I've been using Roam daily for the last month. What I like about Roam is the markdown, the ease of creating bidirectional links, and the automatic daily notes. The daily notes feature particularly eliminates friction for me. I just start writing.
What I don't like about Roam is its slow load time. Also writing does not feel snappy enough - if that makes any sense. I feel a slight delay at times between typing a key and seeing the character on screen. Every time that happens I like Roam a little less. I've never experienced that with an app like iaWriter, and as a result, even for all its lack of features, every time I use iaWriter I like it a little more.
I think what Roam is missing for me, besides improving the writing performance, is a weekly email digest somehow summarizing my notes. This could use the bidirectional links and give me a quick overview of what I wrote for the week. That would help improve my memory and be more introspective.
Every time I read about Roam it sounds like the same note-taking system I've built for myself. I just use a couple of Bash scripts with light text editor integration on top. (The Bash is so I can quickly add the same features to whatever text editor I'm using, today that's VSCode and Vim, but I setup Emacs at one point too. It’s mainly just keyboard shortcuts that call the Bash scripts.)
The basic features are taking a default Markdown mode and adding:
1. Follow a link to a relative file at the cursor
2. Make a link to a relative file out of the selected text (e.g., "my big idea" becomes `[my big idea](my-big-idea.md)`
Daily notes is just open a text file with the current date. (This doesn't include back links or transclusion, but I don't think those are make or break features either?)
All in all this is pretty trivial to add to an existing text editor, and if you do, you get all the power of that text editor: E.g., the best search, comparison, and split view systems anywhere. Plus the longevity of plain text.
I can see the appeal of Roam, especially for audiences that aren't already using an extensible text editors. But for people that do already use extensible text editors, I'm surprised there's not more discussion about how easy these features are to add.
> This doesn't include back links or transclusion, but I don't think those are make or break features either?
I thought this at first as well, but automatic backlinks make a bigger difference than you may guess. It basically turns each document into two things: 1. the document itself, 2. an automatic query of all references to this document that renders automatically as soon as the document is opened. Just like you can have a document with content but isn't referenced anywhere, you can also have a document with no content whose purpose is only to be the place where everything is linked back to, and persistent filtering per document means that you can make it behave exactly how you want. There is no difference between the concept of a saved search and a document, they have been unified.
The biggest value proposition to me is that you maintain each document by itself, only changing things in one place, and other documents/views automatically reflect that. Automatic backlinks with persistent per-document backlink filtering is one way Roam enables this pseudo-DRY thing where you only edit things in one place.
The other big thing it does to enable you to avoid repeating yourself is document embedding: where one line of a document or whole sub-document is referenced and looks like it's embedded in the middle of some other document. Interacting with an embedded document works the same as editing any section (you can still tell it's linked, there are read only modes etc) and the edits actually go through into the original document. Combined with fine-grained sub-document references (e.g. a paragraph or a single todo checkbox) this allows you to create things like temporary daily task lists where when you check off items they actually get marked completed in their related project where you created them, and you don't have to manually synchronize any edits. It's like you get to have your cake and eat it too. For example: do you want to organize your tasks along with the project they're related to, or do you want to organize them in the order you want to do them today? YES.
Not having to repeat myself is by far the best feature for me, and back-references combined with sub-document embedding both enable that to a very large degree. I despise having to maintain the same collection or list in multiple places, or the friction of renaming a document because now you have to fix all the references to it, or even keeping the name of a file in sync with the document header. These are things that I find pure files on disk grates on me enough to abandon the affair.
Thanks for this great explanation! This definitely helps me see the advantages of those specific features. Just to be clear, I think Roam is a great and interesting product, I didn’t mean to diminish it. Frankly, I’m just thankful they’ve rejuvenated the note-taking conversation.
But there’s a separate point I’m still trying to make: What surprises me is that there’s not a separate, parallel conversation happening about the features that can easily by applied to plain text. E.g., the fact that default Markdown supports simple Wiki links like `[A Link](a-link.md)` appears to be lost on an extraordinary number of users of Markdown. These days I come across someone asking for “Markdown with Wiki support” literally almost daily, now that Roam has popularized the feature.
To put it another way: If people were already managing their notes in Markdown Wikis, and then Roam came along and added some additional valuable features that are more difficult to add to plain text, like back links and transclusion, then that would make sense. But what we have is people lamenting that Markdown doesn’t support Wiki links, when in reality, it already has its own form of them...
This conversation is absolutely happening. Athens, for instance, is a clone of roam that is open source. It reminds me, however, of the conversation that happened about IRC vs Slack.
