Tinderbox Forum

How to get PDF files inside Tinderbox?

I concur. I love more than anything how fast my project files got once I removed ALL images and other attachments.

Actually, dragging a pdf into a Tinderbox view will import the pdf. Or, dragging the pdf into a File attribute of the Displayed Attributes table will link the pdf to the note.

It’s not clear which is the desired behavior for a watched file.

1 Like

OK this is awesome.

Indeed, that was my point. Intuition, i.e. an unverified assumption, is not useful here. This is the first time anyone has mentioned synching a non-text Notes note.

It is awesome, but may lead to file bloat.

Personally, I take an alternative approach. I use DEVONThink as my file repository. I then bring my PDF annotations into TBX. See this: Tinderbox Training Video 58 - Annotating with Highlights App, Tinderbox, DEVONThink, and Zotero.

6 posts were split to a new topic: AppleScript - copying $Text between notes

Sorry, I should’ve specified. I just mean the “dragging the pdf into a File attribute of the Displayed Attributes table will link the pdf to the note” part.

I neglected to mention DevonThink. Haven’t used it in a long time.

1 Like

You shouldn’t do that! DEVONthink rules :wink:

1 Like

RE extracting the anchor URL, if you watch my annotations video you’ll see that I pull in the reference URL so that if you ever want to go back to the page an annotation was made on in a PDF you can.


I have come from other TTF apps like Roam, Logseq and Obsidian. Which allow PDF embed and annotations.

I am a #noob here so you might find my questions a bit too simple. But our tools define what and and how we think perhaps? :slight_smile:

This was good and well said.


As noted above, you can embed PDF files in Tinderbox, it is NOT for smaller project. You can also link to them. Depending on what you’re doing, however, an annotations strategy with linking back to the source document may have longer legs/more runway (choose your analogy).

What is TTF? I’m not familiar with this term in this context.

There are no simple questions. We are defined by our experience, which gives us a lens and makes us unique. If you don’t know, you don’t know. If if you’re new to all this, then there is no reason you should know. Also, if you do know, then you know. But know what? You know something that worked for your context, which may not work for someone else’s context. All we can do is share our respective life references design and hope that the pattern can generally work for someone else. Normally it does, more often on the edges it wont, but then that is were adaptation comes in.

This is most certainly the case, and I think one fo the biggest challenges for Tinderbox learning. We need to unlearn what other software has taught us. Not that the other software is wrong, but that there might be a different way, many ways, to accomplish something in Tinderbox. Being able to see a different way may unlock an insight for you, teach you a new skill, who knows? But it will certainly help you move forward.

1 Like

TTF= Tools for Thoughts?

For TTF, I’d thought True Type Font. Anyway I’m equally confused as surely the above is ‘TFT’ or ‘TfT’? Of is it an AI-indexing guess of ‘Thought Tools, For’?

The ‘Tools for Thought’ term itself is unhelpful, a ‘new’ apps like Roam and Obsidian (I don’t know LogSeq) are nothing new but a re-heat for new generations of ideas that have been around since the 1970s-1980s though (presumably for money making reasons) they prefer not to recognise prior art.

The term ‘Tools for Thought’ also seems self-referential as a terms is it seems a label for ‘cool new stuff’. The Mac ecosystem if full of lots of really interesting, deep tools—normally from small/indie developers, at aren’t just doing the underpants gnomes business plan: collect off the things → [undefined magic phase] → success (and riches!). Actually collecting more stuff generally makes more noise, and automatic linkage based just on (possibly ambiguous) world strings. DEVONthink strikes me as a much better example of data collection/curation and plays well with real ‘TfT’ apps like Tinderbox to explore that content.

PDFs are another bad human choice with consequences. They solved print fidelity (recall trying to print someone else’s document pre PDF?) and having a fixed copy of record (albeit much of that is paywalled Adobe ‘secret sauce’). Getting the ‘text’ (i.e. not all the chrome, page numbers, copyright notices) of a PDF. The result is sub-par, even for digitally born (‘printed’) documents. The latter does matter as that mess is what is feed (often unfiltered) in to ‘AI’. No wonder the results are not all we’d hoped for.

