Hacker News Reader

Comment by drvortex | Hacker News Reader

As other's have mentioned, the WebUI is a major no-no.

Evernote's strength lies in its Desktop application and very useful clipping service.

The fact that Evernote uses its own cloud storage instead of say an encrypted folder in Google Drive/Dropbox etc is a negative of the service. Sync is required but implemented in the most undesirable way in Evernote currently.

Paperwork is promising, but seems to have replicated the negative of Evernote, while foregoing the reason why people use Evernote in the first place.

So, if you want to Paperwork to take off:

1. Implement it as an offline Chrome App with Dropbox+Google Drive sync. 2. Implement a web clipping+screenshot extension.

View on HN

> The fact that Evernote uses its own cloud storage instead of say an encrypted folder in Google Drive/Dropbox etc is a negative of the service.

I hate this trend towards supporting only one or another kind of proprietary service.[1] What is wrong with WebDAV? That way, you can use a cloud hosting provider, or self-host.

[1] I'm looking at you Scrivener.

And another proprietary service ("make it an offline Chrome app"). It's the web, why would I want to be locked to a single vendor?

It is the path of least resistance to get to a 'Desktop' experience. And please, what are the alternatives?

Firefox ? I'm sorry, but Firefox lost relevance a couple of years ago now. Firefox is ethically the best , but it sucks in implementation. Firefox is now the broccoli of the web, stale broccoli at that.

ActiveX is pretty cool, and gives you a "desktop" experience. You should look into it if you like letting one corporation lock you into their proprietary version of the web.

a desktop app built with node-webkit would be better than a browser-specific app

How? All you're proposing is replacing one dependency with another.

At least there it's not really any different than shipping a WPF application that's tied to Windows. It's a desktop application, people are used to desktop applications being vendor-specific. I don't like the idea of people getting used to web apps being vendor-specific though. That's not okay. We have a chance to make it right this time. If it runs on your desktop and is vendor-specific, that's a shame but that's expected. If it runs in a browser and is vendor-specific, that's setting a bad precedent that we worked for years to overcome with IE6.

Of course, that's just my personal philosophy. Everyone is different.

Lots of things are wrong with WebDAV, summed up as "many clients with different implementations/interpretations of the spec."

Related: Are there any decent WebDAV servers out there? I don't see any recent development being done on any.

Sure! SabreDAV 2.1 released in November 2014, IT Hit WebDAV Server 3.9.2098 released in December 2014.

And Box is using WebDav too: https://support.box.com/hc/en-us/articles/200519748-Does-Box...

I believe the idea was that the app would simply save data on a local drive that could then be synced with your favorite technology or service.

Paperwork doesn't use its own cloud storage. You can host it either locally or at your own server. No problem using encryption, VPN tunneling etc. here.

The fact that it's not a true cloud service could be seen as a downside. Your data is only stored at one server without failsave mechanisms. And you're responsible on your own regarding the data.

Paperwork doesn't use its own cloud storage. You can host it either locally or at your own server.

This will really limit the potential user base and really eventual growth and uptake of a service like this. When easily accessible "personal cloud" tools like Dropbox and Drive exist why not make them potential service mechanisms? On an open source service like this it seems you may want to have some sort of cloud abstraction service and let other contributors plug in separate cloud services.

Dropbox, Drive, iCloud, etc are not in any way 'personal clouds'. They are very much somebody else's cloud.

Proper personal clouds may be on the horizon when you don't have to become a sysadmin to run one. FWIW, I'm working towards this - http://nymote.org/blog/2013/introducing-nymote/

Many people (myself included) don't trust Drobox, so being able to handle my data myself is regarded as a plus.

(having Condoleezza Rice on the board is just _one_ obvious reason not to trust them)

That makes me think, has anyone tried to build a note taking app based on git? A tricky part for this kind of app is networking and synchronization of notes, and that's all taken care of by git. The "only" thing left would be to implement the mobile and desktop front ends.

I've done emacs org mode on git, and it works. I've thankfully never had to do a manual merge, I wonder what org mode thinks of git merge commentary inside the file. I imagine you'd have to drop out of org mode and into text mode to fix the merge.

I've also used a wiki as what amounts to little more than a note taking app, and it works. There are a couple wikis that can speak git as a backend store.

I can't begin to imagine why someone would build a note-taking app that wasn't build on a standard version control system such as mercurial or git.

Is there a rationale for paperwork not doing this? I'd like to hear it.

Magpie, which I think was on HN a few weeks back, is an attempt at precisely this.

https://magpie-notes.readthedocs.org/en/latest/

I must have missed it a few weeks back, but thanks for mentioning this because it's just what I need!

You might like to take a look at Irmin [1], which seems like it would make it easier to get what you're after. It's also part of a larger stack that will make distributed programming a lot easier when it lands [2].

[1] http://openmirage.org/blog/introducing-irmin

[2] http://nymote.org/blog/2015/brewing-miso-to-serve-nymote/

Better yet -- CouchDB and PouchDB. You can use PouchDB for the browser side, CouchDB for the server, and Couchbase Mobile for the eventual mobile apps. Pros: offline and online; natural fit for document management; git-like revision and conflict management; simple to set up replication for backup. Cons: it's fallen out of favor; nobody seems to like writing the mapreduce code for searching and views.

PouchDB dev here. The map/reduce API is definitely a bit cumbersome, which is why we're replacing it: http://nolanlawson.github.io/pouchdb-find/ :)

Draft (http://draftin.com) is a great version-controlled writing app, though it's not what I would call note-taking. Interesting idea though!

Yes, me. :) I'm intermittently working on an Android app that syncs to Git. I'm using it for my notes, but it's still pretty rough to share widely.

Because you want autosave to the cloud, and autosave to git would be very weighty.

I think the better option would be APIs for desktop app(s), maybe with a stock basic functionality desktop app. Definitely not an "offline" app based on a particular web browser.

As a Firefox user I already dislike Gmail Inbox and WhatsApp (web) being Chrome only. The former has said "Chrome only for now" for quite long already.

It is actually a shame that it is based on PHP. If it was based on pure JS, there was hope to make a desktop version easily using node-webkit or something.

I have to agree. Evernote's Web Clipping feature is a feature that is very important to me. I would also really appreciate the ability to save my data (by default) some place other than Evernote's cloud.

Yeah, as NixNote user, I wouldn't mind switching away from Evernote as a back end. But a web-only UI is a blocker for me.

Is there an open source and self hosting cloud that apps like this can build on. I know Own Cloud exists but is it app-friendly. This requires a mysql db - yuck.

It does not require mysql db. sqlite works fine (at least for smaller systems).

They recommend mysql (especially for bigger systems), but not a requirement.

sandstorm.io

Donno how good it is but they have a cat mascot so thats a start.

That was a joke you mean downvoters.

HN doesn't like jokes. It's dry.

HN has that reputation, but it isn't entirely true.

Some of my highest-voted comments have been pure humor. A few examples, with parent comments included for context:

https://news.ycombinator.com/item?id=8690893 (86 points, and 24 for the followup comment)

https://news.ycombinator.com/item?id=6447007 (43 points)

Even this silly one-liner got 23 points:

https://news.ycombinator.com/item?id=7863205

That one won't make sense out of context, so here's the article it refers to:

http://www.newstatesman.com/business/2014/06/bizarre-secret-...

My point in sharing these is to give some examples of kinds of humor that have gone over well here. Of course as they say, "Past performance is no guarantee of future results."

Reply on news.ycombinator.com