In the intros some people mentioned why they ended up using SilverBullet and I realized this is super valuable information for me to decide where to focus my energy and where to take things next. There are a ton of note taking applications out there so SB too has to pick its niche.
So I’m interested: what are the key selling points of SB for you? Self hosting? Open source? Browser accessible? No lock-in due to it just being markdown files in a folder? Offline capability? Extensibility? Other things?
Speed!!! Silverbullet is damn fast. Browsing through pages, queries… everything
keyboard shortcuts. I can use SB without lifting my hands from the keyboard
I can tweak SB to my needs without much effort. Many tools are based on a philosophy, (Ex: GTD) and they are very difficult to adapt to your own needs. That means that i failed many times following through with many tools because i didn’t manage to click with its philosophy|workflow
browser based, making it completely agnostic to any tool, i.e, there is no need for a specific closed tool (ex: obsidian) nor from any other tool than just a browser (no vim, emacs, vscode)
multi-platform. As a linux and Android user, i love that i am not an outlier and can use SB like any other person.
as it is browser based, i have the same experience both on computer and mobile (very often, other tools have a poor experience in mobile. Example: foam was mostly impossible to use in mobile as it requires vscode)
self-hosted, making it agnostic from server outages (apart from my own)
offline. I love to be able to use it in any circumstances and know that whenever i’ll get a connection, it will sync
git integration, giving me redundancy backups (1st the self-hosted server, then git, and then normal computer backups)
extensibility|power. I cannot start praising the power you are adding to SB, in every possible way
docker alternative
Silverbullet has taken away my G.A.S to find new tools.
This is the one.
This is the biggest strength it has, because finally I am using a tool for working and not working to find a tool that could work for me.
P.S: Things that could make SB even better for me:
Improve Search: from live search features, to fuzzy search…
I have in mind things like broot, one of the best tools i use due to its implementation of fuzzy searching, plus file management
More table features. (justification, multi-line, bullet points within cells, etc)
Thanks so far! To be clear I’m not looking for a full feature list, but am specifically interested in what are the most important ones to you personally.
If it’s everything that’s great, but I doubt everybody cares about all aspects equally
I did outline all the things that are the most important for me
Missing one of them would make me consider whether there would be a better alternative out there.
Ok, maybe the “docker” one is the only one i could live without
I haven’t yet really used silverbullet, but can details what made me stop and look at it.
I am currently using trilium notes and there are 2 things I dislike about it:
many features feel/look like unpolished prototypes
the data is saved in a database
So what’s important to me is, and seems like silverbullet checks all:
web based: I often want to take quick notes from various workstations and I do not want to have to install a client for it, or needing my laptop with me at all time only for taking notes.
text based (md) formatting (and this is where trilium is lacking here): I want to be able to save all my notes in a git repository and explore/edit them regardless which tool I mainly use to create them.
self hosted: my notes may contain private/sensitive information and I don’t want to have them on a 3rd party server.
One major “selling poing” for me is the start page (Home). Most note taking apps inkl Trilium, Joplin and Logseq lack it. There I enter the issues / links to pages that are important right now. Extremely useful!
Then all the other stuff that were mentioned already incl that it is flatfile, based on Markdown files and the PWA for mobiles. And another selling point: the enthusiastic main developer!
(Something that sometimes scares me are all the directives and other stuff etc, it can be very (too) technical sometimes for someone like me who is neither developer nor sys-admin)
Thanks for all of this. Let me draw some conclusions here and see what you think.
My observation is that everybody and their grandmother is creating a notes app these days. There are dozens if not hundreds of them. Not one is great for every audience and every use case, and although SilverBullet is — of course — a silver bullet it too needs to find its niche, or I will go crazy catering to all possible audiences and all possible use cases.
What I’m hearing in this thread, what I see in the chat and the feedback I get is that people appreciate:
The fact that this is open source
The fact that ultimately this is a bunch of files on disk on a server that you self host. This puts you in control, not locked in or dependent on a third party (but also screwed if you don’t do back-ups ) and gives you a legit reason to buy a Raspberry Pi.
Since this is a web app, you can access it from anywhere from a web browser without the need to install anything. This is not only convenient, but even essential for some of us in “hostile corporate environments” where you’re not allowed to install anything locally.
The ability to make it your own: through templates, queries, settings.
Single-user use case: a personal knowledge management system
All of these are things generally important to a pretty technical audience. You don’t need to be a programmer per se, but to get the most value out of what SB today, the idea of writing a query or learning what a template is shouldn’t scare you. And you should be comfortable SSHing into a server to boot up a docker container.
Effectively, the sweet spot that I see right now is that SilverBullet is effectively a “hacker’s notebook” or “hacker’s personal knowledge management system”.
A hacker is a person skilled in information technology who achieves goals by non-standard means.
You use your tools in creative ways. I think is a fruitful environment for technical creativity, and I want to expand it further in this direction.
This would also mean I would deprioritize as in: not do the following (things I have also been toying with in the past):
Shared/cloud-based hosting (so you don’t need to install it yourself). I think I may still add the ability to more easily host multiple spaces (with different accounts) on one server (I’m almost there already), but not more than that. I have done some work on end-to-end encryption, but I’d also deprioritize this, because it’s less relevant in a self hosted environment.
Team-based use cases. Real-time collaboration. Elaborate versioning. There’s a ton of work in this.
Fancy UIs: nice GUIs for settings, visual table editors etc. This is good, because I’m pretty bad at such things.
I am happy you de prioritize shared environments, and features for not so savvy folks.
I like that you focus on the hackers’ niche, which i don’t think is that small. Anyone working with computers can nowadays turn up a container.
As for the sharing features, it is incredibly complicate to convince others to come to your garden. It is exhausting and not very rewarding.
With the ability of using markdown and html rendering I am capable of using SB at work without the need to convince my team to use SB too.
We use Google workspace so i can easily transfer my notes to G.Docs, Gmail or any other tool simply by copy&pasting the markdown, as pretty much any tool nowadays can render markdown.
If not, i copy the html render.
So I am happy that you focus on providing more enhancements to the already powerful tool, with features that would allow all of us to personalize our instance to our needs.
I’d rather have easy sharing between us (Libraries) than sharing with others, like collaborative features with Google workspace.
Hello.
What I miss in the current SilverBullet release?
I work as a system analyst and intensively use the project under discussion .
The developed data including text, many illustrations and diagrams, I publish on the dedicated server in the local network of the enterprise .
Search results and analytics are intended for different project teams and their members .
For me, preferably and rationally :
Provide the remote user with an individual link to a given page without authorization requirements, while excluding access to my working and intermediate materials, analytical reports performed for other commands ;
Form the project web pages suitable for simple transfer (copy / past) into open office packages. The method known to me from other SilverBullet users “Print PDF”, converts text to PDF, but does not display illustrations (I suppose, in my case, this is due to the placement of SilverBullet on a dedicated server without the OC Ubuntu 22 graphic environment) ;
The abundant inclusion in the document of many illustrations, diagrams, tables leads to a frequent re-regnation of the page, which, when you enter the text intensively, leads to a loss of focus on the current edited paragraph ;
Expand the SilverBullet project with the PluntUML plugin (missing Mermaid functionality).
Absolutely!
And I think this plan is a good alternative to most of the other similar project. I’ve see very few ones focusing on self-hosting + markdown + web-client.
Mainly, I think, because most of the other project want to get revenue (understandable) by proposing a service, and the usual route is cloud hosting with paid subscription, or closed source with paid features (e.g. obsidian).
These sound good, but my only hesitation is deprioritizing looking good. You can take that to an extreme in either direction. I know that this can lead down a needless hole in many cases, but some of that can vastly improve usability. Lots of things that are trivial on a keyboard really suck on mobile. At least being about to append rows or columns in a table with a slash command would be nice. And the frontmatter editing would maybe benefit from some polish. Also just because you’re not good at something don’t mean someone else can’t contribute something for it eventually. I agree that obsessing with themes and graph visualization is unnecessary though.
This is probably overkill, but there’s TerminusDB that would solve quite a bit of these problems in collaboration (comments, ‘pull requests’, versioning). This is a db, with a lot of decisions and code (in a bunch of languages, including prolog… and perl). Not suggesting silverbullet should use terminusDB as a DB. But perhaps some good ideas there apply to silverbullet (the append-only db, where you can always go back in time, sounds nice per doc)
That seems like a giant piece of machinery to add on top. Not sure that’s the way to go. I still think there is a lot of values in storing everything as plain text files on disk over using an (admittedly) more flexible database backend:
Data locking is avoided
More hackable: you can use any other external tool to manipulate the content of your space, so not everything thing needs to be built into DB
… and that includes being able to use version control systems like git, which work perfectly with text files
For me it would be interesting to improve the searches.
For example, being able to indicate that all the indicated words appear and that it does not contain any others
For example:
That contains “diary” and “book”
That contains “diary” but not “book”
That it contains “diary” and “book” and not “person”.
Possible notation:
+diary +book
+diary -book
+diary +book - person
Options for:
Case sensitive/insensitive
Accents insensitive, search normalized (“camión” match “camion” and “camión” and “camion” match “camión” and “camion”)
Another improvement is add translations for interface.
Read your documentations and watched your videos several hours. It’s a nice tool but for me the switch from logseq is not possible without better image management (all images in one folder (easier if you move notes,…), a possibility to delete images which are orphaned, image resize in text,…)
