writing a social messaging app aimed at end-users

[Michiel B. de Jong]

hi!

We are thinking again about writing an app for end-users, probably in
the "messaging" corner. The goal would be to write a "killer app" that
is good enough by itself to make people (end-users) sign up for
remoteStorage and sockethub accounts, even if they don't care about the
freedom aspects.

We tried this last year with Libredocs and that helped us remove a lot
of the hurdles and pave the way, even though the resulting app itself
turned out to be too complex for us to maintain. So this year we're
trying again, and since we now have remoteStorage, nimbusbase, and
sockethub, but no good way yet to convince non-developers to use these
technologies, we were thinking a messaging app might solve that
deadlock. another option would be a project management app, as Ray and
Melvin also suggested already, or some sort of editor (either office or
IDE).

does anybody have ideas about this?


Cheers,
Michiel

[Jan-Christoph Borchardt]

Agree about messaging, talked about it with Nick a lot already. Using remoteStorage for storing messages and Sockethub for transport (being able to plug in Email, XMPP or such) and a simple interface reminiscent of private messages. (or Skype, or Facebook messages, or whatever the hell you want to compare it to).

The question is probably who would do it, and most importantly who would maintain it (unlike Libre Docs, unfortunately). Since it would be a flagship unhosted app it would probably make sense for you, Michiel, as being able to work full-time on it to lead this effort and really take care in building a stable product rather than having just another »research project«.

Regarding the other suggestions, I think a project management app wouldn’t have the same scale or usefulness by far. As for editor, you know that Jorin and I are working on Litewrite: http://litewrite.github.com/litewrite which aims to keep things simple but can be used as a full blown notes app.

--

--- You received this message because you are subscribed to the Google Groups "Unhosted Web Apps" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unhosted+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Michiel B. de Jong]

On 2013-03-11 20:38, Jan-Christoph Borchardt wrote:
> The question is probably who would do it, and most importantly who
> would maintain it (unlike Libre Docs, unfortunately). Since it would
> be a flagship unhosted app it would probably make sense for you,
> Michiel, as being able to work full-time on it to lead this effort
> and
> really take care in building a stable product rather than having just
> another »research project«.

okay, let's do it. i'll take the lead in unorganizing it, then. :)
Project kick-off at re:publica, when several of us are in Berlin.
Hopefully that
will give us all some time to get remoteStorage (Niklas) / NimbusBase
(Ray), sockethub (Nick) and useraddress (me) into a shape where this app
can rely on them in production and thereby find which bottlenecks still
exist, so we will know what to fix.

> Regarding the other suggestions, I think a project management app
> wouldn’t have the same scale or usefulness by far.

okay, i think i agree. Ray, what is your stance on that? You seemed to
be leaning more towards productivity app, do you think a messaging app
(addressbook) could also fulfill the role of production-proofing our
products?

> As for editor,
> you know that Jorin and I are working on Litewrite:

> http://litewrite.github.com/litewrite [3] which aims to keep things

> simple but can be used as a full blown notes app.

yes! :)


Cheers,
Michiel

[NimbusBase]

Hi all,

First of all, I'm currently working on a decently complex project management app built on NimbusBase, it not only allows per user storage but also sharing via GDrive; you can invite other people to collaborate with you. Check it out!

https://vimeo.com/61775294 (will be released at Taskstrike.com)

I'm going to open source this after some cleanup and hopefully it'll be like a open version of Podio or Wunderlist. If anyone is interested in contributing to something like this or switching it to to remotestorage, feel free to shoot me a email.

I think the best way to production proof a product is getting users to use it. Move from the realm of experimental technology into real consumer product. I leaned towards project management since I think people will always want to use a new productivity app, especially one with better privacy. 

Messaging is also a good area, if you can also do mobile, a lot of privacy concern there too. I'd be happy to help with interface + do Nimbusbase version if the project really gets underway.

Thanks,

Ray

[Michiel B. de Jong]

On 2013-03-14 14:39, NimbusBase wrote:
> https://vimeo.com/61775294 [1] (will be released at Taskstrike.com)

Cool stuff! not sure if adding remoteStorage to Taskstrike would make
sense, because it would (at least at first) be without the 'share with
any gmail address' feature. but the app is definitely going to be a good
showcase for Nimbusbase!

> Messaging is also a good area, if you can also do mobile, a lot of
> privacy concern there too. I'd be happy to help with interface + do
> Nimbusbase version if the project really gets underway.

