Hacker News Reader

Comment by Multicomp | Hacker News Reader

I find the vimmaxxing message common enough but I disagree with it.

Not because vim vs nano vs emacs etc. Just that plain text is not enough.

I want to be able to encode more of my memory and context to my notes with a minimum of fuss, and therefore want to embed images (gasp!), tables, hyperlinks, and even file or sound embeds in my documents.

Yes, open simple file formats are better than closed complex formats (and doing text vs binary formats has something to say for it as well).

No, I don't like that Onenote's .ONE file specification is super complex and therefore only semi-open.

But I'm not about to lock myself into a DEC dumb terminals 80x24 limitations where I have to learn some greybeard's key bindings for everything, all for the privilege of losing all my ability to embed multimedia combined with said gatekeepers tut-tutting me that I should just make a link in my terminal to the local image file, nevermind that you can't copy/paste an image into a bash terminal and have the file be saved to that directory (ok it probably exists but is some obscure something).

So I watch Xournal++ very closely, specifically [this issue](https://github.com/xournalpp/xournalpp/issues/937) because it will have the total package, multimedia embeds, ink support, text support, file embeds, all in one file that I can syncthing save to wherever else I like.

But until then? Onenote desktop, sorry but not that sorry.

View on HN

> I want to be able to encode more of my memory and context to my notes with a minimum of fuss, and therefore want to embed images (gasp!), tables, hyperlinks, and even file or sound embeds in my documents.

I do this stuff with my notes in Emacs and it's really nice. Emacs is a graphical application and its notetaking facilities are really outstanding. It even supports syntax highlighted, executable, exportable code snippets, as well as all the basic text formatting, hyperlinks, tables of contents, tables, etc. It's a great tool for developers and sysadmins to use for notetaking.

If you know Emacs isn't for you, that's totally fine, but I wanted to throw this out there for bystanders that when someone suggests to build your notetaking workflow around a great text editor, the proposition is not at all to somehow confine oneself to an 1980s terminal.

Same here. I use Emacs to write tons of notes. I rely heavily on tables, hyperlinks, timestamps and tags. Images are also possible, just not part of my workflow. All this is trivial, out of the box and I know it will still work many years down the road.

Even custom hyperlinks and text notes that refer to chunks inside PDFs are easy to do.

My life and work would be impoverished without emacs.

Thank you Richard

I personally dislike emacs, but say the same.

Why?

vim would not be what it is, without emacs to rage against.

vim would not be what it is without having become a shallow emacs.

Emacs is emacs because its extensible, between vim, neovim and vim9, im starting to think people might enjoy the extensibility seeing as that brings vim up to 3 scripting languages to write things in.

vim has become what it purported to hate

I don't think Vim ever hated extensibility. Vim is defined by its modality and vocabulary and grammar, which Emacs lacks without extensions like evil.

Your phrase 'Images are also possible' worries me. But then you go on to say 'All of this is trivial ...', which makes me feel a lot better.

I've had many well-intentioned but ultimately abortive attempts to get on board with Emacs.

So, are images 'possible' or 'trivial'? Not trolling, and I do note that you say it's not part of your workflow.

eehh...

so (out of the box) in org mode if you type [[./file/path/to.png]] it will turn into a link, if you hit Ctrl-c-x-v it will toggle between showing images as a link (which will open in a new emacs buffer) and just showing the images inline.

ALSO you can just drag an image file onto an emacs window and it will display it in a new buffer.

however getting it so that A) images are displayed automatically when you finish typing the [[path]] and/or B) dragging an image into emacs will result in a new link in orgmode... that's probably possible

https://github.com/abo-abo/org-download

This is the package you're looking for. Just drag and drop images into the org mode buffer and you're done. I've been using it for years to take lecture / book notes in org mode where I have to constantly take screenshots and embed them in my notes.

that's part B.

part A is still annoying though. if you have inline images toggled and type in a new image it still turns into a link instead of an inline image.

That's an org mode setting, can be added to the #+startup declaration at the top of the file.

