Thanks for the related links!
It just shows what an amazing job to whomever leads Excalidraw does. They should be the poster child of PLG GTM done right. Fast time to hello world / time to value, doing the small things that make you not just use the product, but love the product. The difference between an MVP and an MLP. It's interesting how it's so rare that it makes so many of us praise them for no apparent reason other than spread the joy. p.s. that issue with worrying about pixel perfect alignment and having them do it "sloppy" so you don't worry about it, I wonder if that was intentional, if it was, genius move, as I think it's a fair assumption that wasting time on making things pixel perfect is something a lot of us suffer from...
It looks like they took the "sloppiness" off the handwriting mode. For me that was always the biggest turn off—I couldn't get over that I was penciling in some hand-written note, and for some reason there were two lines following my cursor, spiraling like twin dragons. It was so bad that my first time using the product I thought it was a bug or that I was using the wrong tool. The handwriting now is just a single line!
Now my only other complaint with excalidraw is that for some reason on my iPad Pro in Safari, the handwriting is incredibly laggy, increasing the longer the stroke is. Which is weird, because in every other browser I've tried (Safari/Chrome/Firefox on various laptops) it's incredibly snappy.
Glad to see things are improving! That's a sure sign of a company that listens to users.
Yeah sorry about the initial version of handwriting, we thought we could do a quick hack and use the multi point line rendering for it but it didn’t work well in practice.
We switched to the awesome “perfect” library from Steve Ruiz instead. Could you create an issue about the lag on ipad pro, we should investigate and fix.
I started playing with containerizing Excalidraw but noticed that there were a bunch of security issues identified by Dependabot.
Given that you already have Dependabot active, would you consider getting the dependencies somewhat up to date? Looks like the pull requests already exist... Thanks.
Dependabot is something like a modern plague of open source, together with the bot that closes issues automatically.
Did you read through the security issues identified, familiar with the codebase enough that you can see that these issues are actually affecting the product?
Most issues Dependabot opens on repositories I'm involved in, points to issues that has no bearing on the actual code, since it's security issues that are involving passing user code to specific functions, or simply regarding APIs that are never used in the first place.
You don't have to be on the newest version of every single library, in most cases it doesn't make any sense to just upgrade for the sake of upgrade.
This. You are just swapping bugs that are known (and that in 99% of the cases do not affect you) with bugs that are not yet known (because of being always on tip release), for the sake of not willing to spend time checking if you actually need to upgrade. It does not correspond to higher security but rather a russian roulette game where you swap bullets with newer variants, all the time.
fyi, i have a different problem where some strokes are missing. when writing, after one stroke, the next one isn't detected, then the magnifier appears, then the next stroke is detected. Admittedly, i'm trying to write like on paper.
The upside IMO is that the sloppiness takes away a lot of the pressure to make things look perfect, align everything, make every box exactly as many pixels wide. I found I could pump out diagrams much faster.
I am not convinced that this is better ('hand drawn diagrams'). We've been drawing diagrams of all kinds for a very long time with no hurdles whatsoever [1]. Millions of diagrams have been drawn by boring-big-corp orgs in PowerPoint. This just reads like a hipster fetish over aesthetics, not substance. And if we're talking about aesthetics, I much prefer Monodraw [2]. I have a bias for having appeal for boring things may be because it works and gets out of the way. Emojis in your readme? No thanks.
That said, the app itself has great UX, especially onboarding. Just straight to drawing mode. Kind of like autodraw.io.
I find it better, in the same way that finding the PowerPoint ‘theme’ that makes everything a bit rough round the edges was also a revelation.
The perfectionist in me won’t let me publish a regular, square-edged diagram unless it’s pixel-perfect. As a result, nothing ever gets published.
Take away that absurd situation, make the whole thing a bit shabby, and the need for perfection disappears. In fact the ability to make it perfect disappears.
The article does not seem to actually mention why Excalidraw is so better over the millions of diagram drawing programs already available for decades (many of them much more featureful). It just seems to mention why Excalidraw is better over your average website.
What I dislike about Excalidraw is the handwriting font. While it might look cute it hurts readability, which is something which I believe shouldn't be a compromise for stuff like technical documents.
All 3 of them are terrible! The handwritten one is really bad, I'd actually prefer Comic Sans. The fixed width font looks very inhomogeneous. The sans-serif is okay, but the serious "a"s clash with the playful style of the drawings.
Sometimes it's not about looking cute, but to signal the 'sketch' phase from a design. When a sketch looks too professional there's going to be a disappointed pm or customer because it looks like it's almost finished when it's not even implemented.
Excalidraw might be good for it's intended purpose but I was introduced to it as an online interview tool and for that it completely sucked.
The most important missing feature at the time was that to doesn't show other user's selections unlike many other collaboration tools. That means what you see is not what you're collaborators see. If you grab some shapes they don't see that you grabbed them. If you highlight a word or item in a list, they don't see you highlighted it.
Further, unlike pretty much all similar software, they don't connect shapes and text. Place a shape, double click, a text box appears in the center of the shape. Drag the shape, the text doesn't go with it. This makes it super inefficient for most common use cases. People might want to separate the text from the shape but I'd argue that's an exception and should not be the default.
> the most important missing feature at the time was that to doesn't show other user's selections unlike many other collaboration tools. That means what you see is not what you're collaborators see. If you grab some shapes they don't see that you grabbed them.
This should work out of the box if you're collaborating. But, we've recently upgraded our collaboration servers and due to ServiceWorker caching it may be your colleague (or you) were connecting to the old one. Try hard refresh your browser.
It's actually a pretty good example of a PWA which can be installed (depending on whether your platform supports it, works well with chrome and windows, sorta works on iOS but more fiddly. Havent tried firefox).
It can be, but I was mostly trying to list a benefit that the "good old" apps didn't have, i.e. not needing to install it.
That said, I'd also misunderstood the grandparent comment: they meant that it's a shame that Excalidraw is one of the few apps nowadays that have these features, even if the other ones also have advantages like not requiring an install.
"Hey, I know you are super busy but I just wanted to check..."
Makes me want to ditch the product on the spot. Yes, I've seen your emails. Yes, I will let you know if I need anything. No, I don't need a zoom call/onboarding video/introductory convo with your marketing team.
Ditching the product is probably the goal of the email, if the stats show that paid adoption happens within x days 98% of the time.
What I think those stats miss is that I will test a product with one email and if I decide to buy for a team will use a new account. So it looks like the conversion happened immediately skewing their sales/marketing strategy.
First of all it's a free, open tool, MIT license, that is also hosted for free.
It loads instantly without an installation step. You just type in the URL and it works.
It's not a "productivity program" specifically as illustrated in the article.
The design is very simple, intuitive and deliberately minimalist, while it does quite a bunch of subtle things under the hood that harmonize with the overall goal of the program.
---
Now there may have been tools kind of like this in the 90's. Not aware of any of them from the top of my head.
I assume kazinator means: In Office 95, you click the Excel icon and it opens to a blank document with no login or registration or anything like that. As did drawing software, 3D modelling software, and so on.
Of course, arguably one did have to pay for it and install it - but that was a one-off thing.
People might be seeing this through slightly rose-tinted glasses.
Yes, you could use Excel 95 by just double-clicking on an icon.
