What if you combined your last 2 emails ? That could be a starting point.
So how about a SURFnet (Open Web) AppStore ? (Maybe just AppGallery ?) With
some automatically created storage by each educational institution for the user ?
Large userbase and one central location for the apps.
And a way to blocking unwanted/unsuitable apps at the AppStore.
Maybe also allow students to submit apps ?
Same here, just brainstorming :-)
>
> cheers!
> Michiel
mozilla open web apps is the way to go, but they're so open that neither mozilla nor anyone else has launched one yet afaik.
Well, I was thinking it would be a good testbed. I'm sure there would be a
lot to learn from such a first appstore.
> also, i think we should try to create a platform where people get donations
> back for the apps they develop. like flattr buttons, etcetera. that's an
> important incentive i think. and we want the apps to be free, not locked
> into a specific country.
>
> but a combination of branding the store and then branding channels inside
> the store, do you think that could work?
>
But yes, support for branded channels might be even better.
Institutions network is so related to any networking project but could
be separated so significantly by realization/compound
that i propose You to decide if Unhosted focuses for this Market
initiative on educational field or on commercial market, even so it is
all,
often - deeply related.
> but look at firefox. if you compile it yourself, then it does not have the firefox logo. it's called "minefield". This stops people from redistributing it, because nobody wants to use a browser that's called minefield if they can have a browser with a cute fox. they give the features for free in minefield, and the brand is "sold" to you in exchange for setting the default search engine (a sponsored link which they get a lot of money for). i was thinking about this yesterday, and called it 'brandware'. it's very similar to donationware (aka nagware), but subtly different.
Nice research and naming.
> we could also do it purely with flattr.
well, for promotion of Free "we" could promote, as well,
some p2p digital concurrency. Just as promotion, i think
- there would be some more-less major system besides
Bitcoin, to the time of launch. +1 for Donation-back platform, though.
> i'm not sure whether we should use the 'unhosted' brand for this purpose. we have to make sure that it's all sound and there is no risk of meddling interests. it should be a separate thing from the unhosted project (keep the development of the standard separate from the operation of this app store). so maybe it's easiest to just brand it differently and keep it entirely separate. but also, if we have an existing brand that seems to work quite well, we might as well exploit it...
i think - the whole idea is
- oh, well, it would be competing task, but it should be amazing enough to try.
I think - if look without all the strategic, marketing staff -
Unhosted is a very good brand to provide the shop - itself.
If look through other side - Unhosted could - in best or worst
traditions of the market be an Open Source Community developed
project with a "reference" implementation. What is worse - is up to You ;)
My opinion for this Market initiative:
As i'd do it for commercial part -- provide the situation where
developers could use the market and sell there
- for free, beyond some point of popularity. Most of it currently
depends on payment services ofc. i won't describe further, there are
many variants, all - not so interesting.
For service part -- as a process of working with market
- a developer should be tested for quality, reliability, semantically
tagged for search etc.
If that would be a part of service - the Market's brand could start to
matter quality,
actually,
one click install for a web app is nice also, ofc. but i hope it would
be a Web Standard that is more than service.
At least - Mozilla could push if People would help and use Open Web
Apps initiative.
i like the Mozilla's approach in the field - so far, but there are
decisions to take, anyway.
also
- make these segmented on the previous step - quality-product
non-Ad-ware non-malware pro-freedom App-makers to have a choice:
- if they Brand themselves or not -- if yes - provide them with a
stylish badge for their company page
along with the actual service of unhosting, moz-apping etc. Provide
the ability to aggregate their existent rating (maybe
not a reviews - because their apps might change to much in process of
your adoption-Service)
Provide the ability to sell directly from their site, ofc.
By that -- Market would provide its Market Brand reputation, Essential
Services and, maybe, Hosting.
- if they not a Brand now
- market developers - could test these apps by "hosting" app into some
"x-quality" sand-box and waiting for feedback,
ofc. there could be manipulation but manipulation could be almost
everywhere :) And that popularity and/or quality-based
scheme work now for Cloud Markets, more or less.
as a variant:
- after the app receives considerable amount of feedback, ratings, it
would became featured app, even - so without
a Brand, i think brand (small-y) would be provided from every developer, anyway.
As for me, personally - i think - without the tagging etc. that is
just doesn't work.. Search - is the base for all Markets of
near-future,
that is what i think. Search as a Search, something - better than you
see now. What i see now - it is a kind - let's guess system.
And if - all that SEO and Semantic Staff - maybe - too hard to ensure
for all the products-web. I'm sure - we could make it for local
Markets.
At least - be best in a particular areas/topics.
> sorry, i meant to say we should be (one of) the first ones to set up a successful mozila open web app store. maybe that wasn't clear from my > wording.
> https://apps.mozillalabs.com/
in my opinion (pretty brief, though, i haven't read all the mail)
what Unhosted needs to decide clearly and for long -
1. if the project means that Data separates from App;
App being hosted in proprietary even if Open Source Cloud;
Data being hosted locally, without manipulation - transferred with
wise system (centralized anyway?)
for stats of, for example - games; with manipulation if it isn't "games"
or
2. if the project going to let the App itself to be hosted with
something like Tahoe-LAFS
by that being a decentralized (as much as possible) hosting system
with a front-end of Market(s)
if 1st - how Locker Project and other major competitors(projects,
related systems that would work in the same field) - relate.
if 2nd - what the system do you see - for people to host these p2p
apps. By that - how would theFNF.org kind
of projects relate - projects like social mesh of various devices.
For 2nd - i'm developing with FNF for XCCC(i have posted some time ago
here - about) social networks,
for Open Augmented Reality, for etc. for the FreedomBases - mesh
routers that being the reliable hosting by-the-way.
By this - i think 2nd is the hardest but ultimately awesome, as for me =)
For 1st - it is realistic in near-time and could show one of the
ideologically(by that - almost absolutely) best possible practice for
existing Clouds markets, by this - ultimately cool and change-making
project (could be both for Mozilla and Unhosted)
PS: Awesome FLOSS Search is a must in any way, i see - Locker is
working for it, local or not - it should be a base for any modern
Social-interaction-related system, i'm sure and could elaborate, but
i'm sure - you know or feel it already.
PPS: "I'm sure there would be a
lot to learn from such a first appstore." +1
> .. other messages from [unhosted] mailing list .. :)
But we must bear in mind that paying developer is going to change the dynamics of people contributing FLOSS. For example, such system may incentivize people to fork projects instead of patching it. Which may possibly harming the community. Search dunc-tank of debian for the arguments of both sides.
Some brainstorming...
Such system should make people feel cool to donate FLOSS projects. May be it could reward people with beautiful badges, stickers, tshirts, avatar or ribbon for their social network accounts, 'brandware', etc. I think kickstarter is a good example of how to make people feel cool to donate.
May be there are two type of donation distribution scheme:
1. Software commune
Users may donate (monthly, yearly or one-off) to a pool, and then money is redistributed to the developers with a transparent formula (by installation count, user voting, number of developers, etc). I think it is in theory a fair system. While popular software developers receive greater reward, certain amount would be given to people doing software for minorities or researching in innovative ideas. This system encourage developers to devote to innovation instead of repeating some popular idea. However the biggest problem of this system is to avoid free riders and cheaters. We couldn't easily verify the contribution of developers when they are located in different part of the world.
2. Flattr for FLOSS
They may choose a fixed monthly donation and the money is divided by how much they vote / use a software. What made a different in this scheme is donor have the "freedom" to choose where the money go. This seems to be a more practical system.
Thoughts?
@edwincheese
>
hi!
>and if you see other ways in which we could make many cool apps compatible during the next six months or so, please also speak up. this is all still just a brainstorm, and i think it's not at all clear yet whether we should actually bring it into practice.
could you comment on
or
- so it would be easier to understand - what actually is in scope of
this market/app-store project?
PS: Awesome FLOSS Search is a must
- so it would be easier to understand - what actually is in scope of
this market/app-store project?
> 1. if the project means that Data separates from App;
means that - all the users data, all the valuable data that being used
and generated with App -
- being separated form the App as a Software itself
>App being hosted in proprietary even if Open Source Cloud;
after that separation - App still being hosted in the Centralized Cloud
that means some fast and reliable centralized cloud server - often -
VPS from a major corporation.
In such a variant - Hack/heist of Users Data could happen while it is
being processed on cloud server,
by administration, for example, maybe it is what could be called -
Man-in-the-middle attack.
Generally - for a "normal" usage - risk is pretty low for User and
pretty average for Developer.
I mean - developers code may be stolen easier than users data.
> Data being hosted locally, without manipulation - transferred with
> wise system (centralized anyway?) for stats of, for example - games; with manipulation if it isn't "games"
i mean - an example - where you are the developer of some Game(App)
that has Ranking Achievements system,
so - people play the App and developer wishes to Rank the Players. So
- the Player Data - during/after the process of Game
- being collected and processed into PlayerS Data (in this example -
Ranking data) - so, finally - we have a PlayerS Data somewhere.