You cannot embed images in Org documents. Or rather, you can, but would need to code the support for this for yourself. On the other hand, displaying images within the Org documents is trivial. You just need to have the image file saved somewhere. Then you insert a link to the image file and, with the default settings, you should be able to see the image rendered inline in the document. You can also generate the image from within the document. You write a ditaa code block, execute it, and the png with your UML (or whatever) is inserted or updated.

On the other hand: no, Emacs is not a WYSIWYG word processor. The need to support terminal as a display backend weights heavily on the pace of GUI improvements. That's because all interface elements need to be renderable on all backends (not strictly true, but the prejudice against GUI-only elements is significant). So if you hope to be able to easily and conveniently display something like this: https://klibert.pl/statics/vw-doc-view.png then Emacs is empathically not the best tool for the job.

It’s as trivial as images in markdown, but there are more optional config options and extensions to, e.g., just paste from the clipboard.

As far as I can tell, embedding images into org documents still requires saving them to a directory and linking them. I can't seem to find a way to get it to actually embed the image data into the document, for example like a base64 data URL or something.

This is the only way I've done it, since my org notes are organized into git-based wikis and that's kinda what I expect from wikis.

Direct embedding binary data in the Org files would be cool for emailing Org files around, though. It looks like Org does support this kind of embedding on export, if you don't mind exporting to other formats when you send documents around.

Base64 images and URLs are possible with extensions, but simple files are better IMO. Getting URLs asynchronously doesn’t work well, and base64 embedding seriously bloats the text file and makes it more difficult to plain text tools on it. (E.g., you’ll get false positive grep results.)

Yea, you will get a lot of Emacs and Emacs with Org mode answers here. To be fair they not wrong :) As a complete emacs newb, start with something like "doom emacs" ! It's brilliant for beginners.

>Emacs is a graphical application

Too few ppl know this.

> Too few ppl know this.

Including me until now. I will examine this emacs of which you speak. No promises but I can at least attempt to run it.

ProTip: Not many ppl like YouTube as a "learning platform" but just look for "doom emacs" on youtube to get the basics. You only a need to "learn" a minimal set of keys+concepts to start to get insane value.

Goodluck :)

Seconding the Doom recommendation. There is no need to become a guru before you can be productive, or to use Emacs for all your coding just to make good use of it for notes. Doom is a great way to quickly get up and running for notetaking with Org mode.

Get used to the basic workflow for a little while and forgive the monospace-ness while you evaluate the workflow. I like taking notes in Gollum wikis, the wiki format used by GitLab and Github. The project management features enabled by default in Doom will kick in for that. Decide how you feel about those.

If you like the workflow, take a look at org-modern and take a few minutes to work out the proportions for variable pitch fonts. Then you can have Org text render in a nice, proportional font that you like and use monospace only for the embedded source code (including shell languages) blocks.

As an alternative to the using git repos to store notes collected into projects, take a look at Org Roam v2 for keeping a topic-based knowledgebase that you can work with without thinking much about the filesystem, and which offers mindmapping features.

That stuff works great for me, has required zero substantial elisp knowledge to get working, and has been natural for me to ease into. I hope you find a workflow you like!

So ? :) How did it go ?

I have images, audio, and backlinks using UUIDs in my terminal emacs. The image support is the worst, as I have to open the image and then it takes over the screen until I dismiss it, but it’s usable.

Agreed 100%. My core workflow is "snap a picture, draw on it." I deal with hardware, and the bandwidth with which this lets me get physical information into the computer is so far beyond any Markdown-based workflow that I can't go back. I suspect this result would hold for many non-software professions.

For now, I've settled on GoodNotes because it's good enough at PDF Backup, OCR Search, and drawing. I have given solid tries to Evernote, ZoomNotes, Concepts, and Notion, and they each are amazingly better than GoodNotes in at least one major respect, sometimes multiple, but none of them are good enough on those zero-compromise requirements and GoodNotes is. FWIW.

EDIT: Oh, while I'm flamebaiting, I also use my ipad camera to take pictures of screens. Life is too short to schlep screenshots between computers, especially the ancient ones that run scientific equipment (lots of 68000, XP on some recent ones). Printer emulators and java applets can get bent. Runs everywhere, my ass. Moore's Law has given us gigahertz and megapixels, and if I can use them to eliminate painful asinine pointless busywork that's my god-given right and nobody is going to convince me otherwise. Sometimes I even use my ipad camera to take a screenshot of something on my PC and toss it into the mix. Sue me.

