This is great. I want this but for much more. I want it to also be a nextcloud and zotero replacement, storing all my documents and books and documenting when I added, opened, edited them. I want it to store all notes that I write. I want it to record and display all browser tabs I open, when I do so, everything I copy and paste, every key I press. I want a record of everything I do in the digital world that is searchable and that can answer the question: "what was I working on 2 weeks ago on this day?" and bring back all the context also.
For obvious reasons this has to be self hosted and managed. I'm not interested in creating surveillance software or technology.
It sounds extreme but whenever I have seen peoples obsidian set ups with heaps of manual and bidirectional linking I always thought that time is the one thing we should look at. If I look up some concept on wikipedia today, there is a higher chance of me looking up related concepts or working on something related to that around this time also.
Well, I can rationally evaluate their product/idea. Their idea itself for that product is fairly interesting which is why I propose a solution one can actually host and control oneself. I don't see why I would need Microsoft to do that for me. It is I who creates the data after all.
I can never find it now, but someone had an idea for a computing system which was purely temporal for every object and then you'd only access outside of temporal by filter.
This is absolutely crazy. Every secret service on this planet will be happy! They have been working so hard to get all these informations about every person. With this software they get everything delivered FOB. The united spies will make a worldwide ad campaign for your software.
> I want it to also be a nextcloud and zotero replacement, storing all my documents and books and documenting when I added, opened, edited them. I want it to store all notes that I write.
Sounds in-scope so far. Long-term, perhaps, and maybe optional add-on features rather than built-in, but we'll see.
> I want it to record and display all browser tabs I open, when I do so, everything I copy and paste, every key I press.
That is possible in theory, but for me personally that's just too detailed. :D I wouldn't need all that granularity, myself.
But hey, the vision is pretty similar. We are generating all sorts of data to document and understand our lives -- we don't even have to deliberately write a journal -- but we have no way of comprehending it. This app is an attempt to solve that.
> That is possible in theory, but for me personally that's just too detailed. :D I wouldn't need all that granularity, myself.
I do see that. I think for that reason it would be cool to support a kind of extension system for arbitrary "collectors". And then solid filtering to filter data.
The vision is definitely similar. I am very pleasantly surprised to see your project. And I also like your ideas/roadmap on the website. I know you are building this for yourself/your family but I certainly would be open to contribute to it.
I feel like nextcloud replacement is out of scope ?
I mostly felt like Timelize was about being *behind* data-generating applications and showing and cross-referencing their data when reading the website.
I think the way is to do some sort of nextcloud extension that puts data into Timeline.
I also saw it tracks "documents" on the website but I didn't try it yet, and I would hope it can use external document sources that are already processing documents like paperless for example (which I am already using and liking)
Well, I don't plan on 1:1 feature parity with NextCloud or any comprehensive cloud suite. But I think in terms of what was mentioned: "storing all my documents and books and documenting when I added, opened, edited them. I want it to store all notes that I write," I think that's in scope.
So yes, Timelinize sits behind your current work flows. It's more of a resting place for your data (which you can still organize and curate -- more features to come in this regard), but I also can see why it might make sense to be one's primary photo library application, in the future, with more development.
As for document storage, this is still WIP, and the implementation could use more discussion to clarify the vision and specifics.
Thanks! Yes, I agree. Someone already implemented a Firefox history data source; I don't think it includes the _content_ of the pages, but that could be interesting.
It's interesting to read such a comment here. I see comments from people on this site frequently who work for Google, Facebook, OpenAI and other such companies where everyday software that doesn't respect users or their privacy is being worked on.
However, if I suggest doing similar data collection and analysis that they already do but in a way that is self hostable/governable and respects my privacy and threat model, it is bad?
This could even allow resetting the state of the computer back in time so you can pick up exactly where you left off, or undo away mistakes regardless of app, etc.
Definitely some privacy concerns but for a self ran open source thing I think it could be really cool
This looks really cool and like something I've been subconsciously looking for!
A couple thoughts & ideas:
- Given the sensitivity of the data, I would be rather scared to self-host this, unless it's a machine at home, behind a Wireguard/Tailscale setup. I would love to see this as an E2E-encrypted application, similarly to Ente.io.
- Could index and storage backend be decoupled, so that I can host my photos etc. elsewhere and, in particular, prevent data duplication? (For instance, if you already self-host Immich or Ente.io and you also set up backups, it'd be a waste to have Timelinize store a separate copy of the photos IMO.) I know, this is not entirely trivial to achieve but for viewing & interacting with different types of data there are already tons of specialized applications out there. Timelinized can't possibly replace all of them.
The problem with any other model, AFAIK, is that someone else has access to your data unless I implement an encrypted live database, like with homomorphic encryption, but even then, I'm sure at some places it would have to be decrypted in memory in places (like, transcoding videos or encoding images, for starters), and the physical owner of the machine will always have access to that.
I just don't think any other way of doing it is really feasible to truly preserve your privacy. I am likely wrong, but if so, I also imagine it's very tedious, nuanced, error-prone, and restrictive.
(Or maybe I'm just totally wrong!)
> - Could index and storage backend be decoupled, so that I can host my photos etc. elsewhere and, in particular, prevent data duplication?
I know this is contentious for some, but part of the point is to duplicate/copy your data into the timeline. It acts as a backup, and it ensures consistency, reliability, and availability.
Apps like PhotoStructure do what you describe -- and do a good job of indexing external content. I just think that's going to be hard to compel in Timelinize.
Agreed! I played with Signal exports for a while but the format changed enough that it was difficult to rely on this as a data source. Especially since it's not just obvious what changes, it's encryption so it's kind of a black box.
That said, anyone is welcome to contribute more data sources. I will even have an import API at some point, so the data sources don't have to be compiled in. Other scripts or programs could push data to Timelinize.
Just to reiterate, one of the main goals of Timelinize is to have your data. It may mean some duplication, but I'm OK with that. Storage is getting cheap enough, and even if it's expensive, it's worth it.
> I just don't think any other way of doing it is really feasible to truly preserve your privacy. I am likely wrong, but if so, I also imagine it's very tedious, nuanced, error-prone, and restrictive.
It's certainly not easy but I wouldn't go as far as saying it requires homomorphic encryption. Have you had a look at what the Ente.io people do? Even though everything is E2E-encrypted, they have (purely local) facial recognition, which to me sounds an order of magnitude harder (compute-intensive) than building a chronological index/timeline. But maybe I'm missing something here, which isn't unlikely, given that I'm not the person who just spent a decade building this very cool tool.
> It acts as a backup, and it ensures consistency, reliability, and availability.
Hmmm, according to you[0],
> Timelinize is an archival tool, not a backup utility. Please back up your timeline(s) with a proper backup tool.
;)
I get your point, though, especially when it comes to reliability & availability. Maybe the deduplication needs to happen at a different level, e.g. at the level of the file system (ZFS etc.) or at least at the level of backups (i.e. have restic/borgbackup deduplicate identical files in the backed-up data).
Then again, I can't say I have not had wet dreams once or twice of a future where apps & their persistent data simply refer to user files through their content hashes, instead of hard-coding paths & URLs. (Prime example: Why don't m3u playlist files use hashes to become resistant against file renamings? Every music player already indexes all music files, anyway. Sigh.)
> Especially since it's not just obvious what changes, it's encryption so it's kind of a black box.
Wouldn't you rather diff the data after decrypting the archive?
> Just to reiterate, one of the main goals of Timelinize is to have your data. It may mean some duplication, but I'm OK with that. Storage is getting cheap enough, and even if it's expensive, it's worth it.
I suspect it will lead to duplication of pretty much all user data (i.e. original storage requirements × 2), at least if you're serious about your timeline. However, I see your point, it might very well be a tradeoff that's worth it.
Correct, Timelinize is not a backup utility, but having the copy of your data in your timeline acts as a backup against losing access to your data sources, such as Google Photos, or your social media account(s), etc. As opposed to simply displaying data that is stored elsewhere.
But yes, I think the likes of what Ente is doing is interesting, though I don't know the technical details.
Data content hashes are pretty appealing too! But they have some drawbacks and I decided not to lean heavily on them in this application, at least for now.
I’m not sure how exactly timelinize stores photos, but you could sync photos as you take them to timelinize, and then, if they are accessible, point immich to the timelinize photos for its use. That would essentially deduplicate your photos.
This is an amazing idea but do I have to run Google takeout every time I want to update the data[0]? Unfortunately that's such a a cumbersome process that I don't think I'd use this. But if my timeline could update in near real time this would be a killer app
Yeah. Major thorn in my side. I spent hours trying to automate that process by using Chrome headless, and it kinda worked, until I realized that I needed to physically authenticate not just once, but every 10 minutes. So, it basically can't be automated since 2FA is needed so often.