The joy of Tinderbox, vs these ‘dump it all in the hopper’ tools is that I—as a human—still have a good degree of control and can make the software work for me rather than me for it.

I’m a little more sympathetic than Mark Anderson to what he calls the “new apps like Roam and Obsidian.” Sure, they’re not unprecedented and their rhetoric is sometimes overblown, but they are trying. We had a few years when too few people were even doing that.

Capturing notes is a necessary first step; a good part of The Tinderbox Way is, after all, simply urging people to make and keep notes. (This was truer of the first edition than the third, but it’s still an iterated and useful message.)

1 Like

:clap: :clap: :clap:

1 Like

good point and true in many aspects - one thing was no part of any solution we had in the last century: it’s called collaboration. That’s at least one strong feature of Roam. The next new part is the user interface. It is not about a better usability but following current trends and expectations. Roam feels a little bit like a simple HyperText engine running on the NeXT in an office at the Cern :wink: But it’s nice that PKM is becoming very popular and that we can expect many good ideas out of this trend.
So said that Apple abandoned HyperCard a long time ago - but that’s a different story.

But do we collaborate now, or merely shout ‘mine’ at each other? My own doctoral research into Wikipedia shows, sadly, that this isn’t really the case—contrary both to intention and expressed aim.

All too often ‘collaboration’ is more like taking a local bar fight and expanding to global scale. Little meeting of minds and lots of shouting at/past each other. Exceptions abound, but not enough to count as the norm. A reason I’ve stayed supporting this community for some 16 years is that it is a place where generally people are looking to learn, share and improve.

I think what has changed in the last 30-odd years is access. Both access to the work of others and our own ability to share.

Hypercard is interesting though arguably quite late to the game. Take a look at Ted Nelson’s pitch to publisher’s Harcourt Brace in 1966 (c.yrs pre Hypercard) to look at some richer ideas. The pitch came to nought but did surface in '67-'68 in Nelson’s collaborations with Andy van Dam to create HES.

Newer is not necessarily better but breathy messianic announcements of many new systems are often describing old wine in new skins. Eye candy UI is common (though Roam goes the eye-searingly ugly route for reasons unclear) but less clear is what is added beyond internet/web access. Not is older better for being that. What’s harder to spot is where improvements (other than in UI/UX terms is being made).

Microcosm, birthed contemporaneously with but separate from the web is interesting. It’s ‘mimics’ (see p.12)—IOW saved trails—and its configurable linkbase stack are ideas ripe for renewal.

So, lots of good stuff hiding in plain sight!

†. Hypertext Editing System. Ted’s Implementation notes can be read here and my clean-up digital version (without images) is here.

‡. Though I saved the app from likely extinction is viewable form, I’m not part of the team that built-it. I just got bored of being taught about it and being unable to see it in use.

1 Like

@mwra wow - that’s a lot of great stuff linked here :wink:
The amount of “newness” is limited and some parts look nice and innovative - but the practical use is often very limited (the Graph view in Roam looks nice but it seems to me like the equivalent to “tag clouds”, very popular until two years ago and useless most of the time).

It is not only the development of new software but the acceptance of those tools that opens the door for collaboration (and I do agree that we are collaborating here using a very old school technology). Still I find my self in trouble to organise my ideas as associative notes. I can’t imagine working with the association structure of someone else - tags are the maximum I can think of. Maybe it is a slow evolution we are looking at - for sure no revolution (like OOP programming in the 70 or HyperText as concept even earlier).

I’m happy with Tinderbox because of the forum here and the help I got - a small step for mankind but a giant leap for myself :slight_smile:


Yes, seems odd. But if we made a little more effort to record—for ourselves—why we associated some things, be it spatial on a map or via links in a more graph-like sense, then both others and computer (AI) assistance would have something to work with. We seem stuck in a world of “guess-the-right-keyword” and word concurrence/distance analysis which has variable per-language performance. Also this word-distance analysis would appear to work better with limited vocabularies (few pairs to measure and alternates to link) which perhaps bodes ill for expressive text in the future.

1 Like

What is “NeXT” and “PKM”?