Obsidian Editor - a Hot Mess?

2022-01-11 11.48.11

Works on my machine. :wink:

2 Likes

As I was quoting the comments made @krocnyc for Excel, I meant in the Mac version of Excel.
Screen Recording 2022-01-11 at 21.13.24

Pretty much all my mac apps do this. There are a few exceptions and one has to decide if the benefits warrant the hassle. Apps that work against the muscle memory of well established workflows or have basic things missing or not working properly like spell checking run the risk of users using something else.

Iā€™m on a mac. It was a choice so I expect developers who want me to use their app to make it mac-like or I probably wonā€™t use it.

I appreciate that some people donā€™t get a choice from the workplace as to what software they want to use and have to make do, but where choice is given non native apps are always going to start from a disadvantage and the more they move away from expected OS behaviour the more theyā€™re going to annoy their users.

I understand where you are coming from.

And itā€™s okay for you to ā€œexpectā€ others to change to suit your preferences.

These are not my preferences. Every OS has its own ā€œpreferencesā€. Some prefer Windows, others Mac or Linux, ChromeOS or BSD. When they choose that OS they would expect the developers to respect those preferences. Not sure Windows users would appreciate the menu bar missing from individual apps and being stuck on the desktop or mac users being expected to use CTRL instead of CMD.

Developers creating non native apps need to take that into account.:grin:

Ah, apologies. Sympathies with that one, it bugs me to no end.

For sure, there are lines that shouldnā€™t be crossed, but I think thereā€™s still a grey area where non-native implementations can still work as good designs.

Non-native app with still-mostly-solid design and the features I need > Native app with great design and without the features I need. If I didnā€™t think that then Iā€™d just use all stock Apple apps. If all devs created native apps thereā€™d probably be a lot less featureful software out thereā€¦

2 Likes

I would totally agree with you and for that reason Obsidian is one of my daily drivers. However, had those guys owning Voodoopad got their act together, I would not be on Obsidian. Voodoopad did 10 years ago what Obsidian does today, but successive sales of the app to new owners that have done nothing with it has killed it. It even spits out a full wiki html site of all your notes, something many of us are eagerly awaiting to do with Obsidian.

1 Like

I thought this was fixed recently. You mean the autocorrect options from macOS System Preferences ā†’ Keyboard ā†’ Text? When I type the expansions Iā€™ve got there in Obsidian, they work.

Although I realize it doesnā€™t help the end-user, @svsmailus, this is (was?) an Electron issue. Obsidianā€™s developers have been pushing Electron to fix it. (e.g., Expose `spellcheck_platform::CheckSpelling` API Ā· Issue #22829 Ā· electron/electron Ā· GitHub)


Odd. The only folder the plugin should be able to read/write to is in Vault/.obsidian/plugins/Minimal Theme Settings. Uninstalling the plugin and deleting that folder shouldā€™ve fixed things. You may want to reach out to the plugin dev directly on their GitHub repo!

More broadly, there has been some chatter about better appearance management settings. The Style Settings plugin is great and shows that thereā€™s a better way. We might also see a ā€œsnippet storeā€ eventually, maybe!


I think parts of this critique are fair, but for what itā€™s worth, us mods have been discouraging people from @ing the Obsidian developers for as long as I can remember. If everyone is shouting, you canā€™t hear anyone. We therefore try to encourage folks to only reach out directly if thereā€™s something critically wrongā€”e.g., data loss.

My experience moderating the Obsidian forum and Discord has been interesting. Managing the thousands of feature requests, help questions, and bug reports that flood in has been arduous at times. And keep in mind that we are only doing it for our love of the appā€”we are not Obsidian staff, just a small team of volunteers.

Iā€™d like to think weā€™re best-in-class, though: I donā€™t know of any other app that provides a similar level of access and transparency in the development process. If anyone knows of one, Iā€™d love to steal their ideas. Nonetheless, it is naturally frustrating when a feature you desire is not coming to fruition, or when bugs arenā€™t getting squashed. One thing Iā€™ve learned from watching Obsidian and other apps develop over the past year, however, is that there arenā€™t easy solutions to these kinds of problems. As youā€™ve said, @anon41602260, success can kill. Adding developers to teams rarely speeds things upā€¦ In fact, every indie app that ā€œgoes mainstreamā€ and starts to add HR seems to slow down. Moreover customer service teams become opaque dead-ends. I personally hate hearing ā€œokay we heard you thanks for the feature request.ā€ So, while there are challenges, Iā€™d like to think weā€™re doing the best anyone can.