> I just use a couple of Bash scripts with light text editor integration on top.
That's where I started. It evolved into a simple CGI program that runs on localhost. That lets me view the markdown files in the browser. When I want to edit/create files, they open in a text editor.
Something that's different is that everything is stored in the tickets database of a Fossil repo (works with the default tickets database of any Fossil repo). That gives me for free (a) a full sqlite database for things like tag querying, and (b) version control and syncing.
The underlying graph and automatic bidirectional linking are not "trivial to add to an existing text editor". Roam is fundamentally different than other note-taking apps. The interface and editing experience are of secondary importance. If backlinks and transclusion strike you as irrelevant, you haven't even scratched the surface w Roam and, respectfully, aren't in a position to dismiss it.
I agree. A note taking app should be high performance and low friction for entering notes, this is one of the key uses of this type of app. If you are walking down a hallway with a coworker and they mention something you'd like to take a note for, the app you use should be fast enough that you can quickly jot that down. If you feel the need to hold the data in your mind or write it on paper because your app is too hard to use, then it's not fast enough.
However, entering notes should also include enough markup to allow quick recall of data at a later date and hopefully allow some intelligence in grouping and analyzing the data (much like your request for digest summaries).
If you haven't already, please send me an e-mail so I can send you the info for the alpha, it's a good balance of what you mentioned!
Love Roam, but I would also love for it to be more like a Wiki. Why does everything have to be a bullet point? I want to save links, I want to write tutorials for myself etc. It's way to limiting to always have to do bullet points.
It embraces the idea that everything is really a graph. Being explicit about the fact that a paragraph is a node allows you do node-level links (called block links in Roam).
I've wondered recently if this simple (but powerful) feature would be enough of a differentiator for a search engine challenger to gain serious market share.
How are you implementing it in practice? I would like to implement it myself but at this point it requires a fair amount of effort to do so. Are you aware of any retail solutions that are similar to buying into an index - some sort of dragon portfolio index?
> Most international trade was historically priced in dollars for two reasons. The second was, until recently, the unique speed and breadth of the Fedwire system.
> The first is the American consumer. When Americans buy, we spend dollars. This puts dollars in vendors’ hands. Those dollars can be reused for trade or invested, the latter supporting dollar financial markets. Both support dollar hegemony, which in turn drives its dominance in global trade and finance.
That has certainly not been my takeaway from history (though those could be more minor and related points)
My understanding is that most international trade was historically priced in dollars because the dollar became the world's reserve currency after Bretton Woods. Why it became the international currency is more complex, but to the best of my knowledge it is because:
1. The US was in strong monetary shape compared to the rest of the world after WWII
2. The US held many foreign assets after that war (was a creditor to the world)
3. The US held most of the world's gold after that war
I'd add that US political system is much more stable and trustworthy unlike countries without democracies and/or corruption perceptions. Presidential powers are far more limited, well scrutinized and things can easily be challenged in courts. If I held US dollars today I would be relatively more certain that it will retain value than other currencies. If all else fails one can count on US to exercise its military might to protect the dollar.
The US's war-might is limited. Look how poorly Afghanistan and Iraq went - the US's last two tests were seemingly ineffective. Syria was another trainwreck.
Internal polarization seems to be very high. Trump has eliminated or not replaced key diplomatic positions. The secretary of the navy was never confirmed and has just resigned. The defense secretary stepped down. A captain of a capital ship was ousted publicly to cheers of support from his crew.
NATO is a trainwreck. France and Germany have expressed desires to move away from US dependence while Trump demands more money.
The US's soft and hard power, to me, is just not fully there anymore.
The war in Iraq and Afghanistan were won almost as quickly as they began. It was the "peace" that was never won.
...but winning the "peace" isn't relevant for USD to be dominant. For that, the US just needs to be able to destroy enemy armies - which it can do globally with impunity.
That simply isn't true - the amount of cost sunk into not achieving objectives was enormous. Destroying the Iraqi army was never going to be the hard part. And how you can say we "won" in afghanistan, even though the taliban still exist and will have a part of the future there, is just...wow.
> ...but winning the "peace" isn't relevant for USD to be dominant. For that, the US just needs to be able to destroy enemy armies - which it can do globally with impunity.
Comments like these make me want to ask people who make them: how did empires ever fall, then?
The US army isn't a perpetual motion machine. It requires soldiers, equipment, and technology.
* 71% of young people between 17-24 are unfit to serve, 31% of that is due to obesity alone
* Equipment requires a strong manufacturing base, much of which has been decimated since World War 2, the skill required for that workforce simply doesn't exist.
