, minimalist structure to my notebook. See #RyderCaroll's book or YouTube channel for details. be warned though, a lot of people have become quite taken with the name "bullet journal", and use it to describe very complicated scrapbooking hobbies that don't have a lot in common with Carroll's actual method
Question for #Python#Programmers:
Getting tired of my #Spyder version that relies on different versions of Python modules than the ones I need for coding…
What is your favourite #IDE for Python (and why?)?
Ninja-IDE doesn't have all the bells and whistles but it was cool. it had a zen-like quality. i haven't used it in a long time though. there was a period where qt4 wasn't working with it or something, and i found Wing IDE. i really like Wing. and i prefer it to PyCharm personally. it feels groovier or something. but i used Wing first and got used to it. i think if you got used to PyCharm, you'd probably prefer it to Wing. these days i just use VS Code
Please help #Coding question 🖥️
I am trying to install #Mountainsort (4) on Windows (10) (I know, I know).
It works on one of my Win10 computers, but doesn’t on any other ones that I’m trying to install this on.
When running ‘pip install mountainsort4’ I get this error: “command ‘some_path_on_my_comp\cl.exe’ failed with exit code 2”
I also get this error: " C:\some_path\Local\Temp\pip-install-kykt_efl\isosplit5_3fce77eb41cf47b4a1d4ca22afae35a2\src\isocut5.h(19): fatal error C1083: Cannot open include file: 'stdlib.h': No such file or directory"
I have googled this to death and tried everything and it just won’t fix itself. The cl.exe path exists. I added it to the environment variables (but maybe not properly?). It’s a Visual Studio function.
Seen this before? Any suggestions? 🙏 #Python#Programming
Edit (and thank you for all answers so far!):
I have tried running it in the visual studio console, it shows the same error.
I have tried running it in as administrator, same error..
I have tried replicating the conda environment from the computer that works onto the others, but it comes up with the same errors as when I install manually. The only detectable difference is that the working install has an older version of conda / anaconda.
I asked the repo owner and they said to use it on Linux (or using the windows subsystem for Linux), but I’d like to try and keep things as simple as possible (both for me and other users), and this is NOT simple. I know it can work on Win10 so I just want to replicate that.
Edit 2: I reinstalled Visual Studio stuff using these instructions
NOW I do not get the error about the missing stdlib.h anymore, and I can actually install isosplit5 which is a dependency of mountainsort, BUT - when I go back to running pip install mountainsort4 it now has new errors! (progress?) They look like syntax errors as described in this issue and they still end in the cl.exe error. At the end it said that it did build mountainsort4 but not isosplit5 and when I try to actually run mountainsort functions they say it's not installed. so Fail.
Edit 4: Am I just using the 'wrong' version of those Visual Studio functions for the code? Like the code has not been adapted to current versions? But how can I know which version the code wants?
Edit 5: if I manage to install isosplit5... why does mountainsort try to 'build the wheels for isosplit5'? Is there a way to tell it to use the one that is already installed?
Edit 6 (Aug 15): I tried my pipeline in the Windows Subsystem for Linux. It had a C++ problem at first, but it worked after I ran sudo apt install build-essential.
Anyone knows about an equivalent to that instruction in Windows?
I think that would fix my problem... if it exists
@carparc Hi! Thanks again for your help on this. I ran the mountainsort install (pip install mountainsort4) from the visual studio developer command prompt (in the conda environment that I previously set up). I got the same error as outside of that command prompt...
I also noticed that the problem was linked to another module, isosplit5, and so tried to install that one separately: I got the exact same error message, which I will paste below in case it's useful...
`C:\some_path_to_temp_folder\pip-install-kykt_efl\isosplit5_3fce77eb41cf47b4a1d4ca22afae35a2\src\isocut5.h(19): fatal error C1083: Cannot open include file: 'stdlib.h': No such file or directory
Any idea what these errors mean? I looked into that temporary isocut5.h file (had to quickly copy it as it disappears during install) and I guess the offending command is
"# include < stdlib.h >" (added spaces to prevent formatting)
But I'm not sure what to do from there.
@elduvelle@odr_k4tana @feditips
I just encountered a time consuming process on 🦣: a paper I boosted got published and I was hoping to replied to you original booted post so make the information continuous...it took me 30 mins to find the post only becus I had an earlier reply that contain images so I can retrieve from my media tool in #tusky, I had tried #tootfinder and #justmytoots neither of it helped me....and unfortunately my instance cannot afford #elasticsearch.😞