Hacker News new | past | comments | ask | show | jobs | submit login

Wow, it was really hard to find out what this software does. I finally found a demo at the end of an hour-long talk from 2 years ago: https://youtu.be/PlAU_da_U4s?t=2687

So it seems to have a bunch of scripts to import data into its database from a variety of sources (including cloud services), and provides a search interface to navigate that historical data. And it has a lot of under-the-hood stuff about replication, and it's entirely self-hosted.

You initial confusion and nice summary made me thought of camlistore which I looked at 4 or 5 years ago. And here it is, renamed!

This is sad because the project looks interesting, it is not the usual quick kludge or vaporware, but the bad or inexistent documentation is a terrible disservice to the project.

Yeah, I've seen this project multiple times over the year, was always interested in it but never managed to figure out what it's for, let alone how to use it.

Perkeep is camlistore, FYI.

They know.

> And here it is, renamed!

Does the first sentence not cover that fairly well?

> Perkeep (née Camlistore) is a set of open source formats, protocols, and software for modeling, storing, searching, sharing and synchronizing data in the post-PC era. Data may be files or objects, tweets or 5TB videos, and you can access it via a phone, browser or FUSE filesystem.

I checked archive.org and that text has been there for a couple of months at least. Looks interesting, I've been in the market for this kind of self hosted backup/replication/tagging/search thing.

> Does the first sentence not cover that fairly well?

Not for me. Almost all of what you copied there seems like intrinsically meaningless fluff. "a set of formats"? "a set of protocols"? "software"! Of course it's software! "for modeling"? "in the post-PC era"?

I would have preferred GP's ~"imports data/files from cloud services into a local database and provides a search interface for it". Because THAT tells me what it does, not any of the other words you posted.

Someone made a really well done cartoon mascot for it, but the copy isn't _helpful_.

> Does the first sentence not cover that fairly well?

It does not.

The first sentence on the site ("Perkeep [...] is a set of open source formats...") describes literally what the thing is, but not at all what it _does_.

Not to slam on these cats, because marketing copy is _hard_. For project collaborators, or open-source dorks who live in this kind of world anyway, the sentence on the homepage is probably perfectly descriptive.

But I agree with my GP post. Reading the homepage I had no idea what Perkeep actually did.

What it _does_ is:

> modeling, storing, searching, sharing and synchronizing [...] files or objects, tweets or 5TB videos, and you can access it via a phone, browser or FUSE filesystem

I mean maybe it could have been more explicit or they could have added more detail, but having this as the first sentence is WAY better than most of the 'professional' landing pages for startups that get posted here. 'Harmonizes synergy and increases your ability to wow your target space with your aspirations', now that's meaningless.

Them: “The project description isn’t clear to me.”

You: “Well I’m sorry it wasn’t clear to you but it was clear to me and better than these other things and here’s why it should have been clear to you.”

If someone tells you something is unclear to them, arguing about it doesn’t change the fact that it wasn’t clear to them.

As an engineer

> files or objects, tweets or 5TB videos

means rsync, curl, and Twitter API integration. It obviously does more than this, since I can throw something together that does that in a few hours. Where is the list of everything it supports?!?

That should be front and center.

I managed to parse it, but there's a great grabber sentence down below that might be right at the top:

"Your data should be alive in 80 years, especially if you are." To which you might add, "We're here to help you make sure that's what happens". Then follow that with the "Things Perkeep believes ..." section.

After that, the mission is clear, how it works is clear (though many people might have -no idea- what 'Open source' is good for). Only Then (IMO) can you get away with going all technical on them!


Not for muggles is a silly cop out. It's very nondescriptive. That "set of protocols and file formats" etc. could do literally anything.

Is it for a nas? Is it distributed? Is it some backup yoke for a cloud service? Do you run it on your own hardware? Is it all of the above? What does your data look like? Where and how are you searching through it? Is that part not for muggles either or is there a frontend? And so on basically forever.

For me it didn’t. That sentence is too vague as to what I as a user will be able to do with this project.

It reads more like a patent description than telling me what it's for.

yeah like a lot of OSS backup products the docs are grandiose without focusing on ease of operation / feature parity with dropbox

Perkeep is really straightforward to operate though. One can just install the dmg, mount the root and put files there. Also there's a web interface. For me the most interesting feature is though how easy it is to setup a new node and to sync to/from it. Like rsync but no need to worry about the parameters. Also as the name suggests, it's far more difficult to loose important data than with Dropbox (and similar products like iCloud) which can quickly happen through wrong usage.

no wonder looking at the funding, is it?

You don't need funding for focusing on ease of operation, it's a mindset. Either you're user-centered, or you're not.

>Either you're user-centered, or you're not.

It really depends on the audience for the product, targeting regular users does require more thought on the UX, etc which almost invariably means more funding.

I don't think the website targets the wrong audience - just the copywriter was so involved in the project they forgot to make what they're offering clear.

When someone says "permanently keep your stuff, for life" do they mean some sort of pay-once eternal backup, like permanent.org? Something censorship-resistant, like Freenet? Something peer-to-peer and distributed-ledger based like Filecoin? Backing up data locked up in cloud services? Converting obscure file formats into ones with more longevity? Bypassing defunct DRM? Activism against civil forfeiture?

Just from the front page it looks like a software stack for running your own archive server?

Why do you think cmd line tools are popular ? Because they are easier to do than a GUI, not because all geeks think the terminal is the only xp that matters.

That seems... offtopic? You can have command line tools that are easy to use, with documentation focused on user-friendliness, and you can have command line tools where figuring out what they even do in the first place takes a lot of effort. And the same is true for GUIs, for what it's worth.

That takes time. Or funding. I have often thought "it would be great if I could suck all my emails and Facebook posts into a local search index." It sounds like this is that. If I wrote it I would have exactly enough time to make sure that all my emails and Facebook posts are in a local index and then I would have no more time for a personal project. Making a nice UI on top of it would easily increase the amount of time fivefold.

Agreed. It's really about benefits vs features. Features are what. Benefits are why; why you should care. Addressing why is essential.

About $30k - https://opencollective.com/perkeep/expenses - which is more than most OSS projects would ever see in their lifetime, so it's not such a good excuse for lacking documentation and UX.

$30K over 13 years, and 98% of that funding was the lead developer and his wife (plus 10% from another developer) giving their own money to another developer.

It looks more like "wealthy Googlers helping out a friend with a short term gig" than "financing a professional product".

Perhaps you could donate your time as a UX specialist to help these folks who are more versed in backend systems and libraries.

I'm more of a backend person myself, and perkeep disappoints on that side too. See my other comment about how much time it took to ingest just one .iso. I don't think this system would handle a 5 TB video file as the authors claim, that's some false advertisement!

I considered it for my use case (archiving and deduping dozens of terabytes / millions of files on several personal NAS boxes), really wanted to use it - as I find some of its ideas pretty cool, but in the end decided it would be simpler to just write something from scratch instead. It took less time to write it than it would have taken perkeep to ingest my data.

Same video is posted at the bottom of their website.

I felt the same but if you go to "Docs" the very first link is then:

"Overview: The original motivation and background for why Perkeep exists and what one might use it for."

And there I found a great description.

it’s a distributed personal content addressed blob object store, with a variety of configurable storage back ends, plus a web interface to make that kind of usable to a human for human things.

it’s like git but for all your stuff.

why would you want to use it? you probably wouldn’t, quite yet. but it’s an interesting attempt at doing something a little more sophisticated than a plain file system.

This comment represents the value of HackerNews comments very well :) Thank you.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact