Fleeing from Fusebase, trying to find new notes-home

Lounge
1

My journey from Fusebase, aka Nimbus, to Joplin, was driven by the disastrous pivot from a reasonably nice notetaking environment to a customer portal environment, where I lost control of my notes. The portal functionality didn't work, and it caused some serious reputational damage with some clients. Much of the damage could have been mitigated if FuseBase exported notes to Google Docs, Microsoft Word, or Markdown. They only gave me HTML or PDF options.

Like many folks, I started with Evernote. When enshittification took over, I tried notion, keep, and one note and returned to Evernote until I discovered Nimbus. Nimbus was an excellent low-friction writing environment. Best of all, it worked with speech recognition (vital if you are slightly disabled like I am). However, as I said above, they changed in a much worse way than the Evernote change.

Here I am, after spending 50 bucks a month for most of the year on the team license and a grand on the license for the portal version, spending hundreds of hours trying to make everything work, trying to recover my notes while attempting to keep grumpy customers happy, feeling beat up and cranky about notetaking environments. I'm highly resentful of subscription services that don't deliver and cost me more than I spent.

I hope that I can get some guidance from those in the know.

Core needs are simple:

  1. Simple notetaking editor. Preferably mWYSIWYG (mostly what you see is what you get). Speech recognition friendly is a double plus.
  2. Easy handling of embedded images.
  3. Easy export into markdown.
  4. Easy export to something like Google Docs or Microsoft Word (for customer copies)

Short-term urgent need

  1. Easy import/conversion of HTML into a useful markdown form[1]

Bigger needs that might be out of scope for Joplin:

  1. Exporting designated notes to HTML for a static website/blog
  2. Sharing subnotebooks/notes with clients (collaborative editing)

Needs/ideal/may be helpful to me.

  1. Task list with page associated with each task for activity log.
  2. Gantt chart

That's my story, and I hope that if Joplin is not a good place for me to settle, we can discuss other options. I have looked at Obsidian, and I found it difficult to physically manipulate the user interface.[2] It also had problems importing HTML output from FuseBase. I have looked at a number of other notetaking applications; they all were not acceptable for a variety of reasons.

Thoughts? Questions?

[1] So far, I've found that I can use https://codebeautify.org/ to convert HTML to markdown. But I haven't figured out where to put images in a way that can be found and moved easily when exporting markdown.
[2] Relatively minor mobility impairments like what I live with is alien to most people but in a nutshell, keyboards: I can no longer count on hitting the key I'm aiming for unless I'm looking at the keyboard and making a conscious effort. With mice, it takes me three or four tries to click on those little tiny targets found in modern GUIs and I can't always double-click accurately. Frequently my hand jerks the mouse between first and second click.

The design constraints on editor are subject for a much longer conversation to be had later.

2

Anything not covered is something others can speak to better.

Core needs

I don't use speech recognition on desktop. Mobile speech recognition includes the Voice typing... feature (see top-right hamburger menu while editing notes) and auxiliary solutions like those from this thread: Voice memos to Joplin text notes

File > Export all > MD - Markdown exports all Joplin notes. To export a notebook or note, right-click the item > Export > MD - Markdown.

For easy export to Google Docs and Microsoft Word, pair existing Joplin export options with an external file format conversion program like pandoc. See this thread about a similar request for context: Plugin Request: Export to RTF or Docx

Short term need

You choose the desired file format upon importing. If the import doesn't work or doesn't satisfy the need in your footnote, someone else can help you out.

Bigger needs

For exporting notes to a blog, check out this Markdown conversion app:

You can also check if the HTML export options meet your needs.

For sharing notes with clients in addition to collaborative editing, check out Pro or Teams plans on the Joplin Cloud plans page.

Nice to haves

I'd use Mermaid for Gantt charts. First, test Mermaid out by first pasting the Mermaid code block from the Diagrams section of the Joplin Markdown Guide into a new note. Then, create a Gantt chart by replacing everything but the first and last lines with the Syntax section code from the Mermaid Gantt syntax page.

View > Zoom In enlarges touch targets.

Hope this helps.

1 Like
3

This does help. At the very least it tells me that Joplin is not a bad place to be as I further extract myself from FuseBase.

I just deleted a fairly long bit of speech recognition and how it works with disabled folks. Then I read an article on HNN about a new whisper-based speech recognition system called Aqua (https://withaqua.com). After 10 minutes of playing with it, it looks like this is the first major improvement on Dragon since NaturallySpeaking came out. It is good enough it should become a part of Joplin's Rich Markdown plugin. It would probably become for a fee at that point because the whisper backend is not cheap.

.

4

Yes, Whisper and the rest of the neural-net based engines are a quantum leap from Dragon NS.
I also looked at the Aqua demo on YouTube. It is impressive but IMHO, combinatorial enhancement to what already exists in the open-source world. I have actually already witnessed a similar (not as polished) private demo, where a LLM was used to correct an existing block of text a couple of times. Add the speech input and iterate over the text correction with the LLM and we have similar functionality. Not to discount their work (in fact, kudos for the promotional video) but, now that the idea has taken form, it is logical to expect similar open-source offerings, probably within a few months.

So on the topic of speech input, to get back to the point, you do not need to pay a cent. Whisper.cpp is free and there are already multiple tools based on it out there (Windows, Linux,Mac).
I am guessing you use Windows, but if you happen to be a Linux user, I would shamelesly suggest to look at my Blurt (for GNOME) or BlahST (most Linux desktop environments) speech-to-text input tools, BlahST being extremely lightweight and efficient for the quality and speed of speech recognition it brings (two hotkeys and a script on top of whisper.cpp). Voluble was also recently added to the Joplin-relevant (text-to speech and accessibility) tools.

5

Thank you for those pointers. I'm currently on Windows because of my dependency on NaturallySpeaking and am planning on moving to Linux as soon as I can get a speech recognition alternative that lets me work as efficiently as the window/NaturallySpeaking combination. I'm sure your tool is nice and useful for many people. The only reason I would choose Aqua over your tool is that I'm spending way too much time in IKEA computer land. Always assembling, never using. When you make your living by the billable hour, spending 10 bucks a month on a package just works without having to do anything is a huge win.

After I recover from the disaster known as FuseBase, I will consider these other options, but I need to get my billable back up.