Show HN: SQLite Online – 11 years of solo development, 11K daily users

4 months ago (sqliteonline.com)

Your tool has been a huge help in the classroom over the past decade, thank you!

Having a quick online link to get students started is really useful from a student motivation standpoint. This allows them to get a sense of the query flow before having to set up their own database or moving on to other DBMSes.

Congratulations on 11 years!

Unsolicited suggestion: since some people are asking "what is this".

Maybe buy a new domain name like below (and direct to your existing from this new url).

  EasyLearnSQL.com
  TrySQLonline.com
  LearnOnlineSQL.com

All of these domain names are available for sale under $10.

And the more descriptive name might allow you to not have to change the UI.

(Very cool project by the way and congrats on 11 years)

  • In my opinion that's not good advice. Over the past 11 years this domain already built up some domain reputation and incoming links. Changing the domain for no good reason won't help with people understanding the use case of the site.

    The domain doesn't really matter so much as you can see with "replit.com", "chatgpt.com" or "stripe.com" which don't explain anything either.

    If you want to invest time I'd suggest:

    - Clean up design (Remove multiple disclaimers, side bar etc.)

    - Add h1/h2 that instantly explains what this is about

    - Have a list of simple examples that can be executed, not just "select * from demo"

    - If you want to increase traffic, take a look at "site:sqliteonline.com" on Google. There's currently only 14 pages indexed, so lots of low hanging fruits to optimize. Could also be extended by having pages dedicated to examples or a topic that people can land on if they search for things like "left join sqlite" etc.

    - Change site title from "SQL Online AiDE - Next gen SQL Editor | SQL Compiler" to something explaining what this is about.

    • i read the parent post's advice as duplicating, rather than changing - e.g., you keep the existing domain, but also the exact same app is available from those other domains.

      But you're not wrong - i dont think the domain has a big effect, and personally i'd rather save $10.

      2 replies →

  • Alternate opinion: these URLs read as "spammy" to me and I would be less likely to click on them.

    It would be apropos to at least have an "about" dialog somewhere, or maybe one of those "quick tour" popups.

Cool project. Congrats for keeping it up for so long!

Could you share some numbers like a ballpark of subscribers?

  • Thank you, unfortunately, almost zero.

    • hey, not to give you "armchair" advice, but I feel like a tool that's existed for 11 years and has 11k daily users is a super serious achievement.

      I'd vicariously love for you to be able to make some/more revenue with this!

      +1 on @redox99's comment that charging in rubles is most probably confusing, and that a flat $10 usd/month would be easier. I also would think that renewal should actually be on by default, not off - if people want the service and/or to support you, having auto renewal off is more of a hassle for them (the customers who want to pay you!) as they'd have to have to... re-enable their service? every 30-90 days?

      and another point I wanted to bring up is that it feels to me like a small text-based advertisement from ethicalads.io (the folks behind the ads on Read the Docs sites) or carbonads.net (btw I have no affiliation to either) could definitely... bring in some not-bad revenue pretty much immediately?

      again, huge congrats on your project and I truly wish you'll be able to find some path to monetization. cheers!

      11 replies →

    • It's weird you show the price in some currency I don't even know what it is (even if it says 10 dollars small next to it).

      Just show and charge 10 USD. You can localize that other currency (Rubles?) if GeoIP matches.

      6 replies →

    • Thanks for sharing. Still valuable for me in any interview if someone says they kept a service up for 11 years. Shows determination.

      1 reply →

It might be a good idea to not attempt an WebRTC connection right away. Since you're only using it for collaboration, there's no need to connect right away.

Privacy minded users might have it disabled by default since it's a good fingerprinting source.

Currently the website doesn't load if WebRTC is disabled.

I get this message:

The site takes a long time to load: is your internet slow?

If you have an old version of the browser update to the latest or use the latest version of chrome.

Close all tabs with the site and reload the last one.

support: z@sqliteonline.com

I am really struggling to figure out what this is or how it provides value.

Edit:

This discussion isn't exactly what I was hoping for. I was looking for ways to better highlight the features or value proposition of this site. Not defenestrate it altogether.

E.g.: A simple modal that says "Welcome to SQLite Online! You can <core value proposition> with this tool." would have radically altered my initial perception.

  • For me: Try queries without the hassle of setting up a database. Learn SQL. Experiment. Etc.

    • I don't understand the people trying to convince others that this tool is useless by saying "just do it this way, duh!". It is useful, even from a rapid glimpse at the website.

      Be kind and sensible.

      2 replies →

  • It's clear that the tool is highly useful to the people who use it.

    That being said, I feel like I'm dumped into the playground without understanding what I am playing with. A few short paragraphs, examples, screenshots, explanations, ect, would go very far.

  • > You can <core value proposition>

    It may be difficult to briefly describe all the website’s capabilities right now, but the key features include:

      * Federated queries across external and internal data sources.
    
      * Using query history as a source for new requests.
    
      * Collaborative access to databases — both server and local, with structure synchronization.
    
      * Automatic chart generation based on queries.
    
      * And much more, including hidden features that are not yet easy to summarize.

  • Agree with your edit saying that there should be a landing message that gives a quick overview. But with in a few moments I was able to see that you can create a database and then start inserting tables/records into it. Seems like a pretty good tool to learn how to create and manage a database without the hassle of having to download sqlite and start testing commands that might be new to you in the CLI

  • As an educator I would've loved to have this last time I was teaching SQL.

    1: No install

    2: Ephermal (just reload if you've messed up?)