In practice, I do a Takeout once or twice a year. (I recommend this even if not using Timelinize, so you can be sure to have your data.)
I thought you could set up an automatic Takeout export periodically, and choose the target to be your Google Drive. Then via a webapp oauth you could pull the data that way. Frequency was limited (looks like it says the auto export is “every 2 months for 1 year”). So hardly realtime, but seems useful and (relatively) easy? Does a method like that not work for your intentions?
Downloading from Google drive doesn't cost anything, does it? Although, I guess you would have to have enough empty space on your Google drive to be able to store the takeout zip, which I think is an acceptable cost
Some kind of companion app that runs on my phone and streams the latest data (photos, location history, texts, etc ) back to the timeline would probably be more tractable for live updates. But that is probably a wildly different scope than the import based workflow. This is very cool regardless.
About 5-6 years ago, Timelinize actually used only the Google Photos API. It didn't even support imports from Takeout yet. The problem is the API strips photos of crucial metadata including location, and gives you nerfed versions of your data. Plus the rate limits were so unbearable, I eventually ripped this out.
But yeah, an app that runs on your phone would be a nice QoL improvement.
Probably not hard. Timelinize's data sources have a standard API with just 2 methods [0], so it should be fairly trivial to implement depend on how accessible Immich is.
To clarify, you don't grant access to Google Photos, you just do the Takeout from https://takeout.google.com to download your data first.
"Because Timelinize is entity-aware, "it can project data points onto a map even without coordinate data. If a geolocated point is known for an entity around the same time of others of that entity's data points, it will appear on the map."
In the context of Timelinize, this is great! Outside of that, this sentence really drives home how much data Google could join on me -- a heavy Android/Chrome/Gmail/Maps w/ timeline user.
What are you planning with weather? Associating entities that have a known location with historical temp/forecast data?
This reminds me of a similar project called HPI [1] that I discovered in 2022, the idea to aggregate one's digital footprints into one single private place really inspires me a lot. Timelinize is almost like a GUI version of HPI, a more sophisticated user product that everyone can use. I will definitely try it, congrats and respect!
P.S. HPI has many existing exporters, so it might be helpful to extend Timelinize's capacity.
You might suggest to users the following use case. “If you want to create a Timelinize data store, but don’t feel that your own local systems are secure enough to safely hold a basket with a copy of every egg in your life, you might consider the following use case, which some of our customers implement. Once our twice a year, update the data store, but store it on an external disk. When the update is done, take the disk offline and keep it in a drawer or safe.”
Also
I always wondered how cool it would be if I could tell some Spotify-like system, “I’m 20 miles away from the lake, we are going to stay in the cabins a week, just like we did 10 years ago. Play me the exact same songs now that played at every turn back then.”
Also
For a name, how about:
ChronEngine
That name seems pretty free and clear from previous use, if you like it grab ChronEngine.com before some squatter does and thank me with a phone call, I would enjoy a quick chat with you.
Also
Your web page might benefit from a grid that lists all the input sources you accept, with hotlinks next to the names that give a pop-up quick summary of what that source is about, and maybe some color coding that shows something like “green = rock solid lately”, “yellow = some users reporting problems”, “red = they just updated their format, it’s broken, we are working on it”. You will/are facing challenges similar to Trillian, a chat client from the early 2000s that would try to maintain ongoing connection compatibility with multiple other chat clients such as AIM/ICQ/MSN. Also, the grid could have a “suggested source sets” filter that helped people find what 5 (for example) input sources they might select for their use style.
Oh and make a list of anybody that says they have done an elaborate something similar with Excel (like me and at least one other person in this thread) and maybe have a discussion with them some time, we/they might have some useful insights.
Let’s hear it for people on the opposite side of the “go fast and break things” coin! My first project took 16 years. My current one I started 28 years ago!
And one more idea from me. I could see your current system to be deliberately kept like it is, but offered as a two-part system, where the second part is, “once you get your timeline built and you have reviewed it carefully, then click here and the local-only LLM will have access to the whole thing.” The two part nature could be a big competitive advantage, helping people to carefully build a LLM system that, for instance, they could then offer to their whole family to peruse, without having to worry too much that it accidentally included information that it should not have.
> You might suggest to users the following use case. “If you want to create a Timelinize data store, but don’t feel that your own local systems are secure enough to safely hold a basket with a copy of every egg in your life, you might consider the following use case, which some of our customers implement. Once our twice a year, update the data store, but store it on an external disk. When the update is done, take the disk offline and keep it in a drawer or safe.”
Sure, I like that.
> I always wondered how cool it would be if I could tell some Spotify-like system, “I’m 20 miles away from the lake, we are going to stay in the cabins a week, just like we did 10 years ago. Play me the exact same songs now that played at every turn back then.”
It's inevitable for an LLM and other tooling like personal assistants to be integrated to this thing, so yeah that sounds like a great use case.
> For a name, how about: ChronEngine That name seems pretty free and clear from previous use, if you like it grab ChronEngine.com before some squatter does and thank me with a phone call, I would enjoy a quick chat with you. ...
Oh and make a list of anybody that says they have done an elaborate something similar with Excel (like me and at least one other person in this thread) and maybe have a discussion with them some time, we/they might have some useful insights.
Not bad actually; and sure, we can chat either way. Feel free to book a time at https://matt.chat.
> Your web page might benefit from a grid that lists all the input sources you accept, with hotlinks next to the names that give a pop-up quick summary of what that source is about, and maybe some color coding that shows something like “green = rock solid lately”, “yellow = some users reporting problems”, “red = they just updated their format, it’s broken, we are working on it”. You will/are facing challenges similar to Trillian, a chat client from the early 2000s that would try to maintain ongoing connection compatibility with multiple other chat clients such as AIM/ICQ/MSN. Also, the grid could have a “suggested source sets” filter that helped people find what 5 (for example) input sources they might select for their use style.
I LOVED Trillian, thanks or the nostalgia. Oh man, they're still alive: https://trillian.im/ (I love the icon...)
A few latin names have been suggested, but nothing has stuck. The problem is they are usually difficult to spell and pronounce, which isn't really an improvement over the current situation :)
FindPenguins is cool! I don't use it myself, but anyone is welcome to implement a data source for it.
This reminds me of Perkeep, but I understand this is more focused on the presentation of the data, while Perkeep was on the data storage.
But maybe they could be integrated, or at least support each other's formats.
Hey - this is awesome. I've been working on a small local app like this to import financial data and present a dashboard, for the family to use together (wife and I). So yeah - great work here, taking control of your data.
I'm curious about real-time data, or cron jobs, though. I love the idea of importing my data into this, but it would be nicer if I could set it up to automatically poll for new data somehow. Does Timelineize do something like that? I didn't see on the page.
Cool, yeah, the finance use case seems very relevant. Someday it'd be cool to have a Finance exploration page, like we do for other kinds of data.
Real-time/polling imports aren't yet supported, but that's not too difficult once we land on the right design for that feature.
I tinkered with a "drop zone" where you could designate a folder that, when you add files to it, Timelinize immediately imports it (then deletes the file from the drop zone).
I have something set up at home: ~/Inbox and ~/Outbox. Anything dropped into ~/Outbox gets rsync'd to rsync.net and mv'd (locally) to ~/Inbox.
Anything in ~/Inbox is "safe to delete" because it's guaranteed to have an off-site backup.
Presumably a fancy management app would queue (or symlink) the full directory structure into ~/Inbox (which would then behave as a "streaming cache")
~/Inbox would effectively be available (read only) on "all machines" and "for free" with near zero disk space until you start accessing or pulling down files.
I use Dropbox to manage ~/Sync (aka: active, not "dead" files).
"Outbox", "Inbox", and "Sync" have been the "collaboration names" that resonated the most with me (along with ~/Public if you're old enough, and ~/Documents for stuff that's currently purely local)
Cool idea. Thanks for sharing. I was really annoyed by the way Google nerfed the maps timeline stuff last year. Obviously this project is way more ambitious than that, but just goes to show you how little Google cares about the longevity of your data.
Nice work. I’ve been frustrated with how closed off location history tools have become lately. This looks like a solid step toward giving people real ownership of their data again. Definitely checking this out.
It was a different era 15 years ago, made some decent inroads on an API based approach with the Locker Project https://en.wikipedia.org/wiki/Locker_(software) but the platforms quickly started gimping the APIs and exports/takeouts were non-existent or too early yet back then, so ran out of steam.
Yep, I encountered the same problem. Initially it was solely API-based, but as they got locked down and paid-only and removed entirely, I went the bulk download route.