Somewhere - means - in centralized "cloud" - where it could be, well,
obtained. Or in decentralized where it also - could be obtained :)
I think - in decentralized the bigger chance - if all such a scheme
would happen at all - it would be encrypted and thought through as
much as possible by these noble FLOSS security developers..
> (centralized any way?)
means - does the hosting of the PlayerS Data should be centralized in
best traditions of current gen "Cloud"
(you could review that PSN hack story as an example of how it
shouldn't be - as a result)
it was 1st variant, for which i had a question
> how Locker Project and other major competitors(projects,
> related systems that would work in the same field) - relate.
? you'v replied
> i would say, by making a person's Locker function as their personal data storage. Just like we already do (in experimental versions) with ownCloud and CouchDb.
thank you, i feel that from a more tight collaboration with Locker
team (pretty ambitious and professional team, as i understand) -
Unhosted could only win.
as for ownCloud -- it would mean the centralized and only centralized,
if something like Tahoe-LAFS or Camlistore or what else is there new
and promising --
if these projects wouldn't collaborate with ownCloud.
I need to say - i see enough of demand - everywhere for a solution of
user-privately or community -- owned Hosting, better if p2p,
decentralized, secured etc.
I want to conclude -- if there are a couple (or triple or some more)
of promising, moreover - serious FLOSS projects from one "sphere" that
everybody is waiting for - why don't collaborate?
ok, i'v described my opinion of situation, now i'll try to describe my
other words:
> 2. if the project going to let the App itself to be hosted with
> something like Tahoe-LAFS
> by that being a decentralized (as much as possible) hosting system
> with a front-end of Market(s)
That variant along with a 1st variant - of self-hosted User Data (if
we are dead serious for our Data and don't want it to be stored
in the Centralized Proprietary Cloud, ever) -- leads us into the need
of home-based-privately-owned Servers, at least for hosting
our own Data, if not for some Apps (Apps and real Hosting for Web -
may be unrealistic today - due to a poor bandwidth in many places,
but it could be changed in a near future, ofc.)
Ok, Private Server - means - let's say - a pretty popular variant of
FreedomBox (i think the reference is about 16gb sdhc storage or kind)
that - let's say - being extended
by theFNF.org and me -- into FreedomBases with some hundreds or
thousands of gigs. Let's don't talk about mesh networking, at all now.
So - let's imagine that we, finally, have a place - where to host and
with a help of Unhosted - we have - what to protect--host.
Here is that couple of variants i try to explain -- 1. variant where
Data being hosted on your and only your server, under the bad.
2. variant - is when - Data is being actively displaced in a
redundant, p2p, over-seeded manner across the neighborhood or city or
etc.
That means - you still - being able to use your data - fast enough -
and it means - if your personal FreedomBase is broken - you won't
lost you precious Data, am, your Life-Long Blog for example.
What i'm explaining - is that - 2nd. variant - if theoretically and
even practically possible now - gives amazing advantage and could
quickly
pay-off it's initial price.
Ok, let's imagine That - it is awesome, all works..
what we could do next...
Market!?
What could really drive an awesome Market that just another awesome
market - as for me - is an awesome currency, or awesome variability of
services that could be provided/exchanged/etc.
Let's imagine that all would work so well - that - well -- not some
billion-user portal, but some small and popular
Web Apps could be and actually - being hosted and/or Unhosted in that
FreedomBases.
Let's say some systems of "independent" digital concurrency or kind of
p2p Ratio Coins
or something - being in use within that systems, unhostments etc.
for such a situation i ask and think about
> decentralized (as much as possible) hosting system
> with a front-end of Market(s)
Front-end - means that i still follow the 2nd variant proposition, it
means that Apps being stored
in a decentralized manner, but still - you are able to provide Awesome
Market on top of that that,
how to make it happen, as for me?
> PS: Awesome FLOSS Search is a must
If combined - an Awesome FLOSS Market and Awesome FLOSS Search --
in a result could give the Awesome infrastructure for all the plethora
of Web Apps to be used.
-
What i feel is a situation where - rather better than later - the new
types of Markets
that being unhosted etc. -- would compete not only and first only -
service, UX, etc.
Let's imagine - the Developer would self host apps without any % rate
from a price for a distributor,
at all,
not a 20% not a 10% not even a 1% (well, ofc. there could be % and anything,
just - i don't like the % ATM :)
The Developer would receive the full - self-stated cost,
and Markets(that's all for Web Apps, i talk) would operate as a Search
portals -
that would, for example - show Ads of other Apps, while you Buy, or
receive the funding in other ways.