    3: Good syntax highlighting

    4: Visual UI to navigate the model

    Why to pay for it though? That's a harder nut to crack, the UI is quite nice compared to many I've seen so maybe sell as an addon for those that provider hosted databases, collaborative spaces or as a desktop app. No obvious slam dunks though.

  • I would have loved this 25-years ago during university.

    Would have made homework (and just learning) significantly easier.

  • Yup, I think it would be a big help if the home page ('/') was a landing page explaining who it's for and why, and why. Is this for students? For prototyping? For quick analysis? Sample data? Importing real data? Use cases are key.

    And then have a big hero button leading to the the actual tool ('/app' or '/playground' or whatever). Maybe preloaded with different sample data depending on the use case.

    Right now, being dumped into a complicated interface with zero explanation is very confusing. (None of this is to criticize the project itself, just to help identify it to the people who might find the most value in it!)

I think the design not that bad but that green disclaimer appears on the first visit is kinda too long and nobody wants to read that in a mobile even in PC, so it might be cartoonized for the user to better understand it :)

Don't localise messages if you don't have anyone to proofread. Browsers have built in translation nowadays that users can activate if they need.

  • Could you please advise where the issue might be?

    • Not the parent, but I see that several messages related to buying a subscription are translated into the locale of my browser. In my language, it just feels a little amateurish. In other languages, perhaps it might contain something totally wrong.

      So it will be safer to just use languages you are comfortable with, like English and Russian. Especially on pages that concern money. :)

      1 reply →

  • Or use the latest llms for translation

    • The sufficiently-late LLM seems a bit like a true Scotsman. After your comment, the OP explained that they did, in fact, use an LLM for translation. No info about whether it was "the latest".

Thank you all! Special thanks to those who offered a fresh perspective and helped reveal the project’s challenges.

> The server is unavailable, contact your network administrator. Please try again in a few minutes.

Also where is it based ? And your payment processor ?

I like to support learning platforms like this financially, when they work of course, but that's going to be a problem if you're based in Russia.

I immediately tried to run .schema and a syntax error was returned.

  • I assume this is because sqliteonline is using the sqlite library, not running the literal sqlite command line application. Per item 7 in the faq (https://sqlite.org/faq.html), the way to get that info would be to run something like the following:

    SELECT * FROM sqlite_schema WHERE type='table' ORDER BY name;

    The website does seem to correctly return the names/schemas of whatever tables you've created if you run the command above, but the editor (incorrectly) adds red squigglies around the command, since I guess it doesn't realize this is allowed.

Amazing that SQLite Online has survived solo for 11 years. What technical or business pivots have kept it alive (and relevant) across changing web stacks and user expectations?

wow that's a lot of users, congratulations. It says sqliteonline but it seems you support other db's as well.

What is the WebRTC connection used for?

  • P2P “Share/Collaborate” mode: the UI text and toasts (“Share”, “Close connect”, “connected”, “No connected.”) plus e.rtc.user strongly suggest a feature where someone “hosts” a DB and others connect directly to run queries/see results live.

Great job and many kudos for the determination to maintain a tool for 11 years!

I thankfully have no use for the tool since I no longer have to code SQL - the world is a better place for it.

It raises the question how many more "bus tools"[1] are there? Tools maintained and developed by a single developer with whom, when hit by a bus, the tool would die.

[1] no offence meant but "bus developers" is the term I learnt, it seems a little cruel to speak of folks being hit by buses - is there something better nowadays?

I'd highly suggest getting a designer, or somehow thinking with more of a product mindset? I fail to understand what it does quickly, which shouldn't happen to a potential customer.

  • 11k daily users is very good even without this so-called "product mindset"!

    • The dev is asking on the site for people to support the development with subscriptions, but they say here they have basically zero subscribers. So 11k daily users hasn't translated to something that people want to actually pay to support. That could change.

  • No, please. Don't inflate the team to something that would need investors. It's cool that an alter-Web can exist without 10 person teams.

    • If you're making enough money, you can hire a designer every now and then to tell you where your UX suffers and how to fix it. No investors necessary.

      (or: just solicit feedback in a space frequented by designers, and harness the power of being wrong on the internet ;)

    • It is very straightforward to hire UX designer in a contract, or even just ask ChatGPT to design an interface that is better than a software engineer's minimum effort (and possibly experience) in UX.

  • And what currency is it in? Seems so odd to not put it in dollars or euros.

    And FURTHERMORE, the $ sign is incorrectly to the right of the numbers. It should be $10. Personally, this shows such a lack of product thinking, and simply hacking away at a tool instead of delivering a service.

    • I mean, it does clearly say it's 1,000₽ which is Russian rubles. Why would the price be listed in dollars if it's not being collected in dollars?

The paid subscription lists this feature:

> No auto-renewal

That's not subscription.

  • I aim to make the product convenient and as straightforward as possible. It's better to extend its use when needed than to worry about when to stop.

  • Sure it is. A subscription's defining trait is continuity of access contingent on periodic renewal, whether manual or automatic. People subscribed to things way before online payments or even credit cards were common. A modern, if niche, example is Tarsnap.com. Once in a while, I get an email from Colin telling me to pay up if I don't want my backups deleted.

A potential sales tip: Going down the list of the $10 plan the first thing I saw was the 300 scripts limit and thought "no way". Pay attention to how supabase does it, basically no limits except compute and storage.

Ok, you also have a popup in your website showing that off.

Cool, but, you're not breaking any Guinness world record.

There are solo founders behind many websites and tools and you might not even know that.