Great, thanks a lot for that! :) I'll be in touch once we get to that
stage, probably in about 6-8 weeks from now.


Cheers!
Michiel

[Sebastian Kippe]

Hi,

> We are thinking again about writing an app for end-users, probably in the "messaging" corner. The goal would be to write a "killer app" that is good enough by itself to make people (end-users) sign up for remoteStorage and sockethub accounts, even if they don't care about the freedom aspects.
>
> We tried this last year with Libredocs and that helped us remove a lot of the hurdles and pave the way, even though the resulting app itself turned out to be too complex for us to maintain. So this year we're trying again, and since we now have remoteStorage, nimbusbase, and sockethub, but no good way yet to convince non-developers to use these technologies, we were thinking a messaging app might solve that deadlock. another option would be a project management app, as Ray and Melvin also suggested already, or some sort of editor (either office or IDE).

With an app so complex and with unfinished implementations like sockethub as dependencies, I see exactly the same problem as last time arise. How about doing a remoteStorage-only app first? remoteStorage is stable now, sockethub is not (not even usable for me in the demos).

Don't get me wrong, this app should be built, and I'm willing to help, because it'd be so great to have it, but let us *please* try to get the now-stable remoteStorage used, before we evangelize the next unripe technology before its time! In the meantime we can still experiment with and use sockethub stuff. The important thing here is *use* daily as opposed to just build it for the sake of having a demo.

Cheers
Basti

[Hugo Roy]

Le vendredi 15 mars 2013 à 12:40 +0000, Sebastian Kippe a écrit :
> but let us *please* try to get the now-stable remoteStorage used,
> before we evangelize the next unripe technology before its time!

I totally agree. I also think we should make it damn simple for people
to host their own remotestorage instance. I'm willing to play the
naive-not-so-technical user here, trying to install a remotestorage
instance on a debian squeeze. Hopefully coming up with a detailed wiki
and why not new features (like LDAP integration… and cohabitation with
other services / not everyone has a VPS or rents AWS)

[Nick Jennings]

Hi Everyone,

Sorry for taking so long to jump in here. I've had some family
matters that have been eating up my time lately, though I've been
following the thread and thinking about the subject a lot. Here's how
I see it.

- As Jan mentioned, he and I already have been discussing a messaging
app since January at Hackerbeach. We worked on a basic mockup for the
layout and discussed how it should behave more or less. A messaging
app is primarily a Sockethub app, secondarily a remoteStorage app
(though could be made to work with Nimbusbase I presume, as well). For
starters, though, this will be a Sockethub + remoteStorage app.

- Meanwhile, a few weeks ago, Michiel, Niklas and myself, (as the 3
currently funded full-time open-source developers) started talking
about all the projects future after our funding ends, and what the
next steps should be in taking things to the next level. We're all on
the same page that the next step here is getting some real first-class
apps created that are not just demos, proof of concepts, or half-baked
research projects that serve no one but ourselves any purpose.

I understand what Hugo and Sebastian are saying about remoteStorage
needing it's own, first-class, demo apps. (However, Sebastian, if
you've tried Sockethub and it's not working, please let me know!).
Though most of the ideas for the first-class apps that we came up with
(messaging, email,

Another thing, for me, is that I need to have this messaging app more
or less working for the Sockethub presentation at re-publica, so
there's something more to demo than code and an ugly testing console.
So, I've already started on a messaging app, though haven't gotten too
far yet due to said unexpected family matters - though I hope to start
picking this back up next week. I'm using Twitter Bootstrap + JQuery
(which came highly recommended by Martin:) + AngularJS. All it does
right now is some config checking, and form modal-windows. At the same
time I'm working on a Sockethub client library to ease handling
WebSocket messages for developers.
https://github.com/silverbucket/dogtalk

I'll move this to the sockethub organization, once it's actually
somewhat working (again, hope to get it up and running next week).

So, if Michiel wants to launch a project at re-publica, that sounds
great, but I need a messaging/contacts app before that. If you guys
want to help on that it would be cool :)

I do agree 100% with what Sebastian and Hugo are saying, that
Sockethub is still new and subject to change, and meanwhile we have
remoteStorage, and no good apps to use it with. We need some
first-class remoteStorage apps urgently. I think a messaging app is
primarily a Sockethub app, so we need to come up with some ideas for a
first-class (and actually useful) app for remoteStorage. Some ideas:
- litewrite (get it stable and get it out there)
- todo/task management
- ... I'm drawing a blank at the moment.

