I don’t want to live in a world where every 6 months I need to add an extra hop to find an album on Spotify, where Google abruptly retires applications I came to rely on, where Microsoft places ads in my taskbar, where Apple keeps me from running the software I want and Amazon acquires companies to let their products, which I use, rot to death.
This also brings up another point which I think is sometimes missed - open vs closed DATA formats. Obsidian is a good example. The software is entirely proprietary but the files it uses are standard markdown, so if the software "rots" it is not nearly as much of an issue.
Contrast this with some inscrutable binary format (cough PSD cough) where its significantly more difficult to ingest the files without reverse engineering the format.
Huge obsidian fan, but I do think it’s interesting how the plug-in ecosystem interacts with the open format idea. The plug-in system is a huge draw for many people, but you can bit rot your own vault by applying plugins that define their own syntaxes. The format is still technically open, but you can inadvertently create your own custom protocol that only works in one place. Probably a worthy trade off for the flexibility and Freedom, but it’s an illustrative example of the influence of human factors on software design.
This is an extremely good point that I don't see mentioned often enough, and is the primary reason I have not settled on Markdown for my notes. Markdown is a pretty thin format, and for anything sophisticated you have to have some kind of extension, and that immediately locks you into that particular tool, even if that tool is open source. For what it's worth, I ended up using TiddlyWiki because it's completely open source and self-contained. And so it doesn't require the upgrade treadmill that other software does. But I'm just as locked-in to TiddlyWiki as other users are to Obsidian since its syntax is unique.
But I'm reasoning from first principles. Is there anyone out there that has used Obsidian extensively with plugins and has successfully migrated to any other tool?
The main reason I don't vocally object to Obsidian is that it's a bootstrapped company, they're passionate about their product, and they don't have a need for ultra-rapid growth because they haven't taken any venture capital funding. They seem like good people with the right incentives.
I get what you're saying though I'm not sure I fully agree that it truly "locks you in" - principally because the plugins are, unlike Obsidian, source readable JS files.
Meaning worst case scenario, you could just adapt these plugins to a different ecosystem.
Markdown transformers/parsers are relatively easy to write. Case in point is I have a custom obsidian plugin that converts a markdown note (related attachments) into a HTML ready for a static site generator.
I will agree this is likely beyond the "ken" of your average user though.
But I guess it really depends on the plugins you use. On a quick glance I have Admonition, Another Quick Switcher, and Mousewheel Image Zoom installed. Most of these are "renderers" rather than extending the regular markdown format.
Yeah, their approach to the business is great; I find it very inspiring and hope to do something similar with my own product if it’s financially feasible.
There's a software directory called "zero data" that aims to catalogue software/sites/etc. that allow you to completely own and control your data in respect to their usage: https://0data.app/
> The biggest challenge, then, lies in re-imagining applications like Google Docs
This is a big one for me. AN APPEAL: Does know here know of a FOSS that offers an equivalent "experience" (visual appeal, time latency) to Google Docs and Google Sheets? Pretty please?
To explain, I have tried NextCloud-based Collabora and OnlyOffice. Neither comes close to the Google Docs & Sheets "real-timeness" and "visual acceptability".
To explain, this is less a rant and more an explanation, so bear that in mind please: For FOSS like Collabora and OnlyOffice, when "actually used" with the intention of replacing "Google Docs and Google Sheets", both
(1) the literal edit-lag (click, type, enter, wait, see screen update) and also
(2) the looks-bad (ugly dark rectangles, fonts, etc.)
are showstoppers, especially when combined.
There are things the FOSS software tries, to work around these issues best as possible. For example, one mode "runs" the doc on the server and hence sends changes to cloud first, hence the lag, while another mode save changes locally since it "runs" in the browser, but then isn't as real-time (if at all) for "other users" . And neither of these issues addresses the simple visual-appearance issues (painting a spreadsheet that doesn't hurt one's eyes).
If there is a FOSS Google Docs & Sheets replacement that is "good enough", the above description in mind, please, pray tell. It need not be feature-complete. Just replicating a Word and Excel '97 type of usability, but self-hostable FOSS with "real-time collaboration that is rather seamless, i.e. not noticeably bad an any way that interferes with actually using it", would be a most amazing find, IMO.
> an equivalent "experience" (visual appeal, time latency) to Google Docs and Google Sheets?
When people say something like this, in my experience they will not be satisfied with anything but a high quality, nearly visually identical knockoff of the product that they're used to, and, after complimenting it, they'll find some intangible reason to dismiss the knockoff (at the least, for being "just a knockoff.")
> (2) the looks-bad (ugly dark rectangles, fonts, etc.)
You may think your "looks bad" is universal, but that's an illusion caused by using the exact same products from the same product lines as the people around you. "Looks bad" is an unfalsifiable excuse. You know you shouldn't be selling yourselves to these companies, but you have to figure out a way to blame it on the only people offering you a way out.