* Cost. US military equipment costs are extremely high. The cost proposition is almost always a losing one for the US in any long term war.
Greater generals have overcome greater differences in strength than between the US and China. Pride comes before the fall.
>That simply isn't true - the amount of cost sunk into not achieving objectives was enormous.
The Afghanistan war is a different story, but the iraq war was won though in messy fashion. US achieved every objective it set out when it first initiated the war. Iraqis now also have a democracy and levels of violence in Iraq are at the lowest levels in a long time. Honestly, looking into the Iraq war doesn't demonstrate the US militaries inefficiency. It shows how far ahead it is of every other country on earth. No other nation would even be able to bring it's military over to the gulf. They just don't have the logistic capability.
The debate should not be on the efficacy of the US military, but on how to use it and I don't think congress has had such a debate since the mid 20th century. They need to pull back the military and then reassess what should be the next steps.
It remains to be seen whether the US objectives in Iraq were successful or not - the recent tiff b/w Trump and Iran shows that many Iraqi's are not pro-USA. Violence may be down, but at extreme cost in both dollars and men, and the problem really never left - it simply shifted to Syria. Not to mention the abandonment of key allies in Iraq, the Kurds. The fact is, US objectives are hard to quantify in Iraq, because people keep changing them to make the US look good.
You're overrating the "logistic capability" of the US here. The US used its network of allies for staging into these regions. Without that network, it would've been much more costly and difficult. On top of that, Iraq had no realistic threat to US dominance at sea. In a war against a real state actor (like China), the US would not be able to operate with impunity. Simply knocking out 2 or 3 capital ships would devastate the US capability to project force.
> and the problem really never left - it simply shifted to Syria.
I think the reason that some say that the Iraq war was never won is because they keep moving the goalposts from the stated mission during 2003.
Make Iraqi military surrender check
Overthrow Saddam Hussein check
Eliminate al-qaeda leadership in Iraq check
Establish new leader check
Establish a democratic system check
Levels of sectarian violence are also low nowadays as a bonus. Did this all take longer than estimated. Yes, but all are objectively completed. The debate is not whether the war was won, but whether it was worth winning which is kind of what you are starting to get into. Anyway, nothing much more to be done in Iraq. It is up to the Iraqi people to now decide their destiny for good or ill.
>You're overrating the "logistic capability" of the US here. The US used its network of allies for staging into these regions. Without that network, it would've been much more costly and difficult.
A network of bases in ally regions is a part of logistical capability.
>In a war against a real state actor (like China), the US would not be able to operate with impunity. Simply knocking out 2 or 3 capital ships would devastate the US capability to project force.
They could, but China has never actually demonstrated capability of doing so. The only time we have seen China in a modern fight was during the Sino-Vietnamese War. They did not fair too well with their next door neighbor. They also haven't been able to test any of their technology since then while the US at least has been able to test some of theirs except their newest big tech.
Also, I don't see US using a fullout assault on China, but more so an embargo meant to starve the China of food and energy imports. China in it's current state would not be able to deal with that.
There is no clear benefit to war anyway so I don't see China and the US getting into a direct war. China fighting with one of it's neighbors and the US helping with supplies, resources, and money is the much more likely scenario.
1-3 are the same thing. The US held most of the world's assets (was a creditor during the war), so became the world's most well capitalized bank. In that capacity, the bank's notes, i.e. dollars, became the world's currency.
Despite a run on the "bank" in the 70's, this arrangement remained.
i keep a www group (~folder) and add web archives and bookmarks to it. it's become a local internet for me.
i first search there for anything i found useful or interesting in the past. that cuts out a lot of the noise i see on google. if i can't find it in devonthink, then i visit google.
i also use devonthink with iawriter to store notes and loose thoughts, but i keep those in a separate group. [2]
The story would be less thrilling, I think: someone bought around 35 BTC, and then later he desperately wanted access to them, because they were worth 35 BTC!
> If you have cash going into this downturn, now's the time to start plotting where you are going to invest it for maximal gains
If much of this money is effectively going into the stock market - either via credit to public corporations or direct stock purchases by the Fed - is that most likely where maximal gains might occur?
i'm working on an auth tool for developers. it's a crowded space. i'm differentiating my tool by focusing very very strongly on convention over configuration. it won't be nearly as configurable as services like auth0 but it should be much more approachable for early-in-their-careers developers.
https://www.youtube.com/watch?v=OMVfrvsIfAo&feature=youtu.be