Looks awesome! Btw, it would be nice if it can also support audio files, I frequently make multi hour long recordings (meetings, presentation etc) and it would be very useful to see pictures and other stuff from a specific time together with the audio from that timing.
This seems like the perfect thing to mix with financial records (ie bank feeds) and a local LLM.
I’m not sure what you’d use it for … exactly … but it could probably reconcile and figure out all your credit card charges based on your message history and location, allocate charges to budgets, and show more analytics than you’re probably interested in knowing.
People with a cloud connected car such as a Tesla could probably get some real “personal assistant” type use cases out of this, such as automatically sorting your personal and business travel kms, expenses and such for tax purposes.
There’s probably other use cases like suggesting local experiences you haven’t done before, and helping you with time management.
I really like the local storage of this. Files and folders are the best!
(When noodling on this, I’ve also been wondering about putting metadata for files in sidecar files next to the files they describe, rather than a centralized SQLite database. Did you experiment with anything like that by any chance?)
Why sidecar metadata files? In general I've tried to minimize the number of files on disk since that makes copying slow and error-prone. (A future version of Timelinize will likely support storing ALL the data in a DB to make for faster, easier copying.) We'd still need a DB for the index anyway, which essentially becomes a copy of the metadata.
Don’t know if you have seen the work DuckDB is doing on ducklake. Maybe there is an overlap in vision for versioning data across multiple data sources - and similar to SQLite it’s not proprietary and easily drilled down on. I’m sorry, don’t have technical knowledge :/
Thank you so much for this! I've also had the idea but, contrary to you, never had the courage nor energy to actually do it. I have some of my old devices backed up with the very first text my (now) wife and I exchanged, and this feels like the perfect use case.
So thank you so much, I now have another project to host on my home server, and a meaningful one.
>(PS. I'm open to changing the name. Never really liked this one...)
Timelinize is great, it explains what the product does
started with overland (was great, but hurt the battery a lot) - there was an iOS app for it. I then just moved on to PSObserver as it was self contained, I could export off my phone. I also augmented with running google maps background for a few years as well. Think I have over 5m data points now. I user Kepler.gl originally + with iPhone photos as well overlayed.
Interesting. But this seems to be limited to the supported data-sources? No generic format to build your own sources? Also, why the focus on a timeline? Not all personal data or tasks are time-related. So is this more about personal data, or a timeview on personal data?
If you want to have collaboration on this, maybe think about splitting it into backend and frontends. With the backend being the storage, and it's import-mechanism, and the frontend being open for anyone to put their own view onto the data, and maybe work with them.
Maintaining personal data has demand here, but so far there is no real unified project catering to a broad range of services and different types of data. So there is a space to fill and a community to build IMHO.
I'm designing an import API that will let any script or program push data to the timeline.
> Also, why the focus on a timeline?
Seemed to make the most sense.
> Not all personal data or tasks are time-related.
Probably true. Do you have examples of what you're thinking of?
> If you want to have collaboration on this, maybe think about splitting it into backend and frontends. With the backend being the storage, and it's import-mechanism, and the frontend being open for anyone to put their own view onto the data, and maybe work with them.
That's how it is right now; it's a client-server architecture. It has a CLI and a JSON API so you can build your own front-ends.
> Maintaining personal data has demand here, but so far there is no real unified project catering to a broad range of services and different types of data. So there is a space to fill and a community to build IMHO.
I agree, this is a huge void that needs to be filled!
> Probably true. Do you have examples of what you're thinking of?
Bookmarks, contacts, notes for example, they are usually not time-based but by context, category or whichever organization their usage demands at the moment. They can be time-based in a journal, but usually I do not remember websites or people by the time I encountered them, so I would not search them this way.
The other question is how personal is personal in this project? Are my IMDB-ratings valid data for this? My Ebook-collection? My Steam-account? Those are usually things I would not manage in a timeline or a map, but with different interfaces and features.
Contacts are already handled as first-class concepts in Timelinize, known as "entities". Entities don't have timestamps, but they have a name and attributes that can be anything, and the relationship between the entity and its attributes can be bounded by time.
Notes are also created and updated/changed at specific times.
> Are my IMDB-ratings valid data for this? My Ebook-collection? My Steam-account? Those are usually things I would not manage in a timeline or a map, but with different interfaces and features.
Could be, if you wanted to timeline those! (Like, when you rated a certain movie. Or when you got or read a certain book. Or when you got a new Steam game.)
Like others I really like the idea and the realisation looks great too!
I might not be the typical user for this, because I'd prefer my data to actually stay in the cloud where it is, but I'd still like to have it indexed and timelined. Can timelinize do this? Like, instead of downloading everything from gphoto, youtube, bluesky, wtv, just index what's there and offer the same interface? And only optionnaly download the actual data in addition to the meta-data?
That's not really aligned with my vision/goals, which is to bring my data home; but to be clear, downloading your data doesn't mean it has to leave the cloud. You can have your cake and eat it too.
The debate between importing he data and indexing external data is a long, grueling one, but ultimately I cannot be satisfied with not having my data guaranteed to be locally available.
I suppose in the future it's possible we could add the ability to index external data, but this likely wouldn't work well in practice since most data sources lock down real-time access via their API restrictions.
A live demo would be great, but I'm not sure how to generate the fake data in a way that imitates real data patterns. That's originally how I wanted to demo things, but the results weren't compelling. (It was a half-hearted effort, I admit.) So I switched to obfuscating real data.
FreeMyCash looks great! Yours is the second financial application I've heard of; maybe we need to look at adding some finance features soon.
I had this same idea for a long time. Even took github.com/center for it (I've since changed how it's being used). Cool to see someone actually achieve it, well done.
Wow that's great! Interesting if it's possible to use not just a folder but like a s3-compatible backend for photos and for db backups as well
(I don't think all my photo/video archives would fit on my laptop, though the thumbnails definitely would, while minio or something replicated between my desktop plus a backup machine at Hetzner or something would definitely do the thing)
I don't think sqlite runs very well on S3 file systems. I think it would also be insufferably slow.
I even encountered crashes within sqlite when using ExFAT -- so file system choice is definitely important! (I've since implemented a workaround for this bug by detecting exfat and configuring sqlite to avoid WAL mode, so it's just... much slower.)
Super interested in this as well (and thank you for Caddy)
How does this handle data updating / fixing? My use case is importing data that's semi structured. Say you get data from a 3rd party provider from one dump, and it's for an event called "jog". Then they update their data dump format so "jog" becomes subdivided into "light run" vs "intense walk", and they also applied it retroactively. In this case you'd have to reimport a load of overlapping data.
I saw the FAQ and it only talks about imports not strictly additive.
I am dealing with similar use cases of evolving data and don't want to deal with SQL updating, and end up working entirely in plain text. One advantage is that you can use git to enable time traveling (for a single user it still works reasonably).
In the advanced import settings, you can customize what makes an item unique or a duplicate. You can also configure how to handle duplicates. By default, duplicates are skipped. But they can also be updated, and you can customize what gets updated and which of the two values to keep.
But yes, updates do run an UPDATE query, so they're irreversible. I explored schemas that were purely additive, so that you could traverse through mutations of the timeline, but this got messy real fast, and made exploring (reading) the timeline more complex/slow/error-prone. I do think it would be cool though, and I may still revisit that, because I think it could be quite beneficial.
Have you heard of XTDB / Bitemporality? The basic idea is to make time 2-dimensional, where each record has both a System Time range and a Valid Time range. Designed as a write-only db with full auditability for compliance purposes.
With 2D time you can ask complex questions about what you knew when, with simpler questions automatically extended into a question about the current time. Like:
"What is the price?" -> "What is the price today, as of today?"
"What was the price in 2022" -> "What was the price in 2022, as of today?"
"What was the price in 2022, as of 2023?"
You probably don't want to just switch to XTDB, but if you pursue this idea I think you should look into 2D time as I think it is schematically the correct conceptualization for this problem.
Yeah, I did actually pursue this for a time (heh), but I might revisit it later. It was too much complexity for debateable value-add, though the value is growing on me.
Thanks for the reply! I'll have to try this out... it almost looks like what perkeep was meant to become.
One interesting scenario re time traveling is if we use an LLM somewhere in data derivation. Say there's a secondary processor of e.g. journal notes that yield one kind of feature extraction, but the model gets updated at some point, then the output possibilities expand very quickly. We might also allow human intervention/correction, which should take priority and resist overwrites. Assuming we're caching these data then they'll also land somewhere in the database and unless provenance is first class, they'll appear just as ground truth as any other.
Bitemporal databases look interesting but the amount of scaffolding above sqlite makes the data harder to manage.