Shift+CMD+5, adjust rectangle, CMD+C

Paste on your iPad

Clipboards are synced through iCloud.

kde connect does that, well with text atleast, you just do copy on your desktop/phone and paste on the other device because it keeps your clipboard in sync. now kde connect is even on IOS so basically all major operating systems now has cross compatibilty via kde connect which is cool.

if i had to share a screenshot from my laptop/phone, it is a matter of taking the screenshot and "sharing" via kde connect. take all of 3-5 seconds and this workflow is now available to everyone

Nice, I'll have to give it a try!

Windows. Thanks, though.

I swear I'm not shilling for Google Keep, but I've found myself using it more and more over the last couple of years; it hits a sweet spot between "limited enough that I don't get distracted" and "advanced enough that I can still express myself".

The biggest praise is its ubiquity. It's just there when I need it. Interesting article + my in-the-moment thoughts? Google Keep-ed. Want to add a reflection at the end of the day with a picture? Google Keep-ed. Need to remember an address in the car? Toss it in the Keep and open it on my phone.

notion/joplin or another note software

Obsidian! #justTryIt

Obsidian changed everything for me. Years of markdown notes and pdf's finally in one place with easy search and synced to my Synology. Svg for the visual stuff. And Vim key bindings. Multiplatform and extendable with plugins.:)

this, the fact thatit is just markdown files means i can add images in the app, should i need and i can use this to add a quick note from the terminal

  function qn ()
  {
    date_today=$(date +%F)
    year=$(date +%Y)
    month=$(date +%m)
    day=$(date +%d)
    FNAME="$NOTES_DIR/Work/journal/$date_today.md"
    #FNAME="$NOTES_DIR/Work/journal/$year/$date_today.md"
    echo "### $(date +%F' '%T)" >> $FNAME
    rlwrap cat >> $FNAME #
    echo -e "\n" >> $FNAME
  }

I've tried a lot, but obsidian works for me

I've never seen `rlwrap` before, thank you. I am adapting this to my workflow. I use Org mode, but as it's still the idea is the same.

Thank you!

Plugins are a HUGE plus.

I can't recommend Obsidian enough. My only complaint is that for a fairly straightforward tool its UI performance wasn't amazing on my previous machine (an early 2015 MacBook Pro so pretty old but no slouch) but on my new machine that's not an issue at all.

I totally agree with you. One Note is fantastic and I would like to use it but not being able to open the notes in linux makes me avoid it. Do you know if there is a way of opening One Note notes in Linux without using the webapp? (I don't even need to edit them!)

In the meantime, I'm using stylyslab write [0] which uses svgz and is a good replacement for OneNote and a compromise I'm willing to make to not be locked in but tbh the OneNote app is just so much better.

[0] http://styluslabs.com/

You are correct, I have write3 installed on my phone and need to use it on the desktop too for that same reason.

I keep having dreams of making a linuxnote for OneNote files since the file format is technically open specs, but as they are super complex and I've barely opened a hex editor before I just don't know what I'm doing enough to actually try parsing the bytes of an arbitrary file yet.

> and therefore want to embed images (gasp!), tables, hyperlinks, and even file or sound embeds in my documents.

And exactly which one of them can't I do in asciidoc?

I guess he means in a wysiwyg way. That's all fun and game till you have to open a different editor for multimedia stuff, render the document and update it "blindly" in the source so everything works...

Doesn't even take until then before it fails for me...wysiwyg formats are traditionally a nightmare for version control systems. All my plaintext notes in Markdown and Asciidoc can be checked in without a problem :-)

I did mean a wysiwyg style yes.

If it is more than step 1 cut, step 2 paste, the tool has too many steps for my workflow, which like other users have posted is mostly a braindump of knowledge or context around several tables, links, and images/screenshots, sometimes with digital ink if I'm on a device with an active stylus.

But for producing formal documents, I have checked out asciidoc before and have respect for its capabilities, I just don't consider it in the running for my particular workflow as I see it as a different tool.

I'm not after version control for my notes though I know many others like it.