However, not only did you have to buy and install the software, which back in the day often meant going to an actual brick-and-mortar store to purchase a shrink-wrapped box labelled "Office 95", but it even required buying a specific type of computer with a specific operating system first.
In this specific case, that doesn't even take into account Microsoft's shady tactics of strong-arming vendors and OEMs into exclusively shipping Microsoft software with their products.
So, arguably, the barrier to entry still was a lot higher than with even the most obnoxious growth hacking tactics we see in many SaaS products today.
That said, I absolutely love the way Excalidraw are "onboarding" users.
It's just worth remembering that not everything went downhill with SaaS products (in fact, quite the opposite) and not everything was better "back in the day".
Dear god - anyone here have to setup licensing servers for high-end 3D software from the past, like Softimage or Alias|Wavefront - those were a bitch, and if you lost connection to the licensing server, you couldnt run the program.
Parent post was not criticizing Excalidraw, but rather highlighting problems with other apps that are so prevalent today. They are saying that Excalidraw is great in comparison to apps that have those bad features, in other words "this is like the good old days".
And what were those tools in the good ole days that had the features enumerated by the parent? Primarily free, open source, hosted, instant open, local storage along with the usability?
Excalidraw feels like a product of this time, and demonstrates some of the things we all enjoy about the web as an ecosystem. People compare it to some yesteryear product because of its purposeful feature set and look / feel, but name something from back then.
Nothing is being contrasted in my comment. The software in question has the above features, and it has "no onboarding. No signup. No confirmation email. No OAuth."
It's just sad that you even have to have that in your checklist of things nowadays. Like it's amazing when software doesn't ask you to make another damned online account and just lets you use it. (And it's good that this doesn't, yay.)
Hosted is a negative feature to me (other than for trying the software). You don't control the software installation.
Tomorrow you could wake up and try to continue your work, to discover that some feature is missing you were depending on, or new bugs are happening that weren't there and so on. Or your local storage is borked because it's not being read correctly by the newly installed code.
It's good if something is hosted for trying out, but you want it to be locally installable so you and your team can depend on it, and evaluate updates before rolling them out, and roll back bad updates.
"Has local storage" is another thing that is sad to see as having to tick off a checklist.
You mean that the companies making insane margins and paying top compensation are exactly the ones responsible for most of this this since they're massively profiting from monetizing consumers through various dark patterns, ads and spying?
The first "onboarding" I can remember, from Microsoft Word 6.0 in (?) 1993.
It would display a dialog when you opened the program. Every. Damn. Time. Displayed a useful tip from the included documentation.
At the time, it felt like a huge, complicated, OLE application, a big departure from the previous version, Word 5.1.
I suspected that some members of the Word 6 product team felt the same about that startup dialog: at a loss to suggest anything productive for that "Useful product tip for today", they wrote down some inane safety tips.
"Welcome to Microsoft Word...
You can hurt yourself if you run with scissors."
I sympathose with the author, bit I shouldn't have to scroll past that to find out what does the tool he is raving about actually do. This stuff can go below that
I originally was going to sarcastically point out that this just means it's at the early stages of the process, when it needs to attract the power users and early adopters. Once they're on board they can start to annoy everyone and take advantage of the lock-in to attract more users while preventing current users from leaving.
I then read that it was free and open, so felt bad, but did notice an ad for Excalidraw+, which basically means my cynical first take is likely to come true eventually, as that is what the market will optimize towards.
OK it's pre-200x and you read about a new productivity app you want to try. You have to check it's compatible with your OS, figure out where and how you can buy it, wait for the diskettes or CD to ship or until you can visit a B&M store, run the installer, then manually enter a product key (fun!).
You'd be lucky to have it all done in less than a week and a few hours of your time. How soon they forget!
You still have to today. A lot of stuff is Mac/Windows/Linux exclusive. Sure, I can run Windows stuff on Linux and vice versa if I'm tech savvy, but Mac stuff requires buying Apple hardware, so that's a massive barrier right there.
>figure out where and how you can buy it
Or you could just download it. The internet existed pre 200x as well.
> then manually enter a product key (fun!)
Honestly, it was fun. Much more fun than the phone-home, always-on telemetry and invasive DRM that act like rootkits & spyware we have today. Good luck playing older games you bought and paid for who's DRM require authentication from the publisher's servers that went offline years ago. Yeah, fun!
Back then, you paid for the product, you got your license and CD key, you owned it forever. No subscription rent seeking model BS.
Outside of free software/shareware, there wasn't a lot of software available for download in the 90s. And compatibility was a lot jankier in that the base systems sold to consumers were insufficient to run a lot of popular software. For folks like you or me, probably not a problem (unless you happened to have a graphics card that wasn't supported) since we'd generally have obscene (by the standards of the day) amounts of hard disk and RAM, but Mom's computer might have 4M of RAM and a 10G hard drive.
Back in the floppy disk era the thing that always drove me nuts was that most (with very few exceptions) software vendors put the software key on the label of disk 1. Which would be in the drive when they asked you to enter the software key. I remember some OS/2 software that put it on disk 2 instead and I wanted to go to the company and kiss whoever it was who decided to do that.
It wasn't as onerous as you make it seem. Certainly that last bit - entering the product key - is way better than what we have today. I can save the key in a text file alongside the installer and the software can be reinstalled completely offline whenever I want.
I still have my Windows 98 product key and installer stashed away, purchased decades ago. And I actually used them a few months ago to stand up a VM to utilize an ancient printer.
Way better for me-the-consumer than the subscription-everything model. When software decides to go that way, I usually find a FOSS alternative or cling to the last standalone version.
I agree with the annoyance of so many modern tools and all the crap to use them, and appreciate ExcaliDraw for not doing it, but what is this talk of productivity programs being some sort of nirvana before 200x?
Should you choose to share a drawing, it would be end-to-end encrypted and sent to the Excalidraw server. The sharing URL contains the ID of your drawing for retrieval, and the encryption/decryption key in the URL hash, which never reaches Excalidraw's servers (thus achieving E2EE).
> deleting all of a website’s script-writable storage after seven days of Safari use without user interaction on the site.
So as long as you access the application once a week, you're safe. Or, alternatively:
> As mentioned, the seven-day cap on script-writable storage is gated on “after seven days of Safari use without user interaction on the site.” That is the case in Safari. Web applications added to the home screen are not part of Safari and thus have their own counter of days of use. Their days of use will match actual use of the web application which resets the timer. We do not expect the first-party in such a web application to have its website data deleted.
Still, horrible idea and horrible implementation, but I think it's known by know that Apple/Safari will always be hostile towards web apps as it cuts into their own App Store revenue.
Your drawing is saved when you leave the page. Dialog boxes are for destructive actions. There is no demo mode - what you're using is the full product.
Totally agree with this. I discovered Excalidraw by chance a couple of years back, and I couldn't believe what I saw. Free (as in beer and as in speech), instantly usable, useful features coming out over time. I converted several people in my company to use it, and to date use it to explain ideas in shared sessions (I have a visual thinking and this really helps me getting concepts through to other people). My secret dream is to be able to build a bunch of tools like this (not for diagramming, for different things) that people love to use.
You mean the times when a developer only had to provide a single version for a single OS instead of what we have today?