> (painting a spreadsheet that doesn't hurt one's eyes)
How fragile are we? Your eyes hurt from spreadsheets and word processors.
> the above description in mind, please
Nobody has time to figure out what you think looks bad, especially when whatever the current dominant, consumer-hostile product is fits the bill. Just use Google Docs and Google Sheets, and don't feel guilty for not being someone you're not.
> When people say something like this, in my experience they will not be satisfied with anything but a high quality, nearly visually identical knockoff of the product that they're used to, and, after complimenting it, they'll find some intangible reason to dismiss the knockoff (at the least, for being "just a knockoff.")
Because the small details actually matter in the experience.
Take Mastodon for example: I can barely see a visual difference after a repost in the icon because the "active" and "inactive" colors are very VERY similar to each other, and there is no change in the number next to the repost icon either. As a result I have to very carefully go near the screen before I repost something, and that makes a pretty annoying experience.
UX/UI in general is the largest issue that FOSS projects have, because the people actually knowing their shit are expensive to hire and the experiments needed to figure out what works better cost serious money if you want to go beyond "run a/b tests and piss the hell out of your users who are unwittingly being used as guinea pigs in a code farm".
> Because the small details actually matter in the experience.
Since when has a large company actually cared about those details?
The last time Microsoft cared was the late 00s. By 2010, Microsoft's UI decisions were squarely in the "doesn't make any sense" camp. Today, they've almost finished removing the last of those sensible late-90s UI decisions (which were, I will note, thoroughly researched), in favour of… vibes? A/B testing? I'm not really sure.
The last time Google cared was around the same time: late 00s, early 2010s. Then they killed Google Specialized Search, flattened their logo, flattened Android 5's UI (the beginning of the end for non-power-users being able to find their way around), and proceeded to flatten everything in sight. Skeuomorphism is a dirty word.
I've seen the same trends from Twitter, Facebook… actually, every big tech property except facebook.com, which has somehow managed to keep its 00s-chic multiple-window interface and colour scheme. If not for the News Feed, it might actually be one of the more usable social media sites!
Nice project! Two bits of feedback regarding your pricing page: 1) you have a typo: "signle" 2) I would left-align your lists (those that start with checkmarks)
Thanks. Yes, Etherpad is the best example with good responsiveness/latency/multi-user experience that I knew of. I had thought Etherpad was too lightweight/minimal, though.
Looking now, I didn't realize Etherpad currently has so many plugins (for things like fonts, etc.). Going to try to customize it and see how it goes. Thanks!
Wow, ethercalc.net might just work, too. Gonna have to give that a real try. Seems there are some long open issue and no recent updates as at https://github.com/audreyt/ethercalc but, it still might be better than alternatives.
This is what you get if you run BrowserBox on your device. You could also self host BrowserBox on your own VPS. Then you have data syncing without needing to be signed in to Google because your data is on your own server, accessible anywhere.
Also we specifically license it free for personal non commercial use.
If you wanted to do this personally for commercial stuff, you can purchase a yearly or perpetual license currently at around $80 and $250 respectively
Most of our clients are organizations, but we do have a few individuals who have personal licenses.
We'll never get to live in this world while people are still obsessed with walled gardens. Apple actively intervenes to keep itself at the center of our lives. Everybody else just lives with it. And this will continue as long as the US grants privileges to the tech sector by refusing to apply standard anti-trust enforcement that it applies routinely to every other economic sector.
Nothing would make me happier than a world without Apple. I'd settle for a world without iMessage.
I mean, there are many proprietary “additions” to SMS. They often don’t work well with each other and telcos often have really odd controls that they place on that line of communication. iMessage still lets you send/receive via SMS if you want all of those caveats.
I was not even aware that the narrative around local-first focused on realtime collaboration. By the time I had come across it, I had been deeply looking at local-first food systems, so local-first software made sense to me for similar reason: resiliency, decentralization, and the foundations for rebuilding local communities. I suppose the roots from CDRT makes the real-time collaboration narrative make sense.
It did just also occur to me something ironic. I had been working on remote-first teams long before COVID, and that is what I insist upon when recruiters come by with local job opportunities.
Looking back now, I don’t think async software development would have worked so well without one of the older local-first software — before CDRT — git.
Sure, we synchronize at a central point of collaboration (github and gitlab). It is having that option to do either that allows for both flexibility and reach.
I used subversion. Subversion required too much synchronization with the central server. The point is, git lets us use github, or not github. Branching is cheap, which enabled a kind of collaboration that subversion did not.
It worked just fine with Subversion. As long as the central server was fast and reliable there were no major synchronization problems. Branching was cheap, we did it every day.
Git is generally better for most workflows. But let's not pretend that it was ever really necessary for remote asynchronous development work.