They could receive the money from Developers - obviously, but not that
% gang-bang, i hope.
Instead - App Markets and FLOSS OS's mobile and desktop - would
compete on UX, on clever ideas,
on something new. Well, that's a kind of my dream.
> I'm not sure what you mean by that. you mean the app store should have good categories, tags, and free text search, so that it's easy to find the app you're looking for?
yes, first of all - the Web Apps should have that tags, categories >> semantic.
There are man groups - still working for Semantic Web, in one or
another way - there are some solutions, now.
> i would say:
> - categories of apps
Search that works on advanced Tag system.. tags that are categories
among other. (it is a very advanced topic,
needs a separate discussion)
> - mirroring of app source code
yep, decentralizing, it could be done with 1. and 2. variant but i
propose the try on 2nd from a start, by this -
helping the FNF/FreedomBoxFoundation.org essential Freedom needs and
fulfilling the commercial potential,
that, in my opinion would be far more fair for developers and by that
for customers.
> - a dashboard / launch pad / start menu where people see their installed apps
yep, awesome UX (i prefer the speed-centralized, though there are many
valuable variants), beautiful UI.
> - sponsored links (whose revenue would go partially to running the store, and the rest to app devs)
variant
> - eventually, a code editor so that you can edit apps from right inside the store
http://ace.ajax.org/ i think - the most promising and works right now.
> - in a next version of our protocol, i think control over the data scopes could be moved from the storage to the store(s)
hm.. decentralizing?
> i think we should always develop the protocol on the basis that there will be multiple app store. but at the same time we'll make sure there is at least one app store, > and that it's awesome. it should be accessible at least over normal http, and if possible also through the content-addressable mesh-routing you propose (if i understood you correctly).
+1, that's what Mozilla's Open Web Apps are about on all fields.. but
without a canonical store :)
PS: i could try to describe it in another way, i understand that my
English syntax isn't easy for the eyes now..
i understand that what i'm talking is rather optimistic but -- i see
the whole new stack of related projects from serious, respected
"hackers",
that are rather optimistic - still - run fine to the moment.
PPS: if anybody wants to help or discuss FreedomBases - please - join
the http://freenetworkfoundation.org/?page_id=275
mailing list and start a topic.. seems like the
http://freenetworkfoundation.org/pipermail/discuss_freenetworkfoundation.org/
is dis-functional now.
Practically - all the possibilities of FreedomHostings kinds are being
evaluated over the
http://wiki.knownelement.com/index.php?title=Data_Ownership work that
is currently in a process of trying the FBX(FreedomBoXes) vision, some
info - http://freenetworkfoundation.org/?page_id=9
@FNF -- I just - keep forwarding the messages i consider as important,
for a whole story of this discussion - please join or observe the
Unhosted Google Group.
Yes, i'v missed that part, though - i'v seen the parts of that
discussion some time ago.
I think - it is a great idea. For sure - such an apps and developers
which would provide the possibilities like this -
at least would lead to "Leaving the Cloud".
If it is a main and principal purpose of Unhosted,
that is cool - way to anti-ChromeOS and in way of Mozilla B2G,
i, personally - very glad that personal, stand-alone computing is still
evolving so good, despite all.
I have posted to Unhosted list about XCCC networks and one of the conditions
was - distributed computing, and i still see it as a crucial state,
without which there
wouldn't be a nice Freedom Networking.
XCCC work leading to FreedomBases as a variant of silent, power
efficient, self-owned
computing,
as for
*silent - it is important for a 24/7 operation, obviously,
*power efficient - see - the
http://green-broadband.blogspot.com/2011/06/could-net-be-killing-planet-one-web.html
for an anti-thesis to the common opinion.
*self-owned - a matter of taste, at least.
So - what's i'm for when talking about Distributed Computing, briefly:
1. You send the data to compute and receive the result..
what may it be? i use the term "compute" here as a term
to describe the any computation.
By that - computing on servers that you mean like an actual
computing of server-side code and computing of your data
on other kinds of servers that may do the computing like indexing
the crawled data for a FLOSS Search.
It is not a simply grid/non grid computing,
rather a mixed solution where the computing could
be done with a cheap, yet efficient portable computers,
think of modern mobiles stack - that being used for FBX's etc.
By that - your mid-range super-powered mainframe with a 600W+ supply
stays for desktop work and staff and your FreedomBases, for example, act
as a home media center, home server, whatever you could imagine for the powers.