I wrote a library to parse plain text notes, sowhat [0] that delivers a lot of semantic and structural information. Notes for me solve lots of problems in my routine that wouldnt work with a markdown implementation: tasks, budgets, time management, reading lists, work log, simple calculations that operate on the global notes environment, etc.

Sowhat handles transactional data well, you could implement double-entry for example with relative ease. Other things include: links, events, quotes, tasks, formulas and a few organizational elements. I combine these elements in different ways depending on the problem.

[0] https://github.com/tatatap-com/sowhat

Rough, hand-drawn diagrams and spreadsheets are where my text-based note taking system falls apart.

Org mode has a table feature that does most of what I'd do in a spreadsheet anyway.

https://orgmode.org/worg/org-tutorials/org-spreadsheet-intro...

The reason I really like text for a lot of note taking is that I find it a lot easier to use compared to a word processor. Word processors tries to be smart about formatting, but never fails to annoy me. I still struggle editing the text of a link regularly.

But spreadsheets applications tend to be more useful than annoying. Doing tables with formulas in text docs are possible, but it seems more annoying than useful. In some instances I enjoy using jupyter notebooks.

iOS notes has changed my life. It's available on my work Mac, home Mac, and iPhone, and is always synced. It has everything I need and nothing more.

I stopped using it when I needed to work in Windows. I use InkDrop these days which is cross-platform.

I like Bear.app (MacOS, iOS, iPadOS) even better. Specifically, I love that it's Markdown-based and has support for code blocks with syntax highlighting. Only problem is that it uses iCloud Drive for syncing and my employer has blocked that. :-/

I agree with you, though: Notes is really quite good. I'm not sure how Notes syncs, exactly, but for whatever reason it's not blocked by my employer.

I think I've just decided to drop Bear for Obsidian on Mac. I need that plugin support (hard to leave VSCode line shifting hotkeys), and the code blocks work so much better.

Hmmmm, thanks for giving me something to check out.

Worth mentioning that it's also available on web at icloud.com. I don't have a mac, but have an iPhone and iPad, and i love the Notes app for quick things and lists that I want to remember - and it's accessible from anywhere.

Yes, same experience, but I wish there was code formatting in Notes. A lot of my notes are code snippets and Notes fails there. I've recently switched to Obsidian with the iCloud drive backup and trying to get used to it.

Beware: there is not an easy migration path out of Apple Notes.

Also sync is not that great for massive datasets: I tried to sync my 22k-ish Evernote notes to Apple Notes and it took days and I never got the same number of notes on all the different devices. If you are currently building your notes you will probably not encounter this issue, though.

I also almost have 20k Evernote notes.

I was considering obsidian, but the devs say that obsidian is designed for a maximum of 20k notes.

So, I'm already on the upper bound, as you are. Probably, obsidian would be slow if I import all my notes.

Evernote, on a Windows 10 Desktop, now works fast, at least for me. I think we undervalue this fact, we undervalue the fact that it is hard to make good note taking apps that can sync and search fast with more than 20k notes. I guess there are users with 60k notes!

Evernote, for all the flak it is getting, is still a remarkable product and I feel like the pricing is fair for what it gives. It ingests everything from everywhere, it indexes and scans everything for a flat fee.

They are currently still catching up to reach feature parity with the old native clients, but at least they are releasing more frequently --which means betting the house on Electron technology is paying off--.

I do not use Evernote any more (finally ended up with plain files in iCloud Drive and EagleFiler to top it off) but I am still following their releases as a paying customer.

Evernote is indeed a great piece of software. They have gone through very difficult times, but now the Desktop app is working very well (not the Android client, which is very slow, but improving), it is fast and despite it lacks some old features, it has great new features, like Tasks. I would not be able to work without it. Each time I thought about moving out, I realized I could not, because of some nice feature. So, Evernote is great and it is here to stay. My feeling is that those who are happy with Evernote do not write blog posts or tweets about it, it is mostly the haters and the frustrated who are writing out there.

Notability will pop your block off

CUA bindings, for the initiated (like me!):

https://wiki.c2.com/?CommonUserAccess

What about running wordpress locally instead? It does sounds like suitable for your needs. A single docker-compose.yml is all it takes.

