@mlemweb hi! Sorry for messaging you out of nowhere. I was listening to LibreLounge and looked the hosts up on the fediverse and on birdsite, which suggested me your profile. I'm an academic in the field of music history and musicology and a free software enthusiast, so I got interested in the "digital humanities" and even more the "free software in academia" mentioned in your profile, as I've been trying to bridge these interests of mine. Do you have any resources to point me to, please?
A lot of the digital humanities work that really engages with free software is done by librarians or IT people rather than the principal investigators.
I was on a panel at last year's Libre Planet on Free Software in Academia that you might find interesting: I led a panel at LP2018 you may be interested in: https://media.libreplanet.org/u/libreplanet/m/free-software-in-academia/
@mlemweb with regards to the panel, I got greatly interested in taking part in whichever discussion group, mailing list or any other form of being in contact with people who are working in this field. By the way, I think DebConf this year will be here in Brazil. Is there anything planned related to these fields we’re talking about?
I've met several individuals at various conferences, but I haven't found a central location yet. I think what we need is more Free Software dialogue at Digital Humanities conferences.
I've never attended DebConf (I use Debian, but I'm not a Debian Developer), so I'm not sure what's planned. I attended LibrePlanet in 2017 and asked around for people working in FS & DH, and the response I frequently got was, 'why don't you do a talk on it', and I did, maybe you could do the same?
@mlemweb don’t you think we could try to create some form of “online gathering point” (I’m sure there’s a better word for this in English, but it’s not coming to me now) for people who are interested in this? I don’t know, a discourse forum, a website, a podcast... I’m really interested in working towards improving this. Most people simply don’t realize the implications...
@teinturs @cwebber @mlemweb @librelounge @emacsen consider yourself part of it already. I see you’re in the field of social sciences. Although I teach musicology/music history, my formal training has all been in history and sociology and actually I feel some colleagues in the field of music treat me as if I were not part of their group vc of that.
“Save” would do git commit -a -m and prompt for a description of what had been done (the description would be important anyway, regardless of Git demanding it). Is there such an editor already?
(There must be a better way to keep everybody in the conversation)
@fredmbarros @wolftune @aminb @eylul @emacsen @librelounge @cwebber @mlemweb @teinturs I'm not that keen on working/forcing people to work in a browser though, but it makes a sort of sense. Has anyone tried GitHub's Desktop (https://desktop.github.com/)? I wonder if the right sort of GUI could at least ease pain for some people.
@_emacsomancer the thing is that @mlemweb and I were discussing about how to make it possible for academics to do version control working on a project for a collaborative history book. Then I think it makes some sense to use a website, as people would find the “base text” online and be able to offer suggestions, alter what’s written etc.
@fredmbarros @mlemweb So the thing I'm (trying to) do is also a collaborative book project, but the book is written in LaTeX, so the website would need to be more complicated to work well. GitLab, I realised yesterday, does allow for editing files directly in the website, which might work as a stopgap.
@mlemweb @fredmbarros Re: getting academics to understand use of vc: for a collaborative project, there are other reasons as well, because it means emailing around numbered versions of the same document, which is exponentially worse than numbered versions of your own single-authored doc. There are tools like Google Docs or Overleaf, but both of these have major drawbacks, particularly the former (the latter is costly).
@_emacsomancer @mlemweb The way regular academics do this is a total mess. I was talking to two colleagues that I disagreed with demanding paper submission in MS Word format and they asked me how wrote my papers. I had barely started the whole plain-text/markdown thing and their eyes got lost and they went like “oh, no, you want me to learn how to code to write my papers!” Asking them to leave Word/Gdocs is already too much. >>
@_emacsomancer @mlemweb That’s why I think a regular editor with Git under the hood would be the best option. Some themes in Atom, for instance, show formatted markdown on the editor itself. If there were those buttons for bold, italic etc. it would do it all. Every time a contributor started an edit or addition, the editor would create a branch for them. “Save” would be a commit and then, at the end, it’d ask “submit?” and “yes” would lead to a pull request.
@aminb @emacsen @librelounge @eylul @fredmbarros @teinturs @cwebber @mlemweb @wolftune
Overleaf: putting the proprietary issue aside, it's pretty bad now compared to v1, where the Git integration actually worked. It doesn't really work anymore - I filled a detailed complaint, but....
Emacs: I was going to suggest something similar - one can set up Emacs to automatically commit and push on every save, for instance. Maybe something like Portacle, but aimed at humanities/social sciences people?