What i'm sure is - without FLOSS Search - not a one of Own-Data
project would it be Apps
data of just some home archive -- won't succeed.
To power the Search as a crawler of let's say - PubSub or PuSH or some
other real-time
stream.. to make it less power consuming to transfer the world
crawling to subscribing with
pushing -- i think - a kind of FreedomBases as a Match-box sized
computer with a stack of HDDs,
!could be - enough. Bandwidth is "problem" but in my country -
100mbit is ok for UA-IX (a kind of Ukrainian
Internet see - http://www.ua-ix.net.ua/eng.phtml
http://ua-ix.net/about_en.phtml)
in other places - there would be better bandwidth -- sooner or later.
But -- to maintain the whole new freedom of Internet -- actual grid
computing for a kind of usual all-crawling
or for other computations - would be needed. That's what i believe,
ofc. - i understand that all this could be
used for brute-forcing, bot-gridding, whatever. But, given the WebGL
and future WebCL in a millions of computers(billions ofc.)
that would be inevitable - anyway. The way to make that kind of
evolution into the evolution for people -
in providing the ability of people's control over the, well, let's
call it resources.
With a clever management, with a Manifests, Ideas -- resources would
be.. at least, have an ability to work Globally,
for benefits of humanity or world at all, i hope.
Without - resources of such a networks for good or bad - would be
gathered alike they are
being gathered now - by Corporations, but only "alike" because it
would be Countries.
I mean - it would be done - anyway. Now you could join the
http://www.w3.org/community/groups/proposed/#cloud
Cloud Computing Community Group (just for example) and try to
make/remake it into Web Standards,
in some bad variant of future - it could be worse for You as an end-user.
My opinion based on experience is that -- you may pay some Corp for
their inland computing and buy a cheap
"screen" in addition or - trust your self and buy a not-so-cheap
server and, maybe a screen for it --
all lead at least into a comparable sum of cost.
I say - in one way - people pay for someone's property in other case
for their own, the result is pretty the same.
Ok, i have disrupted the way of telling a little,
i'll try to fix.
here is the
2. You receive the data to compute and send the result
it is a kind of torrent evolution that could bring exciting new world of
values - based on the people's need. Nice example for me is -
https://secure.wikimedia.org/wikipedia/en/wiki/Parallel_rendering
Renderfarm.fi
excuse me,
i need to go now,
Michiel,
what i wanted to say is that
the projects should collaborate
not only - "i take Tahoe, i take Box,
mix, bump, wow"
The work here and there as i understand
is for a better future, not for a wow product,
i see theFNF.org as a place where all our projects
could be mixed in something - helping the Internet
and humanity at a whole, but the current conditions
is - the later projects start to collaborate, the lesser
time people would have in result.
If you see better places, please - elaborate.
I can't think about ANYTHING more until I have at least 2 apps in a
"market". Let's put our words where our mouth is.
So we have My Favorite Sandwich. Great. Simple. Useful? (no)
So, something useful for me and others on this list, I'm sure.
A non-sensitive data app ( no credit cards ) but still dealing with
personal data & privacy aware app would be...My Birthday Reminder. It
is branded "Datahope.org"
It stores my Friends names and their Birthdays and reminds me via
whatever methods (email, SMS, etc).
I want to prove the unhosted concept by moving My Birthday data
between 3 unhosted storage providers for a week. Moving it to each
one every day. We need this type of testing. Proof. Validation.
The author decides on day 3 that the app has forgotten something very
very critical! The friends' email address of where to send their
birthday card! HORROR. The author quickly needs to add the
functionality and let all his users of the My Birthday Reminder app,
know of the new version, how does the push happen? is there even a
need for a push? (hmm), that allows the new field to let users add the
email address. Does the storage provider automatically extend
additional rights for this new Key-Value being stored? Should it even
care with it's contract with the user's data? What if this Value
being stored exceeds the user's contract of 1 Gig of SPACE? What
checks and balances does the unhosted storage provider really care
about ? Maybe they don't ? Maybe they just have "X AMOUNT OF SPACE
FOR YOU FOR WHATEVER"...blah blah blah
Maybe an even better app test would be...My Test Reminder...(Test
Name, Date, Teacher/Professor, Subject) for those 1 million upcoming
Students ?
"Quit jiving and play"
> i would like to know what other people think of it.I can't think about ANYTHING more until I have at least 2 apps in a
"market". Let's put our words where our mouth is.
[...]
"Quit jiving and play"