Terminals have improved a lot since the dumb. Think your comment would do better without the slight. However, I use CUA keybindings myself.

I guess the tweet about it is a joke but funny though. It says even use Vim for tweets, but then it also says sent using IPhone.

Sounds like HTML to me, or at least the output of it.

Sounds like TempleOS is the choice for you!

On my scale of highly upvoted uninformed opinion, this ranks high up there, next to Bill Gates microchip vaccines...

> On my scale of highly upvoted uninformed opinion, this ranks high up there, next to Bill Gates microchip vaccines...

Is this comment auto-generated? What do you mean?

or you could use xterm's vt240 mode (from 1989 or so) and use emacs or vim extensions that embed sixel data in text files. just because you didn't read the manual doesn't mean it can't be done.

There's a million things with manuals to read. I would rather grab something that just works in an obvious way.

This is oddly sad to me. It's like hearing somebody say that you don't need to read a book when you can just watch the movie adaptation. You're missing out on so much depth and growth potential by just settling for the easiest option.

Maybe I'm just a weirdo. I love reading manuals and discovering interesting nuances in the things I use every day. Or picking some new software at random and learning about it just because it's fun and interesting.

I don't see the comparison holding water. The difference is between a tool that servers a purpose and reading or watching something is that sometimes I just need the tool to get out of my way. The added depth available takes up time and mental resources to learn, potentially far in excess of how regularly I use the tool or how critically I need that functionality.

This is especially true where there's a selection of tools that all work within a similar "market" space for X functionality - some are going to be less complex, some more intuitive for me personally, some fulfill a one-off need and their other functionality is duplicated by another system I know much better (and have put the time into learning).

That shouldn't take away from your enjoyment of reading through minutiae of various systems, but you can't reasonably expect a lot of people to match your enthusiasm for doing so.

Meh. My tools aren't written for other people. They can do whatever they want. But bitching about how a tool they didn't pay for that was originally written for a different purpose requires them to read something is kind of funny.

It's entirely possible there's a community of people out there who don't mind reading about abstractions exported by tools. Especially if it means you'll save time in the long run.

The original post seemed to me to be a mild introduction to the concept that maybe you can get your job done more efficiently with extensible / combinable simple tools rather than waiting for someone to add a feature to Gnome.

It's totally fine if you don't want to do that. If the person in the cube next to you does that, how is this an insult to you?

Maybe you're not the audience for this article.

I prefer both. I like when something works right out of the box. And then I read the manual and find out about all those moments when the defaults were saving me from my foolishness. And then I can unfold the thing's real potential.

Not op but I’ve got 25 years of daily computer use under me. I’m tired of having to read manuals and try to figure out the latest popular app. Software should be better these days, more intuitive to use, more interoperability between applications, data should be even more portable.

This isn’t to say I don’t enjoy learning apps anymore, I just have to guard my time more wisely.

Congratulations on the daily use. I think the point of this article is exactly the opposite of having to read new things every day. It is to read the old things once with the insinuation that the skill will translate to other things as well.

I've used Emacs for 30 years, and still do. I know enough keybindings to get by. Under the hood, everything I do is Linux. I love i3wm and Pop Shell. My main rig uses Linux as the base os (which is mostly used as a hypervisor, for security reasons).

Still, as front-end, I find myself mostly working through Windows VM's, on top of KVM. I pass through most of the hardware (usb controller, the main GPU), and get a native windows experience. Most work happens through Chrome, VS Code and Windows Termnal. I use PowerToys to get some kind of tiling window manager.

There are two reasons for this. One is that my workplace has security setups for Windows and Mac, but not linux. I can grant them full control over the VM I use for work, and can use the standard setup for vpn, etc, as well as ability to use all corporate resources with no pain or hacks.

The other is that a lot of things are more painless. Working across multiple monitors requires less setup, scaling 4k "just works", most hardware and front-end software I use behave as-well or better than on linux (particularly webcam + Teams, that I depend on heavily).

Also, as editor, I still use emacs for some things, when I just want to edit some file in a random directory.