So if I keep ground truth data as text, looks like I'm going to have an import pipeline into timelinize, and basically ensure that there's a stable pkey (almost certainly timestamp + qualifier), and always overwrite. Seems feasible, pretty exciting!
Oh I love this oh so much. It lines up with a lot of things I've wanted or am planning. So I'm definitely going to take a dive into your source for inspiration
One thing I wish for encompasses the idea that I might want to share different slices of my life with different groups
I was looking for a "sharing idea" comment, so it's great to see something is planned.
Being able to share a time-based, geo-based, or X-based slice (all or some of) with someone would be a great way to learn about "close encounters" before or after we met, or even someone else's perspective of a shared trip or experience.
Just a general comment; you've made something truly wonderful (IMHO), and I look forward to seeing where you take this.
As for branding, IMO you could go a bunch of directions:
Timelines
Tempor (temporal)
Chronos
Chronografik
Continuum
Momentum (moments, memory, momentum through time)
IdioSync (kinda hate this one tbh)
Who knows! Those are just the ones that fell out of my mouth while typing. It's just gotta have a memorable and easy-to-pronounce cadence. Even "Memorable" is a possibility LOL
This basically brings your data from the cloud to local-first ! Kudos to your dedication and especially making this open-source for the benefit of everyone!
Oh man, this is awesome. I wanted an answer to Apple's Journal because I personally am using Obsidian (doing my best to tag and ref data). But this can be what I need to access that metadata layer and combine everything
for many years I kept Tiny Travel Tracker running on my android phone, and would periodically import it into os&m maps. It was nice to have the record of all exactly all the places I had wandered to, and not also share that with five eyes.
https://f-droid.org/en/packages/com.rareventure.gps2/
> Recommended: GPU (NVIDIA 20-series or newer; 40-series for faster thumbnails) or Apple Silicon (M2 or newer)
Very nice project, it is something what we all need right now in these days. Just noticed that you recommend Apple Silicon M2 or newer? Is there really a significant difference between M1 and M2? Is it running some LLM model in the background?
I dunno. M1 might be fine, I just haven't tested it with an M1. If you enable semantic features, it does use a large model (not necessarily an LLM) for embeddings; but regardless it will generate thumbnails for images and videos, and transcode videos for playback, so a good GPU is helpful.
Interesting; how easy is it to backup it up somewhere - yes, on a cloud for example - and then restore/sync it on another machine? Is the data format portable and easy to move like this?
Yep -- a timeline is just a folder with regular files and folders in it. They're portable across OSes. I've tried to account for differences in case-sensitive and -insensitive file systems as well. So you can copy/move them and back them up like you would any other directory.
Great idea! Do you plan to integrate with other services? I'm looking for similar solutions to collect feedback and improve my product. Are you already getting feedback from the community?
Got a lot of feedback here so far, plenty to keep me busy. And there will be an import API to let scripts/programs push data to Timelinize, even if it doesn't support those data sources built-in.
Nice one, thanks for sharing. For sure I’ll give it a try.
Have you thought of creating a setup so as to package all libraries and dependencies needed? You have a very nice installation guide, but there are many users who just want the setup.exe :-)
Thank you! Not sure I can package everything because of license requirements. IANAL. I think the container image basically automates the various dependencies, but I didn't create it and I don't use it so I'm not 100% sure.
Basically, I would love to know how to do this correctly for each platform, but I don't know how. Help would be welcomed.
Very nice project! Curiousity question: since you are taking data dumps once-twice a year, and let's say you also copy the photos as well, do you do any updates incrementally or just replace the old one with new dump?
Call it D-FUKT and the purpose would be to show people how absolutely fucked all of their data is. If you list every corp/org that has access to each piece of data, this could be cool
This is just some feedback. But when including product shots on your web site, at least don't use real user data where you have to blur most of the content. It looks tacky and unprofessional.
When announcing a product, the language, license, version number, amount of time under development since inception, and approximate number of contributors are all critical pieces of information to include in an announcement.
Ten years is a long time. Is it just you? What language is it in? Does it run on desktop or does it need a server? What is the license?
This is great. I want this but for much more. I want it to also be a nextcloud and zotero replacement, storing all my documents and books and documenting when I added, opened, edited them. I want it to store all notes that I write. I want it to record and display all browser tabs I open, when I do so, everything I copy and paste, every key I press. I want a record of everything I do in the digital world that is searchable and that can answer the question: "what was I working on 2 weeks ago on this day?" and bring back all the context also.
For obvious reasons this has to be self hosted and managed. I'm not interested in creating surveillance software or technology.
It sounds extreme but whenever I have seen peoples obsidian set ups with heaps of manual and bidirectional linking I always thought that time is the one thing we should look at. If I look up some concept on wikipedia today, there is a higher chance of me looking up related concepts or working on something related to that around this time also.
I think Microsoft has some kind of product that can help you Recall what you were working on?
Well, I can rationally evaluate their product/idea. Their idea itself for that product is fairly interesting which is why I propose a solution one can actually host and control oneself. I don't see why I would need Microsoft to do that for me. It is I who creates the data after all.
The problem with that wasn’t the idea for me, it’s that Microsoft was doing it and they’re a known bad actor
https://activitywatch.net/ gets you part of the way there.
I already use it. It's quite good but more for time tracking!
I can never find it now, but someone had an idea for a computing system which was purely temporal for every object and then you'd only access outside of temporal by filter.
I wish I could find it again.
Might you be thinking of Lifestreams by Freeman and Gelernter?
https://www.cs.yale.edu/homes/freeman/lifestreams.html
https://www.scribd.com/document/18361503/Eric-T-Freeman-The-...
Adjacent:
https://en.wikipedia.org/wiki/The_Humane_Interface#:~:text=N...
Maybe Lifestreams? https://www.cs.yale.edu/homes/freeman/ericchi96.html
Are you, by chance, referring to https://perkeep.org/ ?
This is absolutely crazy. Every secret service on this planet will be happy! They have been working so hard to get all these informations about every person. With this software they get everything delivered FOB. The united spies will make a worldwide ad campaign for your software.
No. It's not like services/products people use of big companies don't already collect to this type of level.
I'm at least proposing that this be open source, self hostable and valuable to yourself.
Also, I know my threat model. Maybe you are James Bond though.
> I want it to also be a nextcloud and zotero replacement, storing all my documents and books and documenting when I added, opened, edited them. I want it to store all notes that I write.
Sounds in-scope so far. Long-term, perhaps, and maybe optional add-on features rather than built-in, but we'll see.
> I want it to record and display all browser tabs I open, when I do so, everything I copy and paste, every key I press.
That is possible in theory, but for me personally that's just too detailed. :D I wouldn't need all that granularity, myself.
But hey, the vision is pretty similar. We are generating all sorts of data to document and understand our lives -- we don't even have to deliberately write a journal -- but we have no way of comprehending it. This app is an attempt to solve that.
> That is possible in theory, but for me personally that's just too detailed. :D I wouldn't need all that granularity, myself.
I do see that. I think for that reason it would be cool to support a kind of extension system for arbitrary "collectors". And then solid filtering to filter data.
The vision is definitely similar. I am very pleasantly surprised to see your project. And I also like your ideas/roadmap on the website. I know you are building this for yourself/your family but I certainly would be open to contribute to it.
Wonderful, I'd love to collaborate!
Dig deep on "dogsheep" from Mr Willison. We're all circling on the same orbit here.
Woah. Hadn’t seen that yet. Link: https://dogsheep.github.io/
I feel like nextcloud replacement is out of scope ?
I mostly felt like Timelize was about being *behind* data-generating applications and showing and cross-referencing their data when reading the website.
I think the way is to do some sort of nextcloud extension that puts data into Timeline.
I also saw it tracks "documents" on the website but I didn't try it yet, and I would hope it can use external document sources that are already processing documents like paperless for example (which I am already using and liking)
Well, I don't plan on 1:1 feature parity with NextCloud or any comprehensive cloud suite. But I think in terms of what was mentioned: "storing all my documents and books and documenting when I added, opened, edited them. I want it to store all notes that I write," I think that's in scope.
So yes, Timelinize sits behind your current work flows. It's more of a resting place for your data (which you can still organize and curate -- more features to come in this regard), but I also can see why it might make sense to be one's primary photo library application, in the future, with more development.
As for document storage, this is still WIP, and the implementation could use more discussion to clarify the vision and specifics.
I also should have mentioned in my comment that I only host nextcloud for myself and my family.
That may be more realistic and in scope than how nextcloud actually supports entire orgs/organizational units.
Timelinize looks rad. Congratulations.
> That is possible in theory, but for me personally that's just too detailed. :D I wouldn't need all that granularity, myself.
Think this can go quite far with just the browsing history & content of viewed webpages.