I think the best investment of time, right now, is to get litewrite
100%. It's been in limbo at "almost ready" for what seems like 5
months now. I'm super, super, overloaded with projects right now, so
can't help out too much on this front, unfortunately. As summer
approaches though, I plan to spend more time on front-end work, while
maintaining and continuing work on Sockethub. Though in reality that
plan is heavily dependant on financial arrangements as well.

So, there are my thoughts, overall though I think we're all on the
same page for the next phase of our project(s) - that we all agree we
need to start making real first-class apps..
Cheers
Nick

[Michiel B. de Jong]

Hi,

On 2013-03-16 03:10, Nick Jennings wrote:
> I need a messaging/contacts app before that. If you guys
> want to help on that it would be cool :)

ok, i'll try to jump in and help you with that during April.

> I do agree 100% with what Sebastian and Hugo are saying, that
> Sockethub is still new and subject to change, and meanwhile we have
> remoteStorage, and no good apps to use it with. We need some
> first-class remoteStorage apps urgently.

i think any app that would be useful enough in itself to make people
sign up for remoteStorage, would naturally use sockethub as well. Even
for instance, to make Litewrite more useful you could add some form of
sharing/publishing, where you already start to want to notify people,
ergo, start to need sockethub. Another example is Ray's app, where
"assign a task to any GMail account" is a feature that makes the app
raise above the level of "just as a demo". And to do that on top of
remoteStorage, you would again need sockethub already.

I totally agree that we should continue working on the remoteStorage
apps that are on https://unhosted.org/apps/ and improve them and use
them and weed out the bugs, but i don't think any of them can be made
into a "killer app" based just on remoteStorage, without adding
sockethub.

I have been thinking about email on smartphones. Reading your email on
the bus is one of the most useful things I found when I first got a
phone with 3G. On the iPhone, the email app is in the bottom row of hot
apps, together with the browser and the dialer and i don't remember what
else. Does anyone now how email works on Firefox OS phones? would you
just have to use webmail?

i think an unhosted email app would be valuable enough to end-users to
make them sign up for sockethub+remoteStorage. Last week i met the CEO
of mailbird, which wants to be "the best email client on Windows". I
think it would make sense to try to build "the best mobile email app on
Firefox OS". you would probably need to build a startup company with a
few dozen employees to do that right, but let's move step-by-step, we
can start with a few people, and see how it goes.

also i was thinking if it would be possible to build an unhosted web
app that is better than iPhone's native mail app. I was thinking maybe
adding in support for rss feeds. Syncing of read/unread status outside
iCloud would also be a unique selling point, were it not that IMAP
already covers that.

i'm not so afraid to use sockethub, if there are any problems with it
then we should be fixing them instead of discarding it. it may take a
bit more time that way, but i don't think we should base our choice of
app on which parts of the necessary code have not been written yet. in
fact, i am a bit scared of relying on useraddress for user search, but
if we narrow our "messaging app" down to an "read email on your
smartphone" unhosted web app, then user search is not such a central
element.

what do you think?

[Jan-Christoph Borchardt]

You’re talking about doing a mail app and a news reader, but we don’t even have a simple notes app ready yet. Let’s get real.

We need to stop moving so fast that the ground we built behind us breaks down again. Libre Docs was short-lived, Friends#unhosted and Shared Stuff also seemed to have had a hype but then not many people actually used it. There are a bunch of small apps which seem nice but no one seems to really be using. We need something simple and solid as a start instead of all the crappy »10 apps by the end of next month« goals and »this idea I have is going to be the killer app for remoteStorage« plans.

Sharing and publishing does not necessitate notifications or the use of Sockethub. We have that plan for Litewrite since a few months: https://github.com/litewrite/litewrite/issues/19 and it would be similar code as for the sharing in myfavoritedrinks.

Someone just has to do it. We’re only 1 developer (Jorin) and 1 designer (me) on Litewrite. I know you’re all busy and such, but if we really want to push forward remoteStorage, we need simple, working, useful, polished apps. At the moment Litewrite is that, or close to becoming that. And don’t be fooled – it’s looks very simple and is just a notes app, but the amount of work needed to make it great is crazy. So any help is really needed.

We can keep playing around and pretend we’re a research project, or we can get real and develop some proper apps useful to people. Or in fact focus on _one_ proper app first. I’m obviously biased because I started Litewrite, but it’s funny how that’s the most done and usable remoteStorage app now. And most importantly, the only one which really has a sustained effort and is maintained for a long period of time.

So I’d say instead of thinking other people will do it, or magically appear out of nowhere with great apps if the docs are good, or trying to get your own small app done alone, let’s focus together on getting this 1 app great.

Who’s in? Check our blocker bug for the next deploy: https://github.com/litewrite/litewrite/issues/136

And then things we want to work on for the next version https://github.com/litewrite/litewrite/issues?milestone=2&state=open

[Sebastian Kippe]

Hi,

> I totally agree that we should continue working on the remoteStorage apps that are on https://unhosted.org/apps/ and improve them and use them and weed out the bugs, but i don't think any of them can be made into a "killer app" based just on remoteStorage, without adding sockethub.

I can think of like 20 from the top of my head. Basically any software that syncs state over multiple devices for starters. I use Sharedy every day for my image sharing needs, and I just need to upgrade and fix it so it doesn't use localStorage at all anymore. I also have some code lying around for a bookmarking app I started (Delicious is still not dead), a first draft of an app dashboard, and more.

Frankly, my biggest problem at the moment is the state of modules, or rather that there is no proper state. If we'd do a hackathon soon, we should focus both on that area and the docs.

> I have been thinking about email on smartphones. Reading your email on the bus is one of the most useful things I found when I first got a phone with 3G. On the iPhone, the email app is in the bottom row of hot apps, together with the browser and the dialer and i don't remember what else. Does anyone now how email works on Firefox OS phones? would you just have to use webmail?

I don't ever use email on the go, while I do use about 15 other apps on the phone. Usage patterns are very different between different people.

> i think an unhosted email app would be valuable enough to end-users to make them sign up for sockethub+remoteStorage. Last week i met the CEO of mailbird, which wants to be "the best email client on Windows". I think it would make sense to try to build "the best mobile email app on Firefox OS". you would probably need to build a startup company with a few dozen employees to do that right, but let's move step-by-step, we can start with a few people, and see how it goes.

My email needs are so well fulfilled it's not even funny. It's pretty much the last app on my whole computer I would switch (not speaking of terminal apps like Vim). Seriously. Also, e-mail doesn't need sockethub or remoteStorage. Everyone can sync everything with just IMAP or use a Webmail client that does that. There's no need for a new solution. It's a solved problem. One solved with open standards, I might add.

> also i was thinking if it would be possible to build an unhosted web app that is better than iPhone's native mail app. I was thinking maybe adding in support for rss feeds.

Make that in a separate RSS app, and yes, that would be awesome. But why build a Frankenapp that does both email and RSS? I'd not use it for the RSS if that's just tacked on.

> i'm not so afraid to use sockethub, if there are any problems with it then we should be fixing them instead of discarding it. it may take a bit more time that way, but i don't think we should base our choice of app on which parts of the necessary code have not been written yet. in fact, i am a bit scared of relying on useraddress for user search, but if we narrow our "messaging app" down to an "read email on your smartphone" unhosted web app, then user search is not such a central element.

Nobody's saying you shouldn't use it or develop new stuff with it. Maybe you're not the prime candidate for building anything stable for other people (no offense), so just keep working on the edge and build cool stuff that we can rewrite and then use as stable software. If you hadn't been doing that, we wouldn't have had remoteStorage; but if Niklas et al hadn't rewritten it, we wouldn't have had a stable remoteStorage. ;)

Cheers
Basti

[Michiel B. de Jong]

hi!

On 2013-03-16 20:13, Jan-Christoph Borchardt wrote:
> There are a bunch of small apps which seem nice but no one seems to
> really be using. We need something simple and solid as
> a start

first of all, i totally agree we should fix Litewrite first. But that's
just business-as-usual, the current blocking bug was (re-)discovered a
few days ago, and Niklas just made some progress on it, i saw.

sorry if that wasn't clear, for me it goes without saying that we will
continue to develop the apps that are on https://unhosted.org/apps/ and
adding new ones to them.

but as you already say, no one seems to really be using them. i want to
do several complementary things to change that:

1. work on them. fixing the bugs, and fixing the bugs and missing
features they expose in remoteStorage/sockethub.
2. use them, even if it's annoying. i have been doing this in varying
degrees since November, and it is often a lot of work. but it also helps
identify what the real bottlenecks are.
3. market research. like asking people which apps they're using or
would use and why (not); thinking about what our unique selling points
are, how to explain to a user that our product is what they want to
have, etc.
4. offer a better first-use experience in general (regardless of which
specific app the user wants to use).
5. develop one app specifically with the purpose of breaking the
chicken-and-egg problem of number of people who have reusable data on
their remoteStorage.
6. worry about funding. we're in a luxury position with 3 full-timers
paid for entirely by NLnet and Wau Holland Stiftung, but that situation
is temporary. we should also think if at some point we'll need more than
3 full-timers, or not. that decision depends on what we decide for all
the previous points.

now you're totally right in pointing out we shouldn't let point 5
overshadow point 1, and of course point 1 comes before point 5 both in
chronology and in urgency, but we need to have a plan in place for all
points.

but yeah, having said that, let's fix Litewrite and use it! :) i just
saw your comment
https://github.com/litewrite/litewrite/issues/136#issuecomment-15019637
that's great news. i have been using my own 'editor' app for
text-writing and for todo lists because i already use it for coding, but
i'll start using more different apps.


Cheers,
Michiel

[Jonas S Karlsson (☯大鱼)]

Hi.

I think progress is great with a collection of apps!

One issue that irks me with them is that I keep having to login in to each of them before you can do anything useful. To be really unhosted, I'd like to first be able to just use it in the browser, store data in local storage in browser, but if I logged in it'd synchronize.

Also the need to log in to storage for each app it's somewhat tedious. Could one make the browser remember and ask permission for each app instead?

I'm currently working on a new programming language. Its implemented in JavaScript, runs inside node.js it's multitasks so it also runs in the browser. Anyway. The language is it's own interactive environment. What's missing is persistency outside the browser. In some sense you could say my language is unhosted too.

I have two things I want with storage, that it's remote storage backing up changes while interacting, but also ability to share data between members online. Not sure how that fits into the remote storage model? I've been toying with the idea of using some interface to a git hub of some sorts...

Jonas

[Jan-Christoph Borchardt]

Jonas,

On Sun, Mar 17, 2013 at 4:33 PM, Jonas S Karlsson (☯大鱼) <j...@yesco.org> wrote:

One issue that irks me with them is that I keep having to login in to each of them before you can do anything useful. To be really unhosted, I'd like to first be able to just use it in the browser, store data in local storage in browser, but if I logged in it'd synchronize.

That should be possible with the apps. I know it works with Litewrite: http://litewrite.github.com/litewrite, let me know if it doesn’t.

 

Also the need to log in to storage for each app it's somewhat tedious. Could one make the browser remember and ask permission for each app instead?

Michiel, your part to tell about »storage-first«.

[Michiel B. de Jong]

On 2013-03-16 20:38, Sebastian Kippe wrote:
> why build a Frankenapp that does both email and RSS? I'd not use it
> for the RSS if that's just tacked on.

yeah, maybe we should keep them separate.

> If you
> hadn't been doing that, we wouldn't have had remoteStorage; but if
> Niklas et al hadn't rewritten it, we wouldn't have had a stable
> remoteStorage. ;)

sure, even if we define our goals in terms of end-user products, i'll
probably keep doing mainly the R&D, but that doesn't mean we have the
luxury to define our goals in terms of R&D goals. The goals should be
defined and measured in terms of how other people benefit from it. So
that's why i brought up the subject of writing an app aimed at
end-users.

maybe starting in April is too early, and we should cover a broader
range of apps first, make each one usable but keep them simple, use them
ourselves, see which one stands out, and then decide based on that. I'll
think about it some more, thanks for the feedback!


Cheers,
Michiel.

[Michiel B. de Jong]

On 2013-03-17 23:42, Jan-Christoph Borchardt wrote:
> On Sun, Mar 17, 2013 at 4:33 PM, Jonas S Karlsson (☯大鱼)

> <j...@yesco.org [1]> wrote:
>> Also the need to log in to storage for each app its somewhat

>> tedious. Could one make the browser remember and ask permission for
>> each app instead?
>
> Michiel, your part to tell about »storage-first«.

storage-first fixes this across browsers, yes. but if you only use one
browser, it should remember everything, you should only have to log in
to each app once in your life. just make sure you don't ever click the
'disconnect' button if you're on your own computer, and that you don't
clear your browser cache. if multiple users share one computer, then
they should each have a user profile at the OS level.

can you describe the steps that caused your browser not to remember
your connected remoteStorage account for an app? which remoteStorage
server are you using?

Cheers,
Michiel

[Paul Frazee]

I'm going to give remoteStorage a go for some local environments. I'll keep you updated if any off them turn out; we've already got a good start on a feed reader (http://localjs.github.com/rssenv). Might help fill out the app list