9 Likes

I think you all have done an admiral job moderating the forums. Cheers!

Thanks. I want to be clear, though, Iā€™m not trying to be defensive here. @anon41602260ā€™s critiques are valid, and Iā€™ve picked up on similar issues over time. I just donā€™t know a better way to run it (yet)!

2 Likes

Thank you @ryanjamurphy ā€“ itā€™s useful to hear the view from the moderatorsā€™ dugout. I think the best things about Obsidian are also the worst things. From the beginning, Erica and Shida created an open, communicative environment for the Obsidian community. But then the app shot up the s-curve and adopters flooded in ā€“ many of whom migrated from Roamcult. So balancing the work with chatter from the peanut gallery probably became, and remains, a difficulty. I think your post hints at that. The only answer to that kind of success is to eventually put up walls and limit communication.

The main good/bad things though are the open API and appearance structures: resulting in the community plugins and themes. The good part is that we can configure Obsidian to an amazing extent. The bad thing is that as the users flooded in, most of them probably have little understanding of managing the complexity created by the the community add-ons. From watching the two OBS forums, I suspect that most of the issues raised are issues with third-party plugins or themes. My own issues are issues with third party themes. For the most part, core Obsidian, what Erica and Shida control, is pretty solid and reliable. The outer shell, created by community contributions, is often very flakey and conflicts arise. This will only get worse. Unless someone buys the company, or abandons the ā€œalways freeā€ philosophy, Obsidian might always lack the resources to tame the contributor cloud.

I donā€™t think itā€™s too dramatic to think that Obsidian, the product with all the community around it, is at its final inflection point. The thread above sort of makes the case: Obsidian is wonderful and messy. The people who want to just work and not put up with messy will stop coming. I donā€™t think OBS will flame out, but I do think the signs point to flat growth and the small part of the market that even cares about OBS getting bored and moving on. The s-curve is probably over.

BTW there are plenty of independent developers that interact openly and effectively with their user base, discussing work in progress, without a gaggle of gate keepers around them. George Browning, Mark Bernstein, Peter Lewis, Greg Pierce, are just a few of them. And they all manage product that are significantly more complex than Obsidian. It can be done.

4 Likes

Weā€™ll see! I have not seen any signs of growth slowing. I definitely think weā€™re seeing some interesting thresholds tested, though:

  1. Will developers of key plugins and themes stick around and keep maintaining, will new devs take over old code, or will community projects languish over time? Thatā€™s the key indicator of health, for me, and I am not worried at the moment.

  2. Will the community support model continue to work? Hereā€™s where Iā€™m seeing stressors. The forum seems less useful over time, and itā€™s hard to track how useful the Discord is. That said, I donā€™t know of a tech companyā€”especially an indie developerā€”that provides good 1:1 support at scale (see below). After a certain point of growth, it seems impossible to get real answers from companies. I think the Omni Groupā€™s model is pretty good and probably one to follow here.

However, as has been discussed in this thread, a key constraint is the complexity of Obsidian. As youā€™ve said, the core app is rock solid. Most challenges users face and seek support for are driven by community add-ons. Most developers donā€™t have to support community complexity like this. The tools that have are usually developer-centric (VS Code, Sublime Text, etc.), and so the users can support themselves. So, I think weā€™re seeing new territory here. What will users tolerate? Time will tell.

These are good pointersā€¦ but the scale is quite different.

I got curious about the volume and velocity of these communities, and the numbers are pretty striking. I added DEVONthink and The Omni Group to the list of comparables. (The stats for Zengobi/Curio and Eastgate/Tinderbox are hidden, so I have excluded them from this analysis, but I imagine they are not far off these numbers.)

Drafts has been around since 2012ish, Keyboard Maestro since 1990 (!), Omni since 1989 (!!), and DEVONthink since about 2001. The communities Iā€™m comparing here have probably been launched later than these dates, but each is still older than Obsidian IIRC.

The Drafts forum has 4677 users. The Keyboard Maestro forum has 7624 users. The DEVONthink forum has 13,504 users. The Omni Group forum has 11597 users.

In the past 30 days, the Drafts forum saw 31 new topics and 436 posts, Keyboard Maestro saw 253 new topics and ~3100 new posts, Omni saw 45 new topics and 784 posts, and DEVONthink saw 279 new topics and ~3600 posts.