Thanks! Yes, I agree. Someone already implemented a Firefox history data source; I don't think it includes the _content_ of the pages, but that could be interesting.
"I'm not interested in creating surveillance software or technology."
That is exactly what you would be doing though
To surveil... myself?
It's interesting to read such a comment here. I see comments from people on this site frequently who work for Google, Facebook, OpenAI and other such companies where everyday software that doesn't respect users or their privacy is being worked on.
However, if I suggest doing similar data collection and analysis that they already do but in a way that is self hostable/governable and respects my privacy and threat model, it is bad?
Na you are right its not bad as a personal tool (I didnt say it was bad) but yea you spying on yourself mate :D
I think you might appreciate this talk, which uses a micro kernel to do a lot of the ideas you’re talking about.
https://www.usenix.org/conference/osdi14/technical-sessions/...
This could even allow resetting the state of the computer back in time so you can pick up exactly where you left off, or undo away mistakes regardless of app, etc.
Definitely some privacy concerns but for a self ran open source thing I think it could be really cool
Oh yeah, mholt is notable for having created Caddy (the webserver). My interest in Timelineize just went up.
Ha! I was about to commend the author for the "install" being a single .exe on Windows. This makes perfect sense now.
That's just the joy of Go. But, I do wish I could have made this one a perfectly static application. I don't know if it is possible.
This looks really cool and like something I've been subconsciously looking for!
A couple thoughts & ideas:
- Given the sensitivity of the data, I would be rather scared to self-host this, unless it's a machine at home, behind a Wireguard/Tailscale setup. I would love to see this as an E2E-encrypted application, similarly to Ente.io.
- Could index and storage backend be decoupled, so that I can host my photos etc. elsewhere and, in particular, prevent data duplication? (For instance, if you already self-host Immich or Ente.io and you also set up backups, it'd be a waste to have Timelinize store a separate copy of the photos IMO.) I know, this is not entirely trivial to achieve but for viewing & interacting with different types of data there are already tons of specialized applications out there. Timelinized can't possibly replace all of them.
- Support for importing Polarsteps trips, and for importing Signal backups (e.g. via https://github.com/bepaald/signalbackup-tools ) would be nice!
Great comment, thanks for the questions.
> unless it's a machine at home,
This is, in fact, the intended model.
The problem with any other model, AFAIK, is that someone else has access to your data unless I implement an encrypted live database, like with homomorphic encryption, but even then, I'm sure at some places it would have to be decrypted in memory in places (like, transcoding videos or encoding images, for starters), and the physical owner of the machine will always have access to that.
I just don't think any other way of doing it is really feasible to truly preserve your privacy. I am likely wrong, but if so, I also imagine it's very tedious, nuanced, error-prone, and restrictive.
(Or maybe I'm just totally wrong!)
> - Could index and storage backend be decoupled, so that I can host my photos etc. elsewhere and, in particular, prevent data duplication?
I know this is contentious for some, but part of the point is to duplicate/copy your data into the timeline. It acts as a backup, and it ensures consistency, reliability, and availability.
Apps like PhotoStructure do what you describe -- and do a good job of indexing external content. I just think that's going to be hard to compel in Timelinize.
> Support for importing Polarsteps trips, and for importing Signal backups (e.g. via https://github.com/bepaald/signalbackup-tools ) would be nice!
Agreed! I played with Signal exports for a while but the format changed enough that it was difficult to rely on this as a data source. Especially since it's not just obvious what changes, it's encryption so it's kind of a black box.
That said, anyone is welcome to contribute more data sources. I will even have an import API at some point, so the data sources don't have to be compiled in. Other scripts or programs could push data to Timelinize.
Just to reiterate, one of the main goals of Timelinize is to have your data. It may mean some duplication, but I'm OK with that. Storage is getting cheap enough, and even if it's expensive, it's worth it.
Thanks for your thoughtful response!
> I just don't think any other way of doing it is really feasible to truly preserve your privacy. I am likely wrong, but if so, I also imagine it's very tedious, nuanced, error-prone, and restrictive.
It's certainly not easy but I wouldn't go as far as saying it requires homomorphic encryption. Have you had a look at what the Ente.io people do? Even though everything is E2E-encrypted, they have (purely local) facial recognition, which to me sounds an order of magnitude harder (compute-intensive) than building a chronological index/timeline. But maybe I'm missing something here, which isn't unlikely, given that I'm not the person who just spent a decade building this very cool tool.
> It acts as a backup, and it ensures consistency, reliability, and availability.
Hmmm, according to you[0],
> Timelinize is an archival tool, not a backup utility. Please back up your timeline(s) with a proper backup tool.
;)
I get your point, though, especially when it comes to reliability & availability. Maybe the deduplication needs to happen at a different level, e.g. at the level of the file system (ZFS etc.) or at least at the level of backups (i.e. have restic/borgbackup deduplicate identical files in the backed-up data).
Then again, I can't say I have not had wet dreams once or twice of a future where apps & their persistent data simply refer to user files through their content hashes, instead of hard-coding paths & URLs. (Prime example: Why don't m3u playlist files use hashes to become resistant against file renamings? Every music player already indexes all music files, anyway. Sigh.)
> Especially since it's not just obvious what changes, it's encryption so it's kind of a black box.
Wouldn't you rather diff the data after decrypting the archive?
> Just to reiterate, one of the main goals of Timelinize is to have your data. It may mean some duplication, but I'm OK with that. Storage is getting cheap enough, and even if it's expensive, it's worth it.
I suspect it will lead to duplication of pretty much all user data (i.e. original storage requirements × 2), at least if you're serious about your timeline. However, I see your point, it might very well be a tradeoff that's worth it.
[0]: https://timelinize.com/docs/importing-data
Correct, Timelinize is not a backup utility, but having the copy of your data in your timeline acts as a backup against losing access to your data sources, such as Google Photos, or your social media account(s), etc. As opposed to simply displaying data that is stored elsewhere.
But yes, I think the likes of what Ente is doing is interesting, though I don't know the technical details.
Data content hashes are pretty appealing too! But they have some drawbacks and I decided not to lean heavily on them in this application, at least for now.
I’m not sure how exactly timelinize stores photos, but you could sync photos as you take them to timelinize, and then, if they are accessible, point immich to the timelinize photos for its use. That would essentially deduplicate your photos.
They are just organized into folders on disk, so that could definitely work!
This is an amazing idea but do I have to run Google takeout every time I want to update the data[0]? Unfortunately that's such a a cumbersome process that I don't think I'd use this. But if my timeline could update in near real time this would be a killer app
[0]: https://timelinize.com/docs/data-sources/google-photos
Yeah. Major thorn in my side. I spent hours trying to automate that process by using Chrome headless, and it kinda worked, until I realized that I needed to physically authenticate not just once, but every 10 minutes. So, it basically can't be automated since 2FA is needed so often.
In practice, I do a Takeout once or twice a year. (I recommend this even if not using Timelinize, so you can be sure to have your data.)
I thought you could set up an automatic Takeout export periodically, and choose the target to be your Google Drive. Then via a webapp oauth you could pull the data that way. Frequency was limited (looks like it says the auto export is “every 2 months for 1 year”). So hardly realtime, but seems useful and (relatively) easy? Does a method like that not work for your intentions?
Will have to look into that. Sounds like it could be expensive but maybe worth it.
You can schedule the takeout to Drive, then use a tool such as rclone (amazing tool) to pull it down.
It should not add any costs except the storage for the takeout zip on drive.
Look at supported providers in rclone and you might find easy solutions for some hard sync problems: https://rclone.org/#providers
> except the storage for the takeout zip on drive.
Yeah, that's the cost I'm talking about. It essentially amounts to paying an extra subscription to be able to download your data [on a regular basis].
I'm a big rclone fan btw :) I'm sure there's some future where we do something like this to automate Takeouts.
Downloading from Google drive doesn't cost anything, does it? Although, I guess you would have to have enough empty space on your Google drive to be able to store the takeout zip, which I think is an acceptable cost
Some kind of companion app that runs on my phone and streams the latest data (photos, location history, texts, etc ) back to the timeline would probably be more tractable for live updates. But that is probably a wildly different scope than the import based workflow. This is very cool regardless.
For sure.
About 5-6 years ago, Timelinize actually used only the Google Photos API. It didn't even support imports from Takeout yet. The problem is the API strips photos of crucial metadata including location, and gives you nerfed versions of your data. Plus the rate limits were so unbearable, I eventually ripped this out.
But yeah, an app that runs on your phone would be a nice QoL improvement.
Is takeout the only way to get the original photos out?
As far as I know, yes. ("Original" is a strong word, but it's pretty close if you don't have space saver enabled / pay for storage.)