Excalidraw can be used on basically everything that has a compatible browser: Windows, almost all Linux distros, macOS, Android and iOS, major BSDs ... and that's not even counting in x86 vs x86_64 vs all the different versions of ARM machines we have now. Try to reach all those people without a tech stack based on web. Sure, you can still get desktop and mobile apps by cramming everything into electron or its mobile counterparts, but then we are back to comments like "I miss the days when desktop apps were less than 100MB"?
I'm agreeing with your overall point, just wanted to share that APE (αcτµαlly pδrταblε εxεcµταblε)[1] actually can create a single binary that runs on Linux + Windows + Mac + FreeBSD + OpenBSD + NetBSD, truly impressive.
That still only gives you the C sdtlib. This is miles away from building a desktop app with a graphical interface. You'd need to add something like SDL on top.
redbean has a https://redbean.dev/#LaunchBrowser function that lets you utilize the highest quality cross-platform gui that's already installed on the user's local machine. Think of it like Electron except with a 1mb install size rather than 100mb. I'm the author and I can tell you cosmo and redbean are ready to meet the needs of many use cases, and if they aren't then I want to know about it.
A lot of people use things like Excalidraw also on their tablets. (mostly iOS and Android) How is this helping?
This whole cosmopolitan-fanboy'ism is completely ignoring my initial statement. What are you trying to bring to the table here? Sorry if I sound harsh, but you are really just proving my point here while probably not even realizing it.
> That only solves 10% of the problem, binary portability. You still need to code for a dozen different graphics, I/O, storage etc system APIs.
Read https://justine.lol/ape.html for a deep dive into how we managed to do it. It mostly just boils down to the numbers file mentioned in the last paragraph. That gets you far enough to do sockets and stdio. That in itself is a big step forward. Is it good enough to make apps like Excelidraw binary portable? Not yet. But it gets us much closer than we were before. Getting the rest of the way there is an ongoing topic of engineering and experimentation. Many people are helping. Further support is welcome from those who share a genuine interest in contributing to our efforts to solve the binary portability problem.
And it can disappear tomorrow if they developers so choose. Maybe instead it would be better to say I miss the days when I had some control over my apps.
Of course I know excalidraw is open source, so someone could just fork it if the developers gave up on it. But this is only true with open source web apps.
There's no onboarding. No signup. No confirmation email. No OAuth. You're just in the product. Are there any flashy overlay tours with annoying animations and pulsing buttons? Is there a modal upselling you on a paid plan[1]? No. You're just in product, able to use it.”
Get rid of your fancy growth marketing people, your advisors with their “request a demo”.
Get rid of your intro page talking features.
Dump your stupid signup before starting.
Just. Fucking. Let. People. Instantly. Use. Your. Full. Product.
Instantly. Not after they give their email. Not after they click three dialogs. Not after they jump through whatever remaining hoops you’d really really love them to jump through before.
Show them, indeed GIVE them your product without the slightest fucking delay.
All of it. Stop arguing about why you’d really love to make them do something else before using your product. Make your marketing person stop insisting the marketing thing gets done first.
Get people into your actual full product within the shortest possible time. Milliseconds if possible.
- Does it actually lead to higher sales? If not, nobody is going to care about our wish for more companies to do this.
- It doesn’t work for many products. How do you just drop someone into a cloud storage product? You could give them some temporary space where they can drop some files if they want, but what do you do if they abandon them? Delete them after some time? Might confuse some customers who didn’t realize they had to actually register an account. Maybe a huge banner with a countdown would work… you start to have to solve a whole lot of problems to make it work
But so many products that are well suited to immediate usage are hidden behind layers of explanation and marketing and signup and other irrelevant barriers.
Says the person with free time to write, freedom to express an opinion, food in their belly, shelter, hot water, a computer in their pocket, access to a free vaccine, yada yada yada.
What the parent comment describes is what people who aren't capable of material or intellectual contribution do to products to create the illusion of value where none exists, or to manipulate users into paying as much as possible regardless of the effective value of the product.
This stifles development of things that are actually useful. It puts money and influence in the hands of people who shouldn't have either, creating bad incentives in the market and within the company. It gives marketing a gloss of false legitimacy, allowing for all the things that have us in a feverish race to the bottom.
Everything you just touted has been achieved despite the Idiocracy fan-fic that seems to be the current mission statement of corporate America.
There are many categories to consider here. Three examples:
- Products that don't require any of this stuff.
- Products that require some form of registration since it stores user data remotely. The person registers so they may secure, modify, or remove that data. The handling of personal data should serve this role exclusively.
- Products that require some form of initial setup and onboarding will facilitate that setup for the vast majority of people. The user should still be able to bypass the onboarding in case their use case doesn't reflect the common use case.
When marketing becomes involved, things become problematic. That friendly offer of assistance to get started usually reflects marketable features rather than any form of genuine training. That collection of personal information is used to harass the person using the product rather than to facilitate the use of the product. (Heck, I have companies contacting me over a decade after I last used their product to try to make a sale.)
> Does it actually lead to higher sales? If not, nobody is going to care about our wish for more companies to do this.
I think this is a really, really interesting phenomena: the human brain (perhaps when sufficiently institutionalized) exhibits some sort of fundamental bias toward disruptive UX that breaks flow and productivity, and this approach must be proven wrong, every time, on a case by case basis, and doesn't fundamentally internalize.
It's crazy: because it doesn't lead to higher sales, it's seen as a bad approach. WHY? "Just because."
> It's crazy: because it doesn't lead to higher sales, it's seen as a bad approach. WHY? "Just because."
If the goal of a company is to create higher sales, then it is a bad approach.
If the goal is to create a great tool used by people and not worry about adoption or supporting a large business through profit, it's probably a great idea!
> Just. Fucking. Let. People. Instantly. Use. Your. Full. Product.
I love that - and that's how I aim to build things now.
Excalidraw has few calls to external domains as well - just Google Analytics and Sentry.io. That's still two that can be reduced, but it's much better than most modern clustertrucks.
Yeah, we’ve been trying to keep it without calling to any external service. Both of them are optional, excalidraw will work offline and/or with an adblocker that blocks those domains.
That said, those two have been really important for the success of the project. Google Analytics is helpful to get a sense of how important it is to people and motivation for the team. The second one helped us fix a lot of edge cases that couldn’t immediately be found via normal testing.
This was an important design goal for Webhook.site, which I'm the founder of. You're instantly in the app, you get the unique URL and you're ready to do. No need to click anywhere. By now I've lost count of how many people have written to me saying they appreciate this aspect of the product.
This is great advice if your goal is to get the maximal number of people using your product.
But it's not exactly clear how you plan to turn that into, say, revenue. Build it and they will come, sure. And the more leads you stuff into the top of the conversion funnel, the more conversions, right? What do you mean 'qualified' leads?
If your product's viability will rest on turning some percentage of those free users into subscribers or purchasers, then you have to do some upsell, at some point. And if they can get into the full product without any obstacles within milliseconds, you're going to have a very, very hard job getting those conversions. Meanwhile you're winning over free users who cost you money and time to serve in handling their support and feature requests, and you're bringing in ever more people who are never going to give you any money.
It is viable to build your product this way if the primary product your business sells is 'convincing investors that by building up loyal active users, and developing rich understanding of those users' behavior, you will somehow create a reservoir of value that can be sold for 10x the amount of money you're asking them to put in to keep your servers turned on for another few months'
And it's probably viable if you aren't trying to be a business at all.