I was comparing git as a local first software, and earlier in the thread, comparing it to local-first food systems. One of the properties of such is resilience. Maybe there are different ideas on what local-first means. However, when that subversion server went down, that halts synchronization. If the network was not fast or reliable enough, it slows things down. Those kinds of interruptions are exactly what was called out in the article posted by the OP.
In theory, perhaps. In the real world of team software development, Subversion server reliability was way down at the bottom of our list of tooling concerns.
And today most organizations drive their automated workflows from a centralized Git service. If that goes down then they can't deliver anything to customers, so they haven't really gained much resilience.
I think I've been operating this way for a long time: My notes, software and wiki are almost all in plain text -- source, markdown or org mode in Emacs, usually -- and open image/video formats (jpeg, mp4) and synced with a replaceable combination of services (git, Dropbox, Syncthing, etc.). It lets me switch the conduits/pipes regularly, to whatever works best for my needs at the time.
When the zombie apocalypse finally hits (or the Internet implodes), I still have everything important, in plain text.
> There are things we can do now—things we need to do now, like video conferencing and real-time collaboration—that couldn’t be done with traditional desktop software.
This is not at all true. You could have desktop software with networking that could do things like real-time collaboration. In fact, that's exactly what happens now in Office; I can collaborate with others on a Word doc right in the desktop app. It's just not how it happened since we decided to focus on webapps instead
There are some good ideas in this document. Locally working software is a good idea. But, I think that it is better to not use formats and protocols that are too complicated. Furthermore, you are not necessarily limited to only one protocol, nor necessarily one file format. Multiple implementations are also possible; someone can make a separate implementation for a different computer system, too, according to their own preferences.
> The biggest challenge, then, lies in re-imagining applications like Google Docs in a context where there isn’t a server acting as a centralized authority, mediating client interactions.
cloud-served software makes client interactions easy because everyone is running the same version of software. In local-served software there's added complexity to make sure different versions of the software can still work together.
Are you suggesting that if a Microsoft Word document is stored on a file server on a network that's disconnected from the internet, two instances of Word running on separate computers can simultaneously edit that document?
That’s actually something Apple does not get enough credit for: notes, pages, numbers, Freeform, reminders, and so on, all work locally first, with synchronisation to the cloud for collaboration.
Although, not true of consumption apps like Apple TV or Music (but you can still sync your local music in that case).
On that note, the recent proliferation of "progressive web apps" has been pretty irritating to me lately.
PWAs give this deceptive illusion that you're installing an actual application, when in reality, it's just a bunch of ephemeral data thrown into the user data folder of the respective browser. Yeah good luck migrating that to another computer particularly if the original website goes down.
There's also a baffling amount of praise around PWAs, but wrap that same damn app in Electron (giving me the ability to run it offline as well as archive the installer in case the website goes down among other benefits) and people lose their collective minds.
It’s almost like what snobby self-important dev communities like vs dislike is as arbitrary as they are in any other group. It’s just that for some reason developers are delusional enough to think that they’re actually smart.
I'm not sure why this is the project that raised this thought, but what's neat about something like this is one could see it being self and privately hosted from one's phone or mobile device, assuming it could run it.
I have been working full-time since early 2024 on a locally running app, where I stick to open file formats and SQLite for storage. I would like to make this "local-first" - I feel I'm already on the way.
It is an open source app to download pervade data and provide fast search, offline access, automatic labels, categories, summaries, etc
Starting with emails, calendars and attachments. Integrations for Slack, LinkedIn, Stripe, databases, even web crawl are coming...
I plan to have real-time collaboration on top to make money (ideally with end to end encryption).
I take privacy very seriously. I do not even have OAuth2 apps for Google to access your data - you have to create those yourself (there is documentation in the app).
reality must surprise me for this to get worse than I fear
however in my view, the underlying problem is not exclusive to software but extends all all kinds of digital assets. in short, digital property has brought out the worst of capitalism; marx had no idea.
assuming it does get a lot worse; human life will abandon city life. I like to imagine this is why Einstein allegedly said that in a fight between humans and technology, humans were going to win. then again, maybe in the future the city will be rid of humans same as we are (were?) trying to destroy all insects
Wait, what? You think my mom backs up her computer? All my kids baby photos got uploaded to iCloud years ago. The drives they were on, including the backups, no longer function. (Actually one works but I forgot the password - because I didn’t have an online password manager then either).
This also brings up another point which I think is sometimes missed - open vs closed DATA formats. Obsidian is a good example. The software is entirely proprietary but the files it uses are standard markdown, so if the software "rots" it is not nearly as much of an issue.
Contrast this with some inscrutable binary format (cough PSD cough) where its significantly more difficult to ingest the files without reverse engineering the format.
My acceptable criteria are either:
- Open source and open format
- Closed source and open format