Syncthing from phone to a directory on PC?
That's what I do. Though I don't then put them into any system. Yet.
How easy would it be to integrate this with immich (instead of needing the access to google photo)?
Probably not hard. Timelinize's data sources have a standard API with just 2 methods [0], so it should be fairly trivial to implement depend on how accessible Immich is.
To clarify, you don't grant access to Google Photos, you just do the Takeout from https://takeout.google.com to download your data first.
[0]: https://pkg.go.dev/github.com/timelinize/timelinize@v0.0.23/...
I did this by creating my own small password manager.
"Because Timelinize is entity-aware, "it can project data points onto a map even without coordinate data. If a geolocated point is known for an entity around the same time of others of that entity's data points, it will appear on the map."
In the context of Timelinize, this is great! Outside of that, this sentence really drives home how much data Google could join on me -- a heavy Android/Chrome/Gmail/Maps w/ timeline user.
What are you planning with weather? Associating entities that have a known location with historical temp/forecast data?
Something like that, yeah. Augmenting public data sets like weather or news to add context to your timeline.
This reminds me of a similar project called HPI [1] that I discovered in 2022, the idea to aggregate one's digital footprints into one single private place really inspires me a lot. Timelinize is almost like a GUI version of HPI, a more sophisticated user product that everyone can use. I will definitely try it, congrats and respect!
P.S. HPI has many existing exporters, so it might be helpful to extend Timelinize's capacity.
[1]: https://github.com/karlicoss/HPI
Yeah, years ago I did some study of HPI. :)
Timelinize's data sources are easily extensible so adding more shouldn't be too hard, HPI included.
Thank you!
Great work. Various ideas here:
You might suggest to users the following use case. “If you want to create a Timelinize data store, but don’t feel that your own local systems are secure enough to safely hold a basket with a copy of every egg in your life, you might consider the following use case, which some of our customers implement. Once our twice a year, update the data store, but store it on an external disk. When the update is done, take the disk offline and keep it in a drawer or safe.”
Also
I always wondered how cool it would be if I could tell some Spotify-like system, “I’m 20 miles away from the lake, we are going to stay in the cabins a week, just like we did 10 years ago. Play me the exact same songs now that played at every turn back then.”
Also
For a name, how about: ChronEngine That name seems pretty free and clear from previous use, if you like it grab ChronEngine.com before some squatter does and thank me with a phone call, I would enjoy a quick chat with you.
Also
Your web page might benefit from a grid that lists all the input sources you accept, with hotlinks next to the names that give a pop-up quick summary of what that source is about, and maybe some color coding that shows something like “green = rock solid lately”, “yellow = some users reporting problems”, “red = they just updated their format, it’s broken, we are working on it”. You will/are facing challenges similar to Trillian, a chat client from the early 2000s that would try to maintain ongoing connection compatibility with multiple other chat clients such as AIM/ICQ/MSN. Also, the grid could have a “suggested source sets” filter that helped people find what 5 (for example) input sources they might select for their use style.
Oh and make a list of anybody that says they have done an elaborate something similar with Excel (like me and at least one other person in this thread) and maybe have a discussion with them some time, we/they might have some useful insights.
Let’s hear it for people on the opposite side of the “go fast and break things” coin! My first project took 16 years. My current one I started 28 years ago!
And one more idea from me. I could see your current system to be deliberately kept like it is, but offered as a two-part system, where the second part is, “once you get your timeline built and you have reviewed it carefully, then click here and the local-only LLM will have access to the whole thing.” The two part nature could be a big competitive advantage, helping people to carefully build a LLM system that, for instance, they could then offer to their whole family to peruse, without having to worry too much that it accidentally included information that it should not have.
Thank you for the great feedback!
> You might suggest to users the following use case. “If you want to create a Timelinize data store, but don’t feel that your own local systems are secure enough to safely hold a basket with a copy of every egg in your life, you might consider the following use case, which some of our customers implement. Once our twice a year, update the data store, but store it on an external disk. When the update is done, take the disk offline and keep it in a drawer or safe.”
Sure, I like that.
> I always wondered how cool it would be if I could tell some Spotify-like system, “I’m 20 miles away from the lake, we are going to stay in the cabins a week, just like we did 10 years ago. Play me the exact same songs now that played at every turn back then.”
It's inevitable for an LLM and other tooling like personal assistants to be integrated to this thing, so yeah that sounds like a great use case.
> For a name, how about: ChronEngine That name seems pretty free and clear from previous use, if you like it grab ChronEngine.com before some squatter does and thank me with a phone call, I would enjoy a quick chat with you. ... Oh and make a list of anybody that says they have done an elaborate something similar with Excel (like me and at least one other person in this thread) and maybe have a discussion with them some time, we/they might have some useful insights.
Not bad actually; and sure, we can chat either way. Feel free to book a time at https://matt.chat.
> Your web page might benefit from a grid that lists all the input sources you accept, with hotlinks next to the names that give a pop-up quick summary of what that source is about, and maybe some color coding that shows something like “green = rock solid lately”, “yellow = some users reporting problems”, “red = they just updated their format, it’s broken, we are working on it”. You will/are facing challenges similar to Trillian, a chat client from the early 2000s that would try to maintain ongoing connection compatibility with multiple other chat clients such as AIM/ICQ/MSN. Also, the grid could have a “suggested source sets” filter that helped people find what 5 (for example) input sources they might select for their use style.
I LOVED Trillian, thanks or the nostalgia. Oh man, they're still alive: https://trillian.im/ (I love the icon...)
Nice project! If you don't like "timelinize" - have you looked at latin names? Perhaps something like Temperi?
In terms of Features, i'd like to see support for FindPenguins. A lot of interesting data (photos, videos, GPS coordinates, text) is already there.
A few latin names have been suggested, but nothing has stuck. The problem is they are usually difficult to spell and pronounce, which isn't really an improvement over the current situation :)
FindPenguins is cool! I don't use it myself, but anyone is welcome to implement a data source for it.
This reminds me of Perkeep, but I understand this is more focused on the presentation of the data, while Perkeep was on the data storage. But maybe they could be integrated, or at least support each other's formats.
Hey - this is awesome. I've been working on a small local app like this to import financial data and present a dashboard, for the family to use together (wife and I). So yeah - great work here, taking control of your data.
I'm curious about real-time data, or cron jobs, though. I love the idea of importing my data into this, but it would be nicer if I could set it up to automatically poll for new data somehow. Does Timelineize do something like that? I didn't see on the page.
Cool, yeah, the finance use case seems very relevant. Someday it'd be cool to have a Finance exploration page, like we do for other kinds of data.
Real-time/polling imports aren't yet supported, but that's not too difficult once we land on the right design for that feature.
I tinkered with a "drop zone" where you could designate a folder that, when you add files to it, Timelinize immediately imports it (then deletes the file from the drop zone).
But putting imports on a timer would be trivial.
I have something set up at home: ~/Inbox and ~/Outbox. Anything dropped into ~/Outbox gets rsync'd to rsync.net and mv'd (locally) to ~/Inbox.
Anything in ~/Inbox is "safe to delete" because it's guaranteed to have an off-site backup.
Presumably a fancy management app would queue (or symlink) the full directory structure into ~/Inbox (which would then behave as a "streaming cache")
~/Inbox would effectively be available (read only) on "all machines" and "for free" with near zero disk space until you start accessing or pulling down files.
I use Dropbox to manage ~/Sync (aka: active, not "dead" files).
"Outbox", "Inbox", and "Sync" have been the "collaboration names" that resonated the most with me (along with ~/Public if you're old enough, and ~/Documents for stuff that's currently purely local)
Cool idea. Thanks for sharing. I was really annoyed by the way Google nerfed the maps timeline stuff last year. Obviously this project is way more ambitious than that, but just goes to show you how little Google cares about the longevity of your data.
Nice work. I’ve been frustrated with how closed off location history tools have become lately. This looks like a solid step toward giving people real ownership of their data again. Definitely checking this out.
Thank you. Yes, I feel the same!
It was a different era 15 years ago, made some decent inroads on an API based approach with the Locker Project https://en.wikipedia.org/wiki/Locker_(software) but the platforms quickly started gimping the APIs and exports/takeouts were non-existent or too early yet back then, so ran out of steam.
Thanks for making this, def will be digging in!
Yep, I encountered the same problem. Initially it was solely API-based, but as they got locked down and paid-only and removed entirely, I went the bulk download route.
Looks awesome! Btw, it would be nice if it can also support audio files, I frequently make multi hour long recordings (meetings, presentation etc) and it would be very useful to see pictures and other stuff from a specific time together with the audio from that timing.