But as generic 'this is how all products should work' advice, it's pretty terrible, unless your main advice is 'don't try to make a business around offering a software product unless you're primarily interested in selling engagement monetization fairytales to investors'.
You monetize when the user has value that they do not want to let go.
Ready to save your new work of art?
Or higher resolution?
Or premium features?
You do it when the user is hooked.
I was not advocating for “no revenue”.
Also it seems self evident that such advice does not apply to all software…. Do you think I’m saying this is how SAP should be sold? No…, obviously the advice applies to software suited to this approach.
‘Here’s the whole product with no tutorial or marketing crap, off you go, have fun’
‘Thanks! Oh wow, this is great!’
‘Glad you like it.’
‘I love how you don’t try to upsell me on things or push any kind of premium paid features, just give me the whole product without limitations’
‘Yeah, that’s our philosophy, let you fall in love with the product’
‘Cool, thanks. Okay, I’m gonna save my work now.’
‘$20, please’
‘I’m sorry what?’
‘$20 to save.’
‘Er… could you have mentioned this earlier? I just spent ages using your awesome product and you just let me in and I didn’t have to sign up or anything, and you never mentioned a premium upsell, so I just assumed…’
‘Yeah, I thought I might be able to trick you like that. We call it “bait and switch”. Isn’t it so much better than all that marketing BS?’
I agree, not allowing to save would be pretty asinine. But having collaboration or limiting the number of documents or the complexity of documents is totally acceptable.
It's not advice for _all_ new products. Excalidraw has a plus plan, and I'd be curious to learn if they're seeing success. I think if you make something highly highly accessible and lovable, there will be opportunities to make money. Not sure! It's just a core belief of mine. I'm also sure there are products like that that can never figure out the pivot to $.
The plus plan has been very successful from my point of view, we’ve been able to bootstrap the business and make our first hire.
The key realization is that while the unauthenticated open source model works really well for individuals, it’s hard for companies to adopt.
They want authentication, security policies, private spaces to collaborate, access control, search… all of this requires having a company behind it in practice and the open source model doesn’t work as well.
The two versions so far work really well alongside each other right now.
That's how the newspaper business approached the internet and it nearly killed them all. You can't do this in a competitive field where operating your platform is expensive. Excalidraw is probably run by a few people not trying to get rich and able to iterate on code at their own pace. That is an ideal situation to keep it simple. It just doesn't work for most businesses.
I feel like this is the benefit of a web app, you can just simply open it without having to go through download process etc. Just being able to open a link and get started or try it out is the best part.
I thought so too, so I actually tried this. With a whiteboard app. And it didn’t exactly work, new users and engagement went down compared to having a “landing page” in front.
Only a relatively few people like you and me really want to dive in without having a story first. Turns out, to my surprise (!) that most people don’t want to dive into your app without reading about it or watching a video first.
Teachers, kids, business users, they all have different questions they want answered before they spend time trying to learn a new app, they want to know if there are reasons this can or can’t be the thing they use from now on. Learning a UI is a time investment, and people want to make sure it will be a good investment before using their time. They don’t want to be tricked into learning your UI for a week or two or more and then find out later it can’t group and copy-paste shapes, or that the price goes up, or that they can’t share they way they need, or that your app doesn’t have a on-premise installation.
Getting in quickly can have a low barrier to entry, I’m with you there, but my experience on the other side of this is that it’s not good to go directly to the app without an intro first, generally speaking.
Aside from this there’s also the fact that users who don’t intend to pay are on average way more demanding, way less patient, and way less understanding about communicating and learning the UI and dealing with bugs or misunderstandings. It was surprising for me to see from the other side how rude the free users are on average compared to the paying users.
> It was surprising for me to see from the other side how rude the free users are on average compared to the paying users.
In my experience, this also applies to cheap vs expensive projects (consultancy). Customers start to get really pleasant and polite when they are charged at least 100K or something like that for doing some custom software development. They get very demanding when it's below 10K.
Might be because you're talking to different sets of customers. People who are cheap vs people who know what they want is complex and are willing to pay for it.
I agree as a user. As a developer and businessperson though this sounds incredibly entitled - "make great software and give it to me instantly for free. I should get to keep using it and you should get nothing."
Shouldn't it be win/win? Both sides giving something to help each other? Software gets tricky I guess since it's fun to develop, and mere attention or praise is enough for many devs. And ads blur the line of "free". But the level of childish entitlement gets a little depressing sometimes.
You need customers before you can make money. It makes sense to me that you first create a user base before you try to upsell them. If the goal is to create a user base, the fewer hurdles the better.
I don’t see it as having to do with entitlement, more as a case of it being easier to standout in a crowd of options when your thing is easier to start with.
I genuinely can't stand the default font, and so many people tend to use it. My dyslexia makes it really difficult for me to read it. Please, please, Excalidraw owners if you're reading this: make the default font something that's much more friendly to those with dyslexia.
As someone without dyslexia, I find that really difficult to read! I'm all for accessibility, but I personally wouldn't want this to be the default - would be great if users with dyslexia had the option to set it as their default though.
Oh sorry I wasn't clear, I get these fonts for my coworker(s) who have accessibility issues, so I just listen to what they say. This open one was their current preference.
To be honest, I don't believe we should adapt defaults to fit a single digit percentage population. I'm not dyslexic and the fonts suggested are horrible to read in my opinion. I'm totally for the option to set different fonts as default so dyslexic users have an easier life, but the global default should not be changed because of this.
There is an important school of design thought that you should design for the people with the biggest handicaps, and by doing that you make it the best for everyone. E.g. If you're making a car, make it easy to get in and out for the elderly and then it'll be joyously easy for everyone else. Another classic example of this is the "Oxo Good Grips" range of kitchen utensils.
You’re applying it over-liberally. That school of thought is to apply accessibility features that maximize accessibility for /everyone/ - eg, automatic door openers.
Features that improve accessibility for a subpopulation while decreasing it for the broader population are /not/ part of that thesis.
The topic of dyslexia friendly fonts falls somewhere in between. Some fonts are considered friendlier than others and are fine for the general population (verdana), others are friendlier to dyslexics but hideous to the general population (open dyslexia). I call it a gray area because they’re ugly, and do not improve readability for the general population, but not overtly inaccessible to the general population.
Wrong, tools on the web should follow basic web design rules. The font should be replaced. You can not have bad UX because somebody thinks it looks cool.
> I don't believe we should adapt defaults to fit a single digit percentage population. I'm not dyslexic and the fonts suggested are horrible to read in my opinion. I'm totally for the option to set different fonts as default so dyslexic users have an easier life, but the global default should not be changed because of this.
This philosophy is based on the red-hot idea that the protecting the needs of someone in the majority is as important as doing it for the minority.[0] It also sets up the false dichotomy, part of the same rhetoric, that it's one or the other, them or me.
It's not the same at all: The majority doesn't need protection: Almost everything is built for and by the majority; the engineers, executives, funders, and reviewers are generally in the majority; everything they do is from that perspective, automatically. Including people who aren't in the room is a real effort. Also, there's productivity, stress, and basic equality: Almost everything people in the majority encounter is built for them. If you are in the minority, your day is filled with obstacles, problems, and ignorance, which not only slows you down but wears you down; it pushes you to the margins and makes you an outsider.
