If you have uBlock Origin - and who doesn't - you can use the 'element zapper' to do targeted killing of such annoying elements. It does wonders for those sticky headers, grey-out overlays and similar nuisances. Bind it to a key - I use CTRL-ALT-Z for 'zap' - and you don't need to leave the keyboard to go on a killing spree.
Haha this is funny. Does this even work for pages that are written with frontend frameworks? Needs a remix for 2024.
Virtual DOM refreshes might bring back the original style if not the elements themselves.
As insane as it sounds what probably makes more sense is to query for and remove the elements, generate a PNG of the page, delete everything in the body, and make sure the only child element of the body tag is your new img tag.
Also since you cannot unload or make undefined all the other js already running you may want to add some defensive event handlers on the body tag with MutationObserver.
At this point I have PyCharm as my IDE, Obsidian as my official "Notes" with a capital N, and about 40 open unsaved text documents in Sublime Text that i just use as scratchpads. Writing down things, acting as a poor man's extended clipboard, etc.
I absolutely love this flow and it works so well for me now.
I recommend you check out “Scratches” in PyCharm (or any IntelliJ IDE). Cmd+Shift+N I think? I type it reflexively, it lets you create a scratch pad of any “type” (Python, PHP, plain text, shell, everything IntelliJ supports) and write code with syntax highlighting and all the bells and whistles you are used to.
They are all saved in a folder on your computer and easily accessible from the “Scratches” section at the bottom of the file explorer.
It’s my go-to for any text manipulation and I think you can even configure it up to run your snippets of code if you want (I’m stuck using CodeRunner cause I’m too lazy to look more into “running” Scratches).
I also use it to paste in information as I’m debugging a problem. Then I can use multiple cursors or regex find/replace to take information and turn it into commands to run. Yes, I know about CLI tools built in to most systems and I use them to create one-liners but sometimes I want a little more direct control and/or I want to build a script from the commands.
> The product is browser first and document history is stored locally
so.. for throwaway documents then?
browser storage is one of the worst data places imagineable. Hard to backup, very hard for user to access, impossible to sync or share. And very unreliable too: many troubleshooting procedures for web apps start with "reset browser profile" or "clear site data"
I assumed it was letting you store it on the local filesystem, not in the browser storage, which I agree is a bad place to store anything. There is one site I use that stores my favorites in the browser store and I had to write two JS bookmarklets so I can import/export my settings on a regular basis because pretty much every time Chrome crashes you can be assured that all my cookies and storage data gets zero'd out.
It sort of does. Not in a strict semantic sense, but I don't think any user of a service that says it stores stuff "locally" expects it to upload that to somewhere else without explicit consent.
It sounds like your interpretation is like if I said "our data is stored on our servers" but that actually meant "our data is stored on our servers, but also on AWS, GCP, Azure, Alibaba and shared with FSB, CIA, CCP, and the public and other parties (including but not limited to the Royal Bhutan Police)".
That would be true, but misleading to the point of being meaningless or (hopefully) potentially criminal.
There is a huge difference between "document history is stored locally" and "works offline".
The local-first movement might sometimes talk about privacy and offline availability in the same breath, but those are very different concerns and if someones says "data is stored locally" then I assume they should check both the privacy and offline boxes but with a huge "I might loose all my data" caveat. If it still uploads my data to some cloud then I would not describe it as "document history is stored locally".
I tried using the product and out of curiosity inspected the editor. The editor is basically a CodeMirror plugin with all the tech heavy-lifting like realtime collaboration is enabled by CodeMirror itself.
It's clearly an overstretch to call this a Google Docs alternative, especially since Google docs is powered by insane levels of engineering. I know because I work on a competing word-processing product.
It's mildly surprising seeing such a shell product receiving millions of dollars in funding and even a published post in TechCrunch!
I'll be swapping over. My usage for Google Docs is little more than pastebin + formating + access control. I suspect this is also probably a huge chunk of their overall use-case as well. In general, the least engineering amount possible to deliver what's actually needed is best IMO!
There once was an incredibly nice product like this called hackpad. It even had a similar theme IIRC. It was a yc company, lovely polished product, and the folks I was founding a startup with loved it with zero effort spent on convincing them.
I just got hedgedoc (nee codimd, nee hackmd [1]) setup for my own purposes. It does have the advantage of splitscreen edit/preview, and it has the option to use vim/emacs keybinds in the editor, but if I'd had stashpad, I probably would not have bothered setting up a selfhost hedgedoc.
+1 for hedgedoc. Only weird thing missing is an admin interface and easy way to check if no one is abusing any of those forgotten public pads that are hanging about.
"only", when 99% of the people are perfectly well served with the featureset of yesterdecade's LibreOffice suite, the three problems being the network effect (if everyone else is on another "network" or standard, you're forced to as well), that it looks like 2002, and that they have to spend 10 minutes on the first day to figure out that some buttons are in different places (I'm lost in Office nowadays, it really is just a matter of habit and not that either of them has an inferior layout). With the only functional defect being "it looks old", a modern style is what you pay 150€ every year for when buying into the Office monopoly. Except as a student: they're happy to help you gain the habit and experience the network effect.
I guess this is why Google Docs has been the first to make a dent: you seamlessly download the software to render the file format upon pageload, so there's no network effect on file formats anymore. Add a modern look, reach into your pockets to make it free until you've established a good market share, advertise for it in your already popular products (costing them nothing), et voila
The real secret we need to know here is how to convince TechCrunch to write about you when you have the most generic product on the planet. Knowing journalists is the secret?
Not sure where the data is actually stored, but assuming the last version is stored in the cloud, how can they prevent users from sharing data e.g. music though the docs? And how can they they stop users from doing that?
Using the free version for anything sensitive might not be a good idea, especially since I can't find anything about their retention policy or something like terms of service.
The lack of needing an account is a (or the) key feature: one I wish more online services copied. How much better and more frictionless would the web be if we weren’t having to log into an account everywhere we went?
If each doc has a unique URL, then bookmarks? Most browsers already do this. I’d be willing to argue that browsers should make sharing bookmarks from one device to the other easier and account-less, too, but that’s a different rant.
> The product is browser first and document history is stored locally, so users can search for docs without querying the server. The company said that while there is no offline support at the moment, it is a feature that the startup will introduce in the future.
>
> You can easily share these docs for others to look at and collaborate with. However, some of the features like read-only sharing are behind a login.
it took you longer to write that comment than it would've taken to get the answer.
data:text/html, <html contenteditable>
It's great for quick notes between tabs or copy/pasting things together into a cohesive comment.