Oh it actually does support audio files :) Haven't tested them in a couple years but a player should appear.
This seems like the perfect thing to mix with financial records (ie bank feeds) and a local LLM.
I’m not sure what you’d use it for … exactly … but it could probably reconcile and figure out all your credit card charges based on your message history and location, allocate charges to budgets, and show more analytics than you’re probably interested in knowing.
People with a cloud connected car such as a Tesla could probably get some real “personal assistant” type use cases out of this, such as automatically sorting your personal and business travel kms, expenses and such for tax purposes.
There’s probably other use cases like suggesting local experiences you haven’t done before, and helping you with time management.
Yeah, I hear this a lot, I would love to have my financials and an LLM integrated as well!
Could make for a very interesting/useful/private personal assistant.
ledger.txt (plaintextaccounting.org), g-cal integration, and Home Assistant are all so close to each other.
I really like the local storage of this. Files and folders are the best!
(When noodling on this, I’ve also been wondering about putting metadata for files in sidecar files next to the files they describe, rather than a centralized SQLite database. Did you experiment with anything like that by any chance?)
Why sidecar metadata files? In general I've tried to minimize the number of files on disk since that makes copying slow and error-prone. (A future version of Timelinize will likely support storing ALL the data in a DB to make for faster, easier copying.) We'd still need a DB for the index anyway, which essentially becomes a copy of the metadata.
Don’t know if you have seen the work DuckDB is doing on ducklake. Maybe there is an overlap in vision for versioning data across multiple data sources - and similar to SQLite it’s not proprietary and easily drilled down on. I’m sorry, don’t have technical knowledge :/
Thank you so much for this! I've also had the idea but, contrary to you, never had the courage nor energy to actually do it. I have some of my old devices backed up with the very first text my (now) wife and I exchanged, and this feels like the perfect use case.
So thank you so much, I now have another project to host on my home server, and a meaningful one.
>(PS. I'm open to changing the name. Never really liked this one...)
Timelinize is great, it explains what the product does
Thanks for the feedback; especially about the name. I'm glad you value the project's goals!
I have been recording my gps every few mins for the past 10years for this exact product.
Looks interesting.
Out of curiosity, how have you been doing this?
started with overland (was great, but hurt the battery a lot) - there was an iOS app for it. I then just moved on to PSObserver as it was self contained, I could export off my phone. I also augmented with running google maps background for a few years as well. Think I have over 5m data points now. I user Kepler.gl originally + with iPhone photos as well overlayed.
Interesting. But this seems to be limited to the supported data-sources? No generic format to build your own sources? Also, why the focus on a timeline? Not all personal data or tasks are time-related. So is this more about personal data, or a timeview on personal data?
If you want to have collaboration on this, maybe think about splitting it into backend and frontends. With the backend being the storage, and it's import-mechanism, and the frontend being open for anyone to put their own view onto the data, and maybe work with them.
Maintaining personal data has demand here, but so far there is no real unified project catering to a broad range of services and different types of data. So there is a space to fill and a community to build IMHO.
> Interesting. But this seems to be limited to the supported data-sources?
Data sources are extensible, they basically just need to implement 2 methods: https://pkg.go.dev/github.com/timelinize/timelinize@v0.0.24/...
> No generic format to build your own sources?
I'm designing an import API that will let any script or program push data to the timeline.
> Also, why the focus on a timeline?
Seemed to make the most sense.
> Not all personal data or tasks are time-related.
Probably true. Do you have examples of what you're thinking of?
> If you want to have collaboration on this, maybe think about splitting it into backend and frontends. With the backend being the storage, and it's import-mechanism, and the frontend being open for anyone to put their own view onto the data, and maybe work with them.
That's how it is right now; it's a client-server architecture. It has a CLI and a JSON API so you can build your own front-ends.
> Maintaining personal data has demand here, but so far there is no real unified project catering to a broad range of services and different types of data. So there is a space to fill and a community to build IMHO.
I agree, this is a huge void that needs to be filled!
> Probably true. Do you have examples of what you're thinking of?
Bookmarks, contacts, notes for example, they are usually not time-based but by context, category or whichever organization their usage demands at the moment. They can be time-based in a journal, but usually I do not remember websites or people by the time I encountered them, so I would not search them this way.
The other question is how personal is personal in this project? Are my IMDB-ratings valid data for this? My Ebook-collection? My Steam-account? Those are usually things I would not manage in a timeline or a map, but with different interfaces and features.
Bookmarks are bookmarked at a specific time.
Contacts are already handled as first-class concepts in Timelinize, known as "entities". Entities don't have timestamps, but they have a name and attributes that can be anything, and the relationship between the entity and its attributes can be bounded by time.
Notes are also created and updated/changed at specific times.
> Are my IMDB-ratings valid data for this? My Ebook-collection? My Steam-account? Those are usually things I would not manage in a timeline or a map, but with different interfaces and features.
Could be, if you wanted to timeline those! (Like, when you rated a certain movie. Or when you got or read a certain book. Or when you got a new Steam game.)
Like others I really like the idea and the realisation looks great too!
I might not be the typical user for this, because I'd prefer my data to actually stay in the cloud where it is, but I'd still like to have it indexed and timelined. Can timelinize do this? Like, instead of downloading everything from gphoto, youtube, bluesky, wtv, just index what's there and offer the same interface? And only optionnaly download the actual data in addition to the meta-data?
That's not really aligned with my vision/goals, which is to bring my data home; but to be clear, downloading your data doesn't mean it has to leave the cloud. You can have your cake and eat it too.
The debate between importing he data and indexing external data is a long, grueling one, but ultimately I cannot be satisfied with not having my data guaranteed to be locally available.
I suppose in the future it's possible we could add the ability to index external data, but this likely wouldn't work well in practice since most data sources lock down real-time access via their API restrictions.
Very cool! I have a sketchy pipeline for exporting my data from Gmaps to my personal site and always thought about building something like this.
This could be a really interesting as a digital forensics thing.
I wanted to give it a try, unfortuantely I fail already at the first screen. Selection a location the backend fails at creating the timeline:
It seems it is trying to parse windows file paths as URIs, and thrown off by the ":" in windows file paths.Just fixed that. Thanks. You can download the latest CI artifact for the patch (the link is on the Install page in the docs).
Thanks a lot, I wasn't really sure if reporting here or on github would be faster, I'll try to report on github in the future!
Love the grind! One suggestion would be to add a demo link with some test data so we can see it in action.
I am also slowly "offlining" my life. Currently, it is a mix of synology, hard drives and all.
I have always thought about building a little dashboard to access everything really. Build a financial dashboard[1] and now onto photos.
[1] https://github.com/neberej/freemycash/
A live demo would be great, but I'm not sure how to generate the fake data in a way that imitates real data patterns. That's originally how I wanted to demo things, but the results weren't compelling. (It was a half-hearted effort, I admit.) So I switched to obfuscating real data.
FreeMyCash looks great! Yours is the second financial application I've heard of; maybe we need to look at adding some finance features soon.
Hmm, I bet you could find appropriately-licensed datasets on Kaggle or HuggingFace that you could repurpose for that.
I had this same idea for a long time. Even took github.com/center for it (I've since changed how it's being used). Cool to see someone actually achieve it, well done.
Wow that's great! Interesting if it's possible to use not just a folder but like a s3-compatible backend for photos and for db backups as well
(I don't think all my photo/video archives would fit on my laptop, though the thumbnails definitely would, while minio or something replicated between my desktop plus a backup machine at Hetzner or something would definitely do the thing)
I don't think sqlite runs very well on S3 file systems. I think it would also be insufferably slow.
I even encountered crashes within sqlite when using ExFAT -- so file system choice is definitely important! (I've since implemented a workaround for this bug by detecting exfat and configuring sqlite to avoid WAL mode, so it's just... much slower.)
Would https://litestream.io/ perhaps help?
Definitely not sqlite-on-s3! Just for the photos and videos, and the periodic db backups
I see... that might make it hard to keep all the data together, which is one of the goals. But I will give it some thought.
Super interested in this as well (and thank you for Caddy)
How does this handle data updating / fixing? My use case is importing data that's semi structured. Say you get data from a 3rd party provider from one dump, and it's for an event called "jog". Then they update their data dump format so "jog" becomes subdivided into "light run" vs "intense walk", and they also applied it retroactively. In this case you'd have to reimport a load of overlapping data.
I saw the FAQ and it only talks about imports not strictly additive.
I am dealing with similar use cases of evolving data and don't want to deal with SQL updating, and end up working entirely in plain text. One advantage is that you can use git to enable time traveling (for a single user it still works reasonably).
Glad you like Caddy!
> How does this handle data updating / fixing?