It's not a zero-sum tradeoff; you can help both groups. And generosity toward those who need it is a virtue; it's about justice and fairness, and their freedom to maneuver in society. I'm not sure what to call generosity toward those who have the power.
[0] Really, rather than majority and minority, we should be saying those with power and those without, but that adds another dimension to the conversation.
We’ve open sourced the font and made a lot of small but important changes to it over the past 6 months. If you are willing to help, it would be great if you could open an issue with specific words that are difficult to read with your dyslexia and we can figure out how to tweak the font to remove the ambiguity. Thanks for bringing this up!
> Are there any flashy overlay tours with annoying animations and pulsing buttons? Is there a modal upselling you on a paid plan[1]? No. You're just in product, able to use it.
This describes my first time experience with Miro. I just wanted to get started, but each new action resulted in another 'tour'. I patiently exited each tour, only to be hit by another tour/upsell/modal/what's new. Within 2 minutes I've now formed my first impression of Miro as being 'infested', and I reflexively microshudder when people talk about it.
First time user experience feels severely underrated by many companies.
You get them even when you’re perma-logged-in to a corporate account. (Ok, they’re “try our new feature” not “upgrade”, but they’re still an obstruction to “getting stuff done”).
My eyesight’s not great. I typically use my computer at about 1.7x with pan&scan following my cursor. It’s not ideal but I can see (and am a career front-end dev with global brands on my CV). ANY kind of attention grabbing device is 1.7x more distracting to me than someone who can use their machine w/out zoom. I rely far more on my memory of where content and controls apprar than a fully sighted person, so having to find a new control (close) for a popup I didn’t ask for is a significant workflow impediment.
It's like how some games will force you to sit through a 10 minute intro, followed by a tutorial level with "helpful" characters telling you how to use all the controls, instead of just dropping you straight into the action.
Being able to run an Adamant nature and still outspeeding most of the meta due to Sand Rush.
Half decent bulk with amazing defensive typing.
Multiple viable sets; you don't know if you're coming up against a defensive spinner or a juggernaut that can get to +2/+2 in one turn with Swords Dance.
Fwiw, Ken Wheeler suggested the name excalibur but when looking for the domain it was already taken but excalibur.draw was suggested. I ended up bringing the two together leading to excalidraw, which was available everywhere.
"Too see what's possible, hit the question mark in the lower right corner" this works, almost...
I have vimium installed so ? displays viumium help. This is a remainder that still there is a gap between native apps and in-browser apps. Yes, it is shrinking, but browser was never designed as the platform the run full-featured applications. So expect the unexpected.
But, true, comparing to other software of this kind this one looks really good.
browser was never designed as the platform the run full-featured applications
Does that mean they shouldn't ever change to become a platform for full-feature applications? Is there a reason why browsers shouldn't evolve and grow and add new capabilities? We don't believe that operating systems should be static once they've been released, or that applications should never add new features that enable them to run scripts or plugins. What's wrong with browsers doing more than Tim Berners-Lee envisioned 30 years ago?
Sure, browsers were once designed to view docs. Now they do more. That's a good thing even if it throws up some design, UX, and security challenges.
> "Too see what's possible, hit the question mark in the lower right corner" this works, almost...
I think they are referring to the actual button in the UI, which is in the button right, not the "?" key on your keyboard which might or might not be in the bottom right in your keyboard.
Anyway, as a user of Vimuim, you're surely used to keyboard shortcuts already being funky, and you know about the "escape" mode it surely offers in order to trigger keyboard shortcuts that conflicts with your own bindings. Just use that instead of complaining as it's your own setup that makes things difficult, not the website owners fault really.
It's a bit like writing something like "Hey, I use a extension that makes all elements background black on page load, and on your website, since the text is already black, makes it very hard to read stuff"
I say this as a vim-keybindings-in-browser user myself too.
Agreed. It was originally developed for desktop use cases and we did a straight port to mobile which surprisingly worked way better than I anticipated.
But we need to do another pass at it, the lack of move tool is really annoying indeed. I also want to add more gestures.
Procreate is so good and we should aim to be a similar experience.
As far as I can tell, https://excalidraw.com/ is already possible to use offline, since it seems to be a PWA.
Just a friendly reminder that "web !== requiring internet 24/7 internet connectivity", many client-side applications works even if you go offline, granted you have loaded the page at least once before.
Try loading the page, disconnecting your internet and then doing a hard-refresh/reload your browser.
277 comments
[ 4.7 ms ] story [ 236 ms ] threadExcalidraw+ - https://news.ycombinator.com/item?id=27034119 - May 2021 (31 comments)
Show HN: It was fun generating version distributions charts for Excalidraw - https://news.ycombinator.com/item?id=25792169 - Jan 2021 (1 comment)
One Year of Excalidraw - https://news.ycombinator.com/item?id=25608336 - Jan 2021 (19 comments)
Show HN: Create a Slideshow with Excalidraw - https://news.ycombinator.com/item?id=25081914 - Nov 2020 (1 comment)
Excalidraw whiteboard – easily sketch diagrams with a hand-drawn feel - https://news.ycombinator.com/item?id=23525648 - June 2020 (54 comments)
Building Excalidraw's P2P Collaboration Feature - https://news.ycombinator.com/item?id=22719576 - March 2020 (1 comment)
Show HN: Excalidraw – Sketch Hand-Drawn Like Diagrams - https://news.ycombinator.com/item?id=22146973 - Jan 2020 (6 comments)
Excalidraw – a whiteboard tool to sketch hand-drawn diagrams (excalidraw.com) - https://news.ycombinator.com/item?id=22104502 - Jan 2020 (3 comments)
Excalidraw – a whiteboard tool that lets you sketch hand-drawn diagrams - https://news.ycombinator.com/item?id=22101381 - Jan 2020 (21 comments)
Kidpix. Now that's a name I've not heard in a long time.
https://kidpix.app/
Now my only other complaint with excalidraw is that for some reason on my iPad Pro in Safari, the handwriting is incredibly laggy, increasing the longer the stroke is. Which is weird, because in every other browser I've tried (Safari/Chrome/Firefox on various laptops) it's incredibly snappy.
Glad to see things are improving! That's a sure sign of a company that listens to users.
We switched to the awesome “perfect” library from Steve Ruiz instead. Could you create an issue about the lag on ipad pro, we should investigate and fix.
Given that you already have Dependabot active, would you consider getting the dependencies somewhat up to date? Looks like the pull requests already exist... Thanks.
Did you read through the security issues identified, familiar with the codebase enough that you can see that these issues are actually affecting the product?
Most issues Dependabot opens on repositories I'm involved in, points to issues that has no bearing on the actual code, since it's security issues that are involving passing user code to specific functions, or simply regarding APIs that are never used in the first place.
You don't have to be on the newest version of every single library, in most cases it doesn't make any sense to just upgrade for the sake of upgrade.
Create a GitHub native tools suite for LTS management of all releases of all languages/platforms, then we are talking about security.
https://github.com/excalidraw/excalidraw/issues/4205
If there’s any other way I can help please lmk.
Will open an issue on github.
That said, the app itself has great UX, especially onboarding. Just straight to drawing mode. Kind of like autodraw.io.
