Could Proton save academics from closed source software? [circular rant and doubts abound]

Communities Blog
, ,
21

Also… check out crossover.

Its more popular on Mac, but basically Codeweavers have a commercial version of wine, and they fund/host and commit the majority of code to the wine project (and are actually contracted by valve to help with proton), so purchasing crossover is a way to contribute.

The offer a free 2 week trial for both Mac and linux versions of crossover.

I’m not affiliated in any way, but these guys have a history of doing good things:

2 Likes
22

Actually, it’s super complicated nowadays with all of these solutions. This is a quick video on how things go sideways with all of these tools: https://www.youtube.com/watch?v=fBG3PIvoX1M

Outside of Office, everything I use I try to keep it either open source or Linux-compatible. For now, I’m stuck with Outlook and Hyper-V (I do find it useful for my use cases) for both non-opensource and linux-compatible, and Obsidian and VMware for non-opensource.

1 Like
23

Good question! I have lived for a long time on a 10mbit/0.5mbit dsl connection, so I’ve always relied on working locally and then using cloud backups once done. Web based software was not practical for a long time and I have not been given a 365 licence by any university I’ve been to so far: only full faculty would get those. I guess I am a bit set in my ways when it comes to working on local software, but I do use my nextcloud collabora from time to time. Until recently though the CWYW of many reference managers (except Mendely I think) had a hard time integrating with web based text editors.

It’s an option, but I’m sure you’re with me when I say that I’d take almost any alternative than having to get a 365 subscription :smiley:

1 Like
24

I really like this channel, I have been following it the past year and it’s always very informative. Thanks!

The only 2 pieces of the office suite I’m stuck with is Word (for the aforementioned reference manager integration) and Access (I use databases for historical judiciary data and mapping epistolary connections and the individual careears of officials). I’m less worried about access because at the end of the day tables can be exported as tables on Libreoffice and all I would have to re-do are masks.

I try to steer my research team mates towards open solutions, generally by setting them up for them. I set up our cloud sharing system on my selfhosted nextcloud, but that was greeted with the same acceptance as an herbalist in 1600s Virginia.

25

Have you tried Joplin as an alternative to OneNote and the like? It generally supports Markdown over Rich Text, though, and there are tons of plugins. I use a modified version of their install script (which just installs or updates their AppImage to my ~/.config/joplin directory instead of the default ~/.joplin or whatever non XDG desktop compliant directory they use.

P.S. I also self-host my own instance of Joplin server, but the developer provides a syncing service.

1 Like
26
27

Try it every year full on - end up heading back to Outlook… I’ll keep trying every year!

Edit: quoting response!

28

Academics often used closed source software because they can get it at a massive discount and because they often do a lot of highly specialized stuff particularly in medicine and engineering

I’m going to be frank FOSS is far too jank to make it past the general computational help with the aid of some FOSS compilers and a programming language. As a grad student I simply cant see you replacing Keysight/Pathwave ADS with a FOSS component. They just are so night and day. I wouldnt go on a save people crusade. It rarely goes well. Just use the best tool for your job

5 Likes
29

I don’t use OneNote per se, but it might be a very interesting tool indeed! I’ll have a look, thanks!!

30

My mental model for software lifetime management goes as follows:

  • Someone’s bright idea leads to an innovative new software product. Said software is licensed according to it’s current biz value and often it’s potential value
  • Over time competitive products emerge with similar biz models, this area of software development matures, standards are set, development moves on from developing main functionality to “smoothing out the rough edges” of already existing functionality. Competition puts pressure on pricing, development of revolutionary extensions is getting harder to fund. It’s harder for customers to justify ongoing payments (for new software versions or subscriptions) at previous or current levels
  • This now established field of software is so well known that open source (with sufficiently open licensing!) software emerges which implements some basic aspects. Usability is initially poor, but there is sufficient industry or community motivation to improve on existing open source base. Eventually, open source projects allow replacing increasingly large parts of commercial software. At this point it doesn’t make a lot of sense for many customers to pay for commercial software.

TLDR:

  • Every user needs to assess what functionality they want to pay for in an ongoing basis.
  • This makes sense for special interest software, such as (maybe) Keysight/Pathwave
  • It doesn’t make a lot of sense for well established software fields. Such as operating systems, office products, etc.
  • In 2024 customers don’t only pay with money but increasingly with data.
  • All commercial software looks for ways to lock in customers - it’s up to customers to escape the lockin

So, some examples to make my point:

  • In the 1970/80/90s operating systems were actively developing in scope, capability and vision. Customers were exited to pay for new user interface innovations (mouse, graphical user interface, etc.). Over the 2000/10s this field matured and merged into only a few viable players. Windows, MacOS, iOS, Android, Linux. In 2020s it’s largely impossible to get customers to pay for OS licenses. Linux is freely available and capable, it’s well supported by an industry that benefits sufficiently from its availablility that there are many contributors that keep the project alive and even advance it
  • In the 1980/90/2000s relational database technology emerged and took the business landscape by storm. Key players of that time are DB2, Oracle, SQL Server. Innovative new players advanced the industry but their ideas eventually were adopted by key players and innovators diminished. Open source relational databases emerged and made a name for themselves (MySQL, Postgres). In 2024 all key use cases are competently covered by open source software and there are few reasons to pay for commercial relational database software.
  • Ditto Office software.
3 Likes
31

I appreciate the sentiment and I share it, but given that my field really doesn’t require any specialised tools (reference managers are generally as tech as you can get), I think there’s value in learning and getting into alternatives.