@aminb @emacsen @librelounge @_emacsomancer @eylul @fredmbarros @teinturs @mlemweb @wolftune You may be interested in what Ricardo Wurmus is trying to do with "Guile Studio" https://git.elephly.net/?p=software/guile-studio.git;a=summary
The idea is to provide a preconfigured emacs that's easy to pick up for non-emacs Guile users in the way that DrRacket is for Racket users. (Or, how spacemacs is for emacs users coming from VIM, but generalized to modern IDE concepts.)
@cwebber @aminb @emacsen @librelounge @eylul @fredmbarros @teinturs @mlemweb @wolftune There are a number of Emacs starter kits listed here: https://github.com/emacs-tw/awesome-emacs#starter-kit Lots of them are likely programming-focussed, but there is Scimax mentioned, and some others which could serve as models.
Yeah, DrRacket is one of the reasons we chose Racket for our DH programming workshops. Since it has its own text editor and interface, we could just jump right in without having to explain emacs or VIM first and half of the tutorial was focused on Scribble, because the markup language applies Racket in a way that's easy to understand and applicable to the humanities student's daily lives
@mlemweb @aminb @emacsen @librelounge @eylul @fredmbarros @teinturs @cwebber @wolftune I agree that Emacs is very intimidating in general, and most of the Emacs starter-packs seem to be aimed towards people used to some IDE or other and for the purpose of writing code rather than some other sort of text content. But I think all of the right pieces are there in Emacs; it's more of question of designing a starter-pack for more general academic purposes and making it more approachable.
There's a good guide from somebody in my niche of the academic world (#socialscience or more specifically #sociology) at http://plain-text.co -- it's opinionated (promoting R, emacs and pandoc) and not that much focused on using SCM for collaboration, but still a good reference to point people towards.
@fredmbarros @wolftune @aminb @eylul @emacsen @librelounge @cwebber @mlemweb @teinturs
The idea is to have an accessible way for people to collaboratively collate information about a topic, right? The wiki format seems a lot more approachable for that purpose, plus has media integration and change tracking already
@jboy @_emacsomancer @fredmbarros @wolftune @aminb @eylul @emacsen @librelounge @cwebber @mlemweb @teinturs
That's not to say eg: mediawiki is perfect, bc it is still slower and more onerous than ideal, but imagine an enhancement to the basic format that is to mediawiki what slack or discord are to IRC. (Minus the walled garden and proprietaryness)
@rubah @jboy @fredmbarros @wolftune @aminb @eylul @emacsen @librelounge @cwebber @mlemweb @teinturs That doesn't seem like ideal solution for a number of reasons. It seems to require internet access in order to work, whereas say a git-based solution allows for offline work (internet only required for syncing changes). And it limits the format of the interactions. I need to collaborate using LaTeX, for instance, not some other format.
Really great ! thanks for sharing, I didn't knew it whereas I am fond of any LaTeX-for-poor-and-lost-humanists...
There is also a very useful LaTeX-for-SHS-users website and free book there... mainly in French ;-). I wrote my PhD with LaTeX thanks to / because of it. https://geekographie.maieul.net/LaTeX
But it wouldn't be the first (or second, or third) tool I'd direct social science or humanities colleagues towards.
I can't think of a single colleague who wouldn't be turned off by #LaTeX source or resources talking about "noobs" etc.
The good news is that with #pandoc, it's also not necessary to take on LaTeX.
Good question – I think that's a great discussion to have! For most I think reference management & version control are most important, followed by something for document preparation, & finally tools for various kinds of data analysis and dataviz.
@_emacsomancer @teinturs @wolftune @aminb @eylul @emacsen @librelounge @cwebber @mlemweb
I think Kieran Healey's plain text guide goes in the right direction, but falsely assumes that regression analysis using #R is a central part of people's workflow. Not true for most social scientists.
There's worse things than Git, sure, but that's not an argument for it being friendly. Git's jargon and way of working are not that approachable, full of pitfalls and confusion for newcomers.
I've heard Mercurial is friendlier, but the mere fact that it's less popular and less-known is a detraction in practice.
Apart from a very few, everybody else in the social sciences seem to think that it’s MS Word or Google Docs to do our work. Some, who tend to be seen as hardcore quantativists, praise some (proprietary) software they use to treat data. The end.
@fredmbarros Do you know of any place where people are having this discussion? As far as I can see, I find it is usually driven by concerns about replicability or by fans of a certain ecosystem like R or emacs. These discussions rarely refer to free software ideals or how to do scholarship with an ethical stack, and they rarely expand our imagination of what scholarly collaboration might look like.
Is that your impression as well? If not, happy to get some pointers!
Welcome to the thread, @jboy! Though I my work is all musicological in a somewhat broad sense, I have a PhD in Sociology. In what field exactly do you work?
Mostly urban and media, but I dabble in other things too. And I work in an anthro department, though they let me keep my professional identify as a sociologist.