[1] https://commons.wikimedia.org/wiki/File:Arpanet_logical_map,...
[2] https://monodraw.helftone.com/
The perfectionist in me won’t let me publish a regular, square-edged diagram unless it’s pixel-perfect. As a result, nothing ever gets published.
Take away that absurd situation, make the whole thing a bit shabby, and the need for perfection disappears. In fact the ability to make it perfect disappears.
Life changing. Sounds daft. Is absolutely true.
The most important missing feature at the time was that to doesn't show other user's selections unlike many other collaboration tools. That means what you see is not what you're collaborators see. If you grab some shapes they don't see that you grabbed them. If you highlight a word or item in a list, they don't see you highlighted it.
Further, unlike pretty much all similar software, they don't connect shapes and text. Place a shape, double click, a text box appears in the center of the shape. Drag the shape, the text doesn't go with it. This makes it super inefficient for most common use cases. People might want to separate the text from the shape but I'd argue that's an exception and should not be the default.
Also agree on missing some kind of laser pointer and or highlighter.
This should work out of the box if you're collaborating. But, we've recently upgraded our collaboration servers and due to ServiceWorker caching it may be your colleague (or you) were connecting to the old one. Try hard refresh your browser.
Since there was no link in the article :)
You mean like pretty much every single darned productivity program before year 200X?
It's sad that this even has to be an issue.
That said, I'd also misunderstood the grandparent comment: they meant that it's a shame that Excalidraw is one of the few apps nowadays that have these features, even if the other ones also have advantages like not requiring an install.
Makes me want to ditch the product on the spot. Yes, I've seen your emails. Yes, I will let you know if I need anything. No, I don't need a zoom call/onboarding video/introductory convo with your marketing team.
What I think those stats miss is that I will test a product with one email and if I decide to buy for a team will use a new account. So it looks like the conversion happened immediately skewing their sales/marketing strategy.
---
First of all it's a free, open tool, MIT license, that is also hosted for free.
It loads instantly without an installation step. You just type in the URL and it works.
It's not a "productivity program" specifically as illustrated in the article.
The design is very simple, intuitive and deliberately minimalist, while it does quite a bunch of subtle things under the hood that harmonize with the overall goal of the program.
---
Now there may have been tools kind of like this in the 90's. Not aware of any of them from the top of my head.
Of course, arguably one did have to pay for it and install it - but that was a one-off thing.
Yes, you could use Excel 95 by just double-clicking on an icon.
However, not only did you have to buy and install the software, which back in the day often meant going to an actual brick-and-mortar store to purchase a shrink-wrapped box labelled "Office 95", but it even required buying a specific type of computer with a specific operating system first.
In this specific case, that doesn't even take into account Microsoft's shady tactics of strong-arming vendors and OEMs into exclusively shipping Microsoft software with their products.
So, arguably, the barrier to entry still was a lot higher than with even the most obnoxious growth hacking tactics we see in many SaaS products today.
That said, I absolutely love the way Excalidraw are "onboarding" users.
It's just worth remembering that not everything went downhill with SaaS products (in fact, quite the opposite) and not everything was better "back in the day".
Excalidraw feels like a product of this time, and demonstrates some of the things we all enjoy about the web as an ecosystem. People compare it to some yesteryear product because of its purposeful feature set and look / feel, but name something from back then.
You insert the CD, install the software, and then it's ready to use. That's it.
To reiterate the quote from the article the GP mentioned:
> There's no onboarding. No signup. No confirmation email. No OAuth. You're just in the product.
It's just sad that you even have to have that in your checklist of things nowadays. Like it's amazing when software doesn't ask you to make another damned online account and just lets you use it. (And it's good that this doesn't, yay.)
Hosted is a negative feature to me (other than for trying the software). You don't control the software installation.
Tomorrow you could wake up and try to continue your work, to discover that some feature is missing you were depending on, or new bugs are happening that weren't there and so on. Or your local storage is borked because it's not being read correctly by the newly installed code.
It's good if something is hosted for trying out, but you want it to be locally installable so you and your team can depend on it, and evaluate updates before rolling them out, and roll back bad updates.
"Has local storage" is another thing that is sad to see as having to tick off a checklist.
Shocking! Who would have thought? /s
It would display a dialog when you opened the program. Every. Damn. Time. Displayed a useful tip from the included documentation.
At the time, it felt like a huge, complicated, OLE application, a big departure from the previous version, Word 5.1.
I suspected that some members of the Word 6 product team felt the same about that startup dialog: at a loss to suggest anything productive for that "Useful product tip for today", they wrote down some inane safety tips.
"Welcome to Microsoft Word... You can hurt yourself if you run with scissors."
I then read that it was free and open, so felt bad, but did notice an ad for Excalidraw+, which basically means my cynical first take is likely to come true eventually, as that is what the market will optimize towards.
You'd be lucky to have it all done in less than a week and a few hours of your time. How soon they forget!
You still have to today. A lot of stuff is Mac/Windows/Linux exclusive. Sure, I can run Windows stuff on Linux and vice versa if I'm tech savvy, but Mac stuff requires buying Apple hardware, so that's a massive barrier right there.
>figure out where and how you can buy it
Or you could just download it. The internet existed pre 200x as well.
> then manually enter a product key (fun!)
Honestly, it was fun. Much more fun than the phone-home, always-on telemetry and invasive DRM that act like rootkits & spyware we have today. Good luck playing older games you bought and paid for who's DRM require authentication from the publisher's servers that went offline years ago. Yeah, fun!
Back then, you paid for the product, you got your license and CD key, you owned it forever. No subscription rent seeking model BS.
Maybe I'm just old and cranky.
Outside of free software/shareware, there wasn't a lot of software available for download in the 90s. And compatibility was a lot jankier in that the base systems sold to consumers were insufficient to run a lot of popular software. For folks like you or me, probably not a problem (unless you happened to have a graphics card that wasn't supported) since we'd generally have obscene (by the standards of the day) amounts of hard disk and RAM, but Mom's computer might have 4M of RAM and a 10G hard drive.
Back in the floppy disk era the thing that always drove me nuts was that most (with very few exceptions) software vendors put the software key on the label of disk 1. Which would be in the drive when they asked you to enter the software key. I remember some OS/2 software that put it on disk 2 instead and I wanted to go to the company and kiss whoever it was who decided to do that.
Well - as recently as a couple years ago - even if it was a web-based tool, you would have to check which browser was supported.
Things have gotten better now that Microsoft has deprecated the entire IE-chain and moved to Chromium-based Edge. Things "just work".
I still have my Windows 98 product key and installer stashed away, purchased decades ago. And I actually used them a few months ago to stand up a VM to utilize an ancient printer.
Way better for me-the-consumer than the subscription-everything model. When software decides to go that way, I usually find a FOSS alternative or cling to the last standalone version.
Being prompted to create an account to use a word processor didn't come until later.
Should you choose to share a drawing, it would be end-to-end encrypted and sent to the Excalidraw server. The sharing URL contains the ID of your drawing for retrieval, and the encryption/decryption key in the URL hash, which never reaches Excalidraw's servers (thus achieving E2EE).
> deleting all of a website’s script-writable storage after seven days of Safari use without user interaction on the site.