But for development work, I've switched to VS code. The ability to ssh into any remote VM, container, etc, and be able to work directly on a remote server, complete with gitlens, debuggers, file browsers, shell access, etc, is a killer app for me. I also love the pane management, which provides the tiling window manager functionality I love from i3/Pop Shell. (Power Toys is ok, but not quite as efficient for me as i3.) Emacs CAN be set up to do those things, but that takes more effort, I feel, and still not the same user experience.

+1. I use emacs as my daily driver. Plus some tools I wrote myself to evaluate C# and JavaScript. Every now and again I hop over to a jetbrains tool.

I was slightly disappointed by i3 and xmonad and built my own WM in lisp.

In the old days we used to write our own tools by extending and combining existing tools. Now people seem to just download python and use whatever library comes with it after a day or two of experimentation

Or you could read the manual for one app and use it for a long time.

It’s oddly sad to me that you (and so many others) see interfaces adopting more intuitive and inclusive idioms and reducing the barriers to entry while becoming even more powerful as regressions.

Software is interesting and fun in different ways to different people; most people find tools accomplishing their goals with as little overhead as possible far more fun than reading technical docs. Not saying you’re wrong to like reading manuals, but it’s not a moral failure to have a cognitive profile, or heck, even a schedule that’s meaningfully hindered by your daily toolbox having reading prerequisites.

They are not inclusive. They are dumbed down, and by dumbing them down they are less powerful. Or at least less powerful than they could have been.

I don't get that allergy to reading. For a long time it has been accepted that you had to read a manual before using something relatively complex, as not everything can be expressed intuitively, and because intuition is not really universal.

The dumbing down has other reasons, mainly mobile ones. It's sad to see.

Fwiw, there's a pretty decent discussion of the "expert/trained" interface vs the "beginner" interface in Theirry Bardini's book about Douglas Engelbart.

Engelbart assumed people wanted an interface for "trained experts" -- the idea being computer applications were important enough for people to invest time in learning new concepts to make their work more efficient.

Larry Tessler and others argued you should present a beginner's interface based on existing metaphor. Larry eventually went on to work at PARC and Apple where his work inspired (and frustrated) generations of computer users.

Boarding noted that Tessler (and others in PARC) needed a raision d'etre inside the Xerox organization and partially used "building document handling systems for secretaries" to justify their budget. Kay was well known at the time for wanting to use computers in education, so it's no wonder the interface Jobs saw at PARC was one tweaked for beginners.

Interesting anecdote, but a vast oversimplification of beginner and expert needs that was formed in the infancy of computer interfaces.

> For a long time it has been accepted that you had to read a manual before using something relatively complex, as not everything can be expressed intuitively, and because intuition is not really universal.

Do you know how many pages of software documentation the average person will read in their entire life? Zero. Why? They don't have to, and imposing that requirement yields no benefit. I'm always astonished by the perennial hubris required to glibly assert chemical engineers or historians or teachers or physicists or lawyers or doctors or librarians or literature professors (or even other developers who want to concentrate on other tasks) only avoid software tools with prerequisite study because they're too lazy to read.

The purpose of software is to help people more easily solve their problems. The bigger the barrier to entry— prerequisite study for example— the harder it is for most people. That's acceptable if the complexity is genuinely necessary to afford or augment expert usability in purpose-built software, but it almost universally compensates for ham fisted interface design. Interfaces that stop expert users from doing things efficiently and only afford beginner workflows are also poorly designed. Expert vs dumbed down is a false dichotomy.

For many FOSS developers and users, tolerating counterintuitive interfaces turned into a badge of honor, and that turned into an aesthetic preference. It's not the default state of software, it does exclude non-technologists, and is why FOSS alternatives are still alternatives for every application not specifically designed to be used by technologists, even though they're free. Every single one. Every professional photographer on the planet who relies on Adobe Photoshop would jump for joy if Gimp even came close to being a sufficient replacement. That preference is also why developers don't generally make significant interface decisions in any professionally managed software project.

Not having a technologist's domain knowledge or priorities isn't a contemptible moral failure. Having it doesn't come with design expertise— learning design takes a lot of study and practice, but I'm guessing you have other things you'd prefer to concentrate on learning. Which is fine. Just realize that assuming documentation is sufficient to guide users through software is primarily informed by your not understanding software usability rather than everybody else misgauging the importance of software documentation.

Reply on news.ycombinator.com