In the advanced import settings, you can customize what makes an item unique or a duplicate. You can also configure how to handle duplicates. By default, duplicates are skipped. But they can also be updated, and you can customize what gets updated and which of the two values to keep.
But yes, updates do run an UPDATE query, so they're irreversible. I explored schemas that were purely additive, so that you could traverse through mutations of the timeline, but this got messy real fast, and made exploring (reading) the timeline more complex/slow/error-prone. I do think it would be cool though, and I may still revisit that, because I think it could be quite beneficial.
Have you heard of XTDB / Bitemporality? The basic idea is to make time 2-dimensional, where each record has both a System Time range and a Valid Time range. Designed as a write-only db with full auditability for compliance purposes.
With 2D time you can ask complex questions about what you knew when, with simpler questions automatically extended into a question about the current time. Like:
You probably don't want to just switch to XTDB, but if you pursue this idea I think you should look into 2D time as I think it is schematically the correct conceptualization for this problem.Docs: https://docs.xtdb.com/concepts/key-concepts.html#temporal-co... | 2025 Blog: https://xtdb.com/blog/diy-bitemporality-challenge | Visualization tool: https://docs.xtdb.com/concepts/key-concepts.html#temporal-co...
Yeah, I did actually pursue this for a time (heh), but I might revisit it later. It was too much complexity for debateable value-add, though the value is growing on me.
Thanks for the reply! I'll have to try this out... it almost looks like what perkeep was meant to become.
One interesting scenario re time traveling is if we use an LLM somewhere in data derivation. Say there's a secondary processor of e.g. journal notes that yield one kind of feature extraction, but the model gets updated at some point, then the output possibilities expand very quickly. We might also allow human intervention/correction, which should take priority and resist overwrites. Assuming we're caching these data then they'll also land somewhere in the database and unless provenance is first class, they'll appear just as ground truth as any other.
Bitemporal databases look interesting but the amount of scaffolding above sqlite makes the data harder to manage.
So if I keep ground truth data as text, looks like I'm going to have an import pipeline into timelinize, and basically ensure that there's a stable pkey (almost certainly timestamp + qualifier), and always overwrite. Seems feasible, pretty exciting!
Oh I love this oh so much. It lines up with a lot of things I've wanted or am planning. So I'm definitely going to take a dive into your source for inspiration
One thing I wish for encompasses the idea that I might want to share different slices of my life with different groups
Sharing is planned. Long term roadmap but definitely on my list!
I was looking for a "sharing idea" comment, so it's great to see something is planned.
Being able to share a time-based, geo-based, or X-based slice (all or some of) with someone would be a great way to learn about "close encounters" before or after we met, or even someone else's perspective of a shared trip or experience.
Just a general comment; you've made something truly wonderful (IMHO), and I look forward to seeing where you take this.
Thank you!
Sounds really cool. I’ve been wanting something like this. Kudos for building it!
I don’t see the link to the rep on on first glance of the linked site, so linking it here: https://github.com/timelinize/timelinize
As for branding, IMO you could go a bunch of directions:
Timelines
Tempor (temporal)
Chronos
Chronografik
Continuum
Momentum (moments, memory, momentum through time)
IdioSync (kinda hate this one tbh)
Who knows! Those are just the ones that fell out of my mouth while typing. It's just gotta have a memorable and easy-to-pronounce cadence. Even "Memorable" is a possibility LOL
-suggestions from some dude, not ChatGPT
Dateline (with the Dateline NBC theme song playing quietly in the background while you browse your history and achievements)
Momenta
Scribe? As in the person who writes the timeline.
This basically brings your data from the cloud to local-first ! Kudos to your dedication and especially making this open-source for the benefit of everyone!
Oh man, this is awesome. I wanted an answer to Apple's Journal because I personally am using Obsidian (doing my best to tag and ref data). But this can be what I need to access that metadata layer and combine everything
This is really cool! I’ve always wanted to do something similar, but didn’t want to invest the time, so thanks for seeing this through!
for many years I kept Tiny Travel Tracker running on my android phone, and would periodically import it into os&m maps. It was nice to have the record of all exactly all the places I had wandered to, and not also share that with five eyes. https://f-droid.org/en/packages/com.rareventure.gps2/
> Recommended: GPU (NVIDIA 20-series or newer; 40-series for faster thumbnails) or Apple Silicon (M2 or newer)
Very nice project, it is something what we all need right now in these days. Just noticed that you recommend Apple Silicon M2 or newer? Is there really a significant difference between M1 and M2? Is it running some LLM model in the background?
I dunno. M1 might be fine, I just haven't tested it with an M1. If you enable semantic features, it does use a large model (not necessarily an LLM) for embeddings; but regardless it will generate thumbnails for images and videos, and transcode videos for playback, so a good GPU is helpful.
Interesting; how easy is it to backup it up somewhere - yes, on a cloud for example - and then restore/sync it on another machine? Is the data format portable and easy to move like this?
Yep -- a timeline is just a folder with regular files and folders in it. They're portable across OSes. I've tried to account for differences in case-sensitive and -insensitive file systems as well. So you can copy/move them and back them up like you would any other directory.
Matt, I am so impressed by how far this has come. WRT naming, what about "Journey".
OTOH, I always thought just simplifying to "Timeline" was the obvious move?
Thanks man! Good to see you here. Nice suggestion, I have a list on a text doc here, I'll add this one :)
Great idea! Do you plan to integrate with other services? I'm looking for similar solutions to collect feedback and improve my product. Are you already getting feedback from the community?
Got a lot of feedback here so far, plenty to keep me busy. And there will be an import API to let scripts/programs push data to Timelinize, even if it doesn't support those data sources built-in.
It reminded me of Camlistore (Perkeep). I've always liked that kind of project since I'm terrible at keeping my personal data organized.
Yeah, big fan of Perkeep. I think it's well-engineered. (Better than Timelinize probably! Just missing some important features for me.)
Nice one, thanks for sharing. For sure I’ll give it a try.
Have you thought of creating a setup so as to package all libraries and dependencies needed? You have a very nice installation guide, but there are many users who just want the setup.exe :-)
Thank you! Not sure I can package everything because of license requirements. IANAL. I think the container image basically automates the various dependencies, but I didn't create it and I don't use it so I'm not 100% sure.
Basically, I would love to know how to do this correctly for each platform, but I don't know how. Help would be welcomed.
Beautiful app. Surprised to see JQuery for your frontend; brings back good old memories.
Ha, thanks it’s actually AJQuery, just a two line shim to gain the $ sugar. Otherwise vanilla JS.
Very nice project! Curiousity question: since you are taking data dumps once-twice a year, and let's say you also copy the photos as well, do you do any updates incrementally or just replace the old one with new dump?
Timelinize doesn't import duplicates by default, so you can just import the whole new Takeout and it will only keep what is new.
But you have control over this:
- You can customize what makes an item a duplicate or unique
- You can choose whether to update existing items, and how to update them (what to keep of the incoming item versus the existing one)
For a name, how about 'Rain Barrel' ? Your own personal cloud
I am remembering Perkeep (nee Camlistore): https://perkeep.org/
This looks so cool ! I will try it ASAP !
I've always wanted this but not enough to build it. I wonder if I can integrate this with my Monica instance. Thank you! I'm going to try it.
Will be curious how you use it. I plan to integrate local LLM at some point but it’s still nebulous in my head.
That's great, I've been running a timeline of my life in excel, I wonder if this could replace it.
I've been basically doing this for years via a private mastodon instance. Very nice to see!
ballast.
i want to forget the past except for keeping the warmest moments in my heart.
i know others will feel differently.
that said, beautiful presentation and website!
Thank you!
And yeah, I feel ya. This lets you delete things too.
Amazing project. In the era of AI I can see the software like this being used daily.
Yeah I totally want this. How much data are we talking about on average?
Is it open source?
Yes https://github.com/timelinize/timelinize
Call it D-FUKT and the purpose would be to show people how absolutely fucked all of their data is. If you list every corp/org that has access to each piece of data, this could be cool
This is just some feedback. But when including product shots on your web site, at least don't use real user data where you have to blur most of the content. It looks tacky and unprofessional.
On the contrary, I think it's even better to use real data instead of fake data. Using fake data looks tacky and unprofessional.
So you'd prefer to look a screenshot where half of the content is blurred out?
Yes, absolutely moreso than a demo that is faked.
Congratulations, good work, and good luck, "TickTock" might be an appropriate name.
When announcing a product, the language, license, version number, amount of time under development since inception, and approximate number of contributors are all critical pieces of information to include in an announcement.
Ten years is a long time. Is it just you? What language is it in? Does it run on desktop or does it need a server? What is the license?
outside of pictures living in the past bro