Also, about software being given out at a discount, it really depends on what we’re talking about and where in the world we’re talking about. As a PhD student 9 years ago I got 0% discount on EndNote because my university (or any back then) would consider reference management tools a resource. Zotero has had some luck in very limited institutions, and that’s it.

I would never suggest replacing lab software or highly specialised tools with FOSS just for the sake of FOSS, but in the general feel of the post, could we see a future for non linux native academic software run in linux, thanks to the attention now given to Prontesque tools? Could we move more basic tools off MS and closed source? And of course by we, I mean who is willing or who could actually appreciate the savings - not everyone who works in academia is faculty, so not everyone gets the benefits of discounts or priviledge accounts.

32

I disagree here. I used to agree to this point about office software, generally, but then I tried to write an academic paper in LO. Failing that, I tried OnlyOffice, but it’s far more barebones than LO. In my opinion, FOSS office software still lags behind Microsoft and possibly even the online Google Office Suite.

2 Likes
33

I agree here, but also there is FOSS software that excels at what it does: such as Blender, OBS studio, and the Linux Kernel. I wish more projects could replicate what makes those FOSS projects so successful. FOSS projects surrounding OSes are different. The Linux kernel is already successful, it’s just the projects behind the wider ecosystem of desktop tools, applications, and desktop-oriented user space utilities - the desktop environments, if you will; those projects are in a jank state that holds back wider-spread Linux desktop adoption.

34

There are many choices for MS Office alternatives. I wrote my dissertation in open source LaTex in the 90s (at that time MS Office wasn’t even an option, yet). I think kids these days would choose the Google suite over MS.
I cannot imagine why anyone would fail writing an academic paper in LO, but then again everyone can make their own choices.
That’s my point about a mature software ecosystem.

35

Perhaps I am wrong. Ultimately, I caved and used my University’s Office 365 subscription on an old laptop where I run Windows. I’ve always had to keep a Windows machine around since we’re explicitly forbidden from taking university exams on Linux as per the student handbook. Perhaps the issue, in my case, is an education thing. In high school, I learned Microsoft Word, not LO; later I got familiar with GApps because of my school district’s adoption of Chromebooks. For university work, I didn’t often have time to pour through a LO tutorial, so when I have to figure out how to do something in LO that I generally know how to do in Word; it’s pretty frustrating - especially if you’re on a deadline. Most of the simple things I have had to do for school are relatively simple supporting documents and whatnot for some wider piece of software I made. In this case, it had strict style requirements, and I kept having to Google how to do this and how to do that; finally, I caved when I couldn’t figure out how to center text vertically in LO. A lot of things I saw seemed to imply that LO didn’t support that. In Word, it’s just a few clicks.

One day, I want to write a book, and maybe then I’ll go through a full-blown LO tutorial.

1 Like
36

well for 1. merit should be the only form of determination of someone’s weight in a community is where you start. A lot of FOSS projects are currently going down hill because politics from our world is starting to interfere. Projects who leaders cared about the project and really put their back into it are replaced often now with someone who has some wild agenda and its hurting our projects regardless of the agenda. The agenda should be software not politics

I mean one does not need to look far in the news to see what Im talking about

1 Like
37

I agree, I think that’s why so many people are excited about Cosmic. It brings a structured system to Desktop Environment development that will basically not be hindered by infighting among developers and ideological identities. They’ll do what companies do best: they’ll develop products to create value for their customers which creates a fixed goal that leaves no room for extraneous endeavors.

1 Like
38

oh yeah its gonna be a big mainstay. im looking forward to cosmic in some ways but I worry the rust language spec is not stable nor mature enough for the task.

39

I’m curious as to what the wording is here: does it explicitly restrict you to Windows, explicitly forbid Linux (but nothing else), or what?

In the first case, too bad. In the second case, if technically possible, try FreeBSD or OpenBSD. Otherwise, it depends.

40

I understand this choice: in the corporate world you’ll be using Office (particularly Word, Excel, PowerPoint, and Outlook) 99% of times, so it makes sense to become proficient in it early on. It’s an unfortunate monoculture perpetuating itself: you learn to use what everyone else already uses, leading to everyone continuing to use it…

Another, albeit far smaller, monoculture. At least you got to learn an alternative, and then you already know two sets of Office products, which should in principle make it easier to learn yet another.

I remember that feeling when typesetting stuff for university in LaTeX, now about 25 years ago. I half-assed and quick-hacked a lot of stuff just to get it done.

You don’t say how long ago this was but I guess the following thread would have the answer:

Unless you want to self-publish, don’t bother, the publisher will take care of the gory typesetting details for you.

1 Like