So as long as you access the application once a week, you're safe. Or, alternatively:
> As mentioned, the seven-day cap on script-writable storage is gated on “after seven days of Safari use without user interaction on the site.” That is the case in Safari. Web applications added to the home screen are not part of Safari and thus have their own counter of days of use. Their days of use will match actual use of the web application which resets the timer. We do not expect the first-party in such a web application to have its website data deleted.
Still, horrible idea and horrible implementation, but I think it's known by know that Apple/Safari will always be hostile towards web apps as it cuts into their own App Store revenue.
Example of a saas that does this?
Excalidraw can be used on basically everything that has a compatible browser: Windows, almost all Linux distros, macOS, Android and iOS, major BSDs ... and that's not even counting in x86 vs x86_64 vs all the different versions of ARM machines we have now. Try to reach all those people without a tech stack based on web. Sure, you can still get desktop and mobile apps by cramming everything into electron or its mobile counterparts, but then we are back to comments like "I miss the days when desktop apps were less than 100MB"?
¯\_(ツ)_/¯
- 1, https://justine.lol/ape.html
We're also working to expand into OpenGL UX too. You can see our latest progress at https://github.com/jacereda/cosmogfx and the screenshots are here https://github.com/jart/cosmopolitan/issues/35#issuecomment-... which we're working hard to polish and get merged into the mainline. If you read threads like https://github.com/jart/cosmopolitan/issues/309 you can see we're committed to making sure your gui apps just work even on traditionally underserved platforms like OpenBSD because our goal is to help you reach the broadest audience possible.
Please consider supporting the project!
This whole cosmopolitan-fanboy'ism is completely ignoring my initial statement. What are you trying to bring to the table here? Sorry if I sound harsh, but you are really just proving my point here while probably not even realizing it.
I didn't expect anything of value or meaningful as a response anyway. Thanks for confirming.
> That only solves 10% of the problem, binary portability. You still need to code for a dozen different graphics, I/O, storage etc system APIs.
Read https://justine.lol/ape.html for a deep dive into how we managed to do it. It mostly just boils down to the numbers file mentioned in the last paragraph. That gets you far enough to do sockets and stdio. That in itself is a big step forward. Is it good enough to make apps like Excelidraw binary portable? Not yet. But it gets us much closer than we were before. Getting the rest of the way there is an ongoing topic of engineering and experimentation. Many people are helping. Further support is welcome from those who share a genuine interest in contributing to our efforts to solve the binary portability problem.
Of course I know excalidraw is open source, so someone could just fork it if the developers gave up on it. But this is only true with open source web apps.
Everything:
“ You're in #
There's no onboarding. No signup. No confirmation email. No OAuth. You're just in the product. Are there any flashy overlay tours with annoying animations and pulsing buttons? Is there a modal upselling you on a paid plan[1]? No. You're just in product, able to use it.”
Get rid of your fancy growth marketing people, your advisors with their “request a demo”.
Get rid of your intro page talking features.
Dump your stupid signup before starting.
Just. Fucking. Let. People. Instantly. Use. Your. Full. Product.
Instantly. Not after they give their email. Not after they click three dialogs. Not after they jump through whatever remaining hoops you’d really really love them to jump through before.
Show them, indeed GIVE them your product without the slightest fucking delay.
All of it. Stop arguing about why you’d really love to make them do something else before using your product. Make your marketing person stop insisting the marketing thing gets done first.
Get people into your actual full product within the shortest possible time. Milliseconds if possible.
- Does it actually lead to higher sales? If not, nobody is going to care about our wish for more companies to do this.
- It doesn’t work for many products. How do you just drop someone into a cloud storage product? You could give them some temporary space where they can drop some files if they want, but what do you do if they abandon them? Delete them after some time? Might confuse some customers who didn’t realize they had to actually register an account. Maybe a huge banner with a countdown would work… you start to have to solve a whole lot of problems to make it work
This stifles development of things that are actually useful. It puts money and influence in the hands of people who shouldn't have either, creating bad incentives in the market and within the company. It gives marketing a gloss of false legitimacy, allowing for all the things that have us in a feverish race to the bottom.
Everything you just touted has been achieved despite the Idiocracy fan-fic that seems to be the current mission statement of corporate America.
In the webapp ecosystem, those transaction costs take the form of annoying signup pages and mail spam. Could be worse.
- Products that don't require any of this stuff.
- Products that require some form of registration since it stores user data remotely. The person registers so they may secure, modify, or remove that data. The handling of personal data should serve this role exclusively.
- Products that require some form of initial setup and onboarding will facilitate that setup for the vast majority of people. The user should still be able to bypass the onboarding in case their use case doesn't reflect the common use case.
When marketing becomes involved, things become problematic. That friendly offer of assistance to get started usually reflects marketable features rather than any form of genuine training. That collection of personal information is used to harass the person using the product rather than to facilitate the use of the product. (Heck, I have companies contacting me over a decade after I last used their product to try to make a sale.)
I think this is a really, really interesting phenomena: the human brain (perhaps when sufficiently institutionalized) exhibits some sort of fundamental bias toward disruptive UX that breaks flow and productivity, and this approach must be proven wrong, every time, on a case by case basis, and doesn't fundamentally internalize.
It's crazy: because it doesn't lead to higher sales, it's seen as a bad approach. WHY? "Just because."
If the goal of a company is to create higher sales, then it is a bad approach.
If the goal is to create a great tool used by people and not worry about adoption or supporting a large business through profit, it's probably a great idea!
I love that - and that's how I aim to build things now.
Excalidraw has few calls to external domains as well - just Google Analytics and Sentry.io. That's still two that can be reduced, but it's much better than most modern clustertrucks.
That said, those two have been really important for the success of the project. Google Analytics is helpful to get a sense of how important it is to people and motivation for the team. The second one helped us fix a lot of edge cases that couldn’t immediately be found via normal testing.
But it's not exactly clear how you plan to turn that into, say, revenue. Build it and they will come, sure. And the more leads you stuff into the top of the conversion funnel, the more conversions, right? What do you mean 'qualified' leads?
If your product's viability will rest on turning some percentage of those free users into subscribers or purchasers, then you have to do some upsell, at some point. And if they can get into the full product without any obstacles within milliseconds, you're going to have a very, very hard job getting those conversions. Meanwhile you're winning over free users who cost you money and time to serve in handling their support and feature requests, and you're bringing in ever more people who are never going to give you any money.
It is viable to build your product this way if the primary product your business sells is 'convincing investors that by building up loyal active users, and developing rich understanding of those users' behavior, you will somehow create a reservoir of value that can be sold for 10x the amount of money you're asking them to put in to keep your servers turned on for another few months'
And it's probably viable if you aren't trying to be a business at all.
But as generic 'this is how all products should work' advice, it's pretty terrible, unless your main advice is 'don't try to make a business around offering a software product unless you're primarily interested in selling engagement monetization fairytales to investors'.
Ready to save your new work of art?
Or higher resolution?
Or premium features?
You do it when the user is hooked.
I was not advocating for “no revenue”.
Also it seems self evident that such advice does not apply to all software…. Do you think I’m saying this is how SAP should be sold? No…, obviously the advice applies to software suited to this approach.
‘Thanks! Oh wow, this is great!’
‘Glad you like it.’
