Just got bcced with an editor decision on a paper I reviewed. I do not agree with the other reviewers, but this is not going to evolve in a petty argument.
What annoys me, is that one of the reviews shows clear signs of written by a LLM. And it mainly re-iterates the claims of the paper. 🤔
Developers here? I want to get rid of VSC. For various reasons. One is the automatic inclusion of Copilot. And the rather annoying (often faulty) code completions.
So, I tried pulsar (https://docs.pulsar-edit.dev/ ). Nice. Almost the feel of a good old editor plus a few goodies.
VSCodium is perhaps an alternative, too. Though the installation on Debian seems cumbersome.
Thing is, there are so many alternatives. And I really do not know what I want. Obviously a good editor. Easy to use. Syntax highlighting plugins for my favourite languages (C++, Python, Shell, Snakemake, Rust, possibly a few more). A few IDE features (although, I prefer git from the command line).
So I am asking for opinions, not hints or pointers.
No hastags, and please, no flamewar.
Oops, this turned unexpectedly and unplanned into a thread.
Today, I will copy a Pratchett e-book onto my e-book reader. My son is eager to read. He is discovering this author now.
He is an only-child, so in my household, from my perspective the gender ratio is quite balanced. Not so at my former workplace. Not so in my scientific domain. I work in #HPC and the fact that there is an action group "Women in HPC" tells you there is need for it.
Here, on Mastodon, I actively seek to follow women to get not just the male perspective. I gave up on other balances (there just aren't so many non-Europeans in my vicinity and that's ok, if I want to focus on my region. Which is not to say, I do not follow non-European accounts, it is just impossible to balance the European vs. non-European ratio.). Yet, following interesting women is a bit harder than following interesting men, simply because there are fewer (in the research domains I am interested in and at the same time active on Mastodon).
So, self-diagnosis: I like it when women seem to like what I write (boosts or likes or replies). Thank you.
This seems to be another issue of the Fediverse. However, inclusive and open it wants to be, misogeny and racism are omnipresent in our societies and hence on social media. It's better here. Not perfect. We need to improve on it.
That, too, is an asset we can capitalize on.
4/n
Also, looking at my new "followers" over time, I see what is to be expected: A big jump, when joining my instance. Then less. A peak at time with more activity, then less again. 🤷♂️
And I know, I lost some over time. Some were bored by the choice of topics or the language mix. I really do not know. Some tried Mastodon and now their accounts are dormant. And this is quite alright. No one has to choose the Fediverse, no one has to like social media in general!
3/n
BTW I could to more to "boost" my reach. But I think it is really not that bad (test for yourselves: https://rstockm.github.io/fediwings/ ).
After all, this Fediverse adventure of mine is not just "a hobby". But I write "with little time" in my profile for a reason ...
2/n
Well, I do not have much Twitter experience. Jumped on that ship, just before it was sunk.
But I get it: "Science-Twitter was so much better!"
Yes, yes. Now, there are Threads, Bluesky, LinkedIn, ...., and the chaotic Fediverse. People are realizing the inevitable: Less activity here, in the Fediverse. This happens, when people follow the herd. Whether the grass is really greener elsewhere remains to be seen, though.
Yesterday, I learned that an organization I am affiliated with is posting on #LinkedIn . Presumably for better reach. 🙄
It is true: People will not sign up for a mastodon server, just because it's there. In scicomm we have to choose: Where is my audience? How can I reach it?
So much, so trivial. We also have the option to shape that outcome. By not only posting on enshittified media, by using fediwalls or mastowall (which allow accountless views), by choosing which QR-codes to share, by forgoing ai slop images, etc. etc. etc.
1/n
My first poster since more than 10 years.... #MOPRED symposium Frankfurt, Germany.
5/5
Yes, postfactual populism is dangerous. It is a slippery slope towards outright Nazism. Many societies, right now, are under thread.
Let us stand and work together to counter this thread. Work together, strive for reform, etc.
Most of all: Being together means not being alone! To me this means getting the quantum of courage I desperately need.
4/n
We were many: not just in the one city - https://www.deutschlandfunk.de/tausende-protestieren-fuer-pruefung-eines-verbotsverfahrens-gegen-rechtsextrem-eingestufte-parteien-108.html
3/n
2/n
Dear Bubble,
the far right is on the rise - not just in Germany. But that a Nazi party won an election in a German country state, 80 years after WWII, made the headline in many papers. And it concerns many of you.
I have many friends and acquaintances abroad. (Most of you I haven't seen for too long. That's life, I suppose. And no excuse.) Most of you have been to Germany and you know that this country is more colourful than it used to be.
Yesterday, I have been on one of many demonstrations which took place at the same time. I want to share some impressions to let you know that not all hope is lost and that the civil society is holding strong.
I will start with showing my "best of" collection of protest signs.
edit: I attempted a translation in the ALT texts.
#farRight #Germany #antifa #Mainz
1/n
Anthropic offers one year free access to Claude for academics. Of course, this means groups are jumping on the #AI train, not to miss the "opportunity".
I wonder: If it is unavoidable, then are there free alternatives with +/- equal benchmarks?
To be honest, I think many groups will a) enter a stage of #vendorlockin and b) #decompetencing. However, if it cannot be avoided, it might be better to have "b" without "a".
I did some research already, but am curious what "my" bubble will have to say.
#academicchatter #spam If an email to you starts with "I hope this email finds you well", you
Today, I have accomplished almost nothing. Had a nice meeting. Its result is work. That's a worker's life, I presume. Even in academia. 🤷♂️
However, one thing annoyed me. I have a workflow, which - among other things - counts genomic reads. I was using a tool called "salmon" for it. Was dealing with so-called "long reads". Only to discover, that an update now causes the tool to tell me: "I am cowardly refusing to work as before. That flag does not do what it did in the past. Use this other tool: <link>".
Seriously???
Just returned from my holidays. Well, from the actual being-away part. One more week to go. 🌴
I could not resist and peeked into my software issues: Again one user who wrote that "Claude" helped him.
Alas, I have to assume, that no human Claude was involved. I would not mind, but I hardly understand the description. Plus, it is a wall of text. Plus, the code suggestions to fix it are identical with the issue. 💥 🤬
Seriously, I would like to help more people with their research going forward. Yet as it seems, a concise and brief description is too much to ask for these days. 🤷♂️
RE: https://fediscience.org/@rahmstorf/116945592188564262
We already entered a stage where it's too late for many things: Loss of productive soil, agricultural stability in many regions, etc.
It is now time to act to avoid total disaster. Yes, phrasing it this was this is not scientifically modest. But is it not too late to be pleasant with our dear politicians? Should we not bluntly ask "How dare you?" or "Do you actually hate your (grand) children?"?
@saranjeet this is one use case for our onboarding slide set 😉 . Hope, you found a #nanopub use case, too. Perhaps, next time, I will propose a SparQL 101 set 🤔
... and this is how it will +/- look like in the final slide set. I might extend it, so that it fits better in a 16:9 background.
Onboarding. Cartoon image for courses. Idea after https://en.wikipedia.org/wiki/Dropping_the_Pilot
Turn's out simple printer paper is good to scan and awful to sketch on.