Obsidian was launched around 16 months ago. The forum has 25,556 users and the Discord has 49,557 users. In the past 30 days, users have posted 786 new topics and ~5700 new posts on the forum. I donā€™t want to even look into the Discord, because that place is pure chaos.

Itā€™d be neat to look into per-user contribution and so on, but I only have those stats for Obsi. From what I can see, Keyboard Maestroā€™s per-user activity is wild: many members have published thousands of posts. Omni has a Slack team thatā€™s pretty active, too, but Iā€™m not in it at the moment. Otherwise I think these communities are comparable in nature, but not in scale.

Hereā€™s the numbers in a table:

Age (years) Number of forum users Number of topics in the last 30 days Number of posts in the last 30 days
Drafts ~10 4677 31 436
Keyboard Maestro ~30?! 7624 253 3.1k
DEVONthink ~20 13504 279 3.6k
Omni Group ~30?! 11597 (+ ? on Slack) 45 784
Obsidian <2 25556 (+ 49,557 on Discord) 786 5.7k

Obviously, quantity is not qualityā€”I bet Keyboard Maestroā€™s community would show that having fewer, more engaged users is probably better for those who are involved.

Another consideration is that of all of these devs, only Obsidian supports platforms beyond Apple. That makes support harder. (It would be cool to look at some other cross-platform indie devs, but I canā€™t think of any!)

I donā€™t really have any big conclusions to draw, except to show that Obsidianā€™s community growth has been meteoric. My question now is what the community will look like when itā€™s as old as these others.

11 Likes

Interesting analysis. Does it include the Chinese forum?

As you point out, Obsidian ā€“ the company through its public forum ā€“ is basically supporting the work of third party developers of plugins and themes, who may or may not be interested or active in supporting their own work.

Also, I think all the other apps we mentioned have off-line support offers ā€“ not that any of them have big staffs for that kind of work, but they do take an unknown number of requests by direct contact with ā€œsupport@ā€¦ā€. Whereas, OBS does everything in the forums.

Something will need to change ā€“ the support numbers you mention are wild and unsustainable.

1 Like

This is really helpful Ryan and I sympathize with how difficult it must be to support all of this. But I think my original post is still valid, to me Obsidian is a subpar markdown editor. Do you not agree with that?

Today I did all of my editing in Obsidian just to force myself to really use it. I think everything I posted in the original still holds. I do have to say the live preview mode works pretty well, so thatā€™s an improvement. I would just encourage the developers to make the Markdown editor best in class.

1 Like

It strikes me that the comparison for Obsidian might be Wordpress: very usable, solid core app; huge and untamed/untamable ecosystem of themes and plugins.

Wordpress has survived and even thrived. I do t know the ecosystem well enough to say whether thatā€™s because of this model or in spite of it. But I have no doubt there are lessons for Obsidian one way or the other.

2 Likes

I find Obsidian is a first-rate Markdown editor. I just reviewed the original post in this topic. Keyboard shortcuts for formatting? I donā€™t need them; formatting by characters, the Markdown way, is easier for me. Errors in header sizes are not a problem Iā€™ve encountered.

Obsidian is not for everyone, and neither is Markdown.

4 Likes

Sure, today as I was using it, I did just type in the characters for the headings and that works fine. But when you get used to using Cmd-1, Cmd-2, etc, it is a bit slower.

Just now I was using Obsidian to collect a bunch of weblinks as I was researching something. I know I can type in the 4 link characters by hand, but I find using Cmd-k faster and easier (which is supported in Obsidian), donā€™t you?

No big deal, but I would like an app I use for hours a day to have a good number of keyboard shortcuts to help with my productivity. This is the Mac Power Users forum, right?

The thing is, you can add those, no?

Not saying your opinion is wrong; I am not completely satisfied with Obsidianā€™s editor either. Some things about Vim support are a little flaky and having a way to temporarily ā€œmaximizeā€ a pane would be nice. But personally, I think the #1 selling point of Obsidian by a mile is the organizational features, so it kind of makes sense that the devs would focus substantially more energy on that, and satisfy those who want better Markdown experiences with their interoperable filesystem structure.

3 Likes

100% agree nobody is going to mistake Obsidian for a Mac-assed app.

Worth noting though that the Things Theme is a big improvement in that direction compared to Obsidianā€™s default theme & most of the other third party themes