‘I love how you don’t try to upsell me on things or push any kind of premium paid features, just give me the whole product without limitations’
‘Yeah, that’s our philosophy, let you fall in love with the product’
‘Cool, thanks. Okay, I’m gonna save my work now.’
‘$20, please’
‘I’m sorry what?’
‘$20 to save.’
‘Er… could you have mentioned this earlier? I just spent ages using your awesome product and you just let me in and I didn’t have to sign up or anything, and you never mentioned a premium upsell, so I just assumed…’
‘Yeah, I thought I might be able to trick you like that. We call it “bait and switch”. Isn’t it so much better than all that marketing BS?’
The key realization is that while the unauthenticated open source model works really well for individuals, it’s hard for companies to adopt.
They want authentication, security policies, private spaces to collaborate, access control, search… all of this requires having a company behind it in practice and the open source model doesn’t work as well.
The two versions so far work really well alongside each other right now.
How did you get people to try the free version? What worked?
Only a relatively few people like you and me really want to dive in without having a story first. Turns out, to my surprise (!) that most people don’t want to dive into your app without reading about it or watching a video first.
Teachers, kids, business users, they all have different questions they want answered before they spend time trying to learn a new app, they want to know if there are reasons this can or can’t be the thing they use from now on. Learning a UI is a time investment, and people want to make sure it will be a good investment before using their time. They don’t want to be tricked into learning your UI for a week or two or more and then find out later it can’t group and copy-paste shapes, or that the price goes up, or that they can’t share they way they need, or that your app doesn’t have a on-premise installation.
Getting in quickly can have a low barrier to entry, I’m with you there, but my experience on the other side of this is that it’s not good to go directly to the app without an intro first, generally speaking.
Aside from this there’s also the fact that users who don’t intend to pay are on average way more demanding, way less patient, and way less understanding about communicating and learning the UI and dealing with bugs or misunderstandings. It was surprising for me to see from the other side how rude the free users are on average compared to the paying users.
Or create a youtube channel for it.
No need to spoil the tool for everybody with marketing cruft.
In my experience, this also applies to cheap vs expensive projects (consultancy). Customers start to get really pleasant and polite when they are charged at least 100K or something like that for doing some custom software development. They get very demanding when it's below 10K.
Shouldn't it be win/win? Both sides giving something to help each other? Software gets tricky I guess since it's fun to develop, and mere attention or praise is enough for many devs. And ads blur the line of "free". But the level of childish entitlement gets a little depressing sometimes.
I don’t see it as having to do with entitlement, more as a case of it being easier to standout in a crowd of options when your thing is easier to start with.
You are also advancing the art. Which is an objective good.
Defaults matter a lot.
Feel free to jump in on it
https://opendyslexic.org/
Notice how parts of the letters are weighted. That helps with identifying which part of the letter is bottom.
And its monospace equivalent for coding is OpenDyslexicMono
Is it maybe just a matter of getting used to it?
Otherwise sure, the bewt option would be probably for people to use their own font and make it easy to do so (@browsers)
Features that improve accessibility for a subpopulation while decreasing it for the broader population are /not/ part of that thesis.
The topic of dyslexia friendly fonts falls somewhere in between. Some fonts are considered friendlier than others and are fine for the general population (verdana), others are friendlier to dyslexics but hideous to the general population (open dyslexia). I call it a gray area because they’re ugly, and do not improve readability for the general population, but not overtly inaccessible to the general population.
This philosophy is based on the red-hot idea that the protecting the needs of someone in the majority is as important as doing it for the minority.[0] It also sets up the false dichotomy, part of the same rhetoric, that it's one or the other, them or me.
It's not the same at all: The majority doesn't need protection: Almost everything is built for and by the majority; the engineers, executives, funders, and reviewers are generally in the majority; everything they do is from that perspective, automatically. Including people who aren't in the room is a real effort. Also, there's productivity, stress, and basic equality: Almost everything people in the majority encounter is built for them. If you are in the minority, your day is filled with obstacles, problems, and ignorance, which not only slows you down but wears you down; it pushes you to the margins and makes you an outsider.
It's not a zero-sum tradeoff; you can help both groups. And generosity toward those who need it is a virtue; it's about justice and fairness, and their freedom to maneuver in society. I'm not sure what to call generosity toward those who have the power.
[0] Really, rather than majority and minority, we should be saying those with power and those without, but that adds another dimension to the conversation.
https://github.com/excalidraw/virgil
This describes my first time experience with Miro. I just wanted to get started, but each new action resulted in another 'tour'. I patiently exited each tour, only to be hit by another tour/upsell/modal/what's new. Within 2 minutes I've now formed my first impression of Miro as being 'infested', and I reflexively microshudder when people talk about it.
First time user experience feels severely underrated by many companies.
My eyesight’s not great. I typically use my computer at about 1.7x with pan&scan following my cursor. It’s not ideal but I can see (and am a career front-end dev with global brands on my CV). ANY kind of attention grabbing device is 1.7x more distracting to me than someone who can use their machine w/out zoom. I rely far more on my memory of where content and controls apprar than a fully sighted person, so having to find a new control (close) for a popup I didn’t ask for is a significant workflow impediment.
Being able to run an Adamant nature and still outspeeding most of the meta due to Sand Rush.
Half decent bulk with amazing defensive typing.
Multiple viable sets; you don't know if you're coming up against a defensive spinner or a juggernaut that can get to +2/+2 in one turn with Swords Dance.
Stab EQ with decent coverage.
"Good" is an understatement and a half.
I have vimium installed so ? displays viumium help. This is a remainder that still there is a gap between native apps and in-browser apps. Yes, it is shrinking, but browser was never designed as the platform the run full-featured applications. So expect the unexpected.
But, true, comparing to other software of this kind this one looks really good.
Does that mean they shouldn't ever change to become a platform for full-feature applications? Is there a reason why browsers shouldn't evolve and grow and add new capabilities? We don't believe that operating systems should be static once they've been released, or that applications should never add new features that enable them to run scripts or plugins. What's wrong with browsers doing more than Tim Berners-Lee envisioned 30 years ago?
Sure, browsers were once designed to view docs. Now they do more. That's a good thing even if it throws up some design, UX, and security challenges.
I think they are referring to the actual button in the UI, which is in the button right, not the "?" key on your keyboard which might or might not be in the bottom right in your keyboard.
Anyway, as a user of Vimuim, you're surely used to keyboard shortcuts already being funky, and you know about the "escape" mode it surely offers in order to trigger keyboard shortcuts that conflicts with your own bindings. Just use that instead of complaining as it's your own setup that makes things difficult, not the website owners fault really.
It's a bit like writing something like "Hey, I use a extension that makes all elements background black on page load, and on your website, since the text is already black, makes it very hard to read stuff"
I say this as a vim-keybindings-in-browser user myself too.
But we need to do another pass at it, the lack of move tool is really annoying indeed. I also want to add more gestures.
Procreate is so good and we should aim to be a similar experience.
I’d suggest a 20-pixel gutter on either side of the page.
Looks like an interesting product. Thanks!
I wouldn't mind an offline version. I usually need to do this kind of work on planes.
Just a friendly reminder that "web !== requiring internet 24/7 internet connectivity", many client-side applications works even if you go offline, granted you have loaded the page at least once before.
Try loading the page, disconnecting your internet and then doing a hard-refresh/reload your browser.