Touchstone can be extended in a number of nifty ways - they are:

• Input Adapters - get data from different data sources. Usually a .NET Project (e.g. The Feed Reader)
  
• Output Adapters - display data in various ways. Usually a .NET Project (e.g. The news ticker or popup alerts)
  
• Subscription Plugins - Little bits of XML (found in the \subscriptionplugins folder) that add tabs to the Subscription Helper so that you can subscribe to sites easily. Even sites that don't provide RSS by using Dapper (E.g. LinkedIn, Facebook, Myspace etc)
  
• ActionExtensions - Little bits of XML (found in the \ActionExtensions folder) that allow you to send items to various social sites (E.g. Digg, Twitter, Tangler, Del.icio.us etc)

Learn more about them on the wiki at http://touchstone.stikipad.com and suggest ideas for extensions here!

What was your question about Tangler again?

does Particle support Tangler ? to keep me informed about new plugin development

oh right 

No the tangler guys have not put RSS in yet

And their pages are not static so we can't even monitor them for changes 

but they assure me they will soon

is there any wiki for developers as it was with Touchstone ?

not yet - that's my next stop

interested in helping to write it?

will try to write my findings. but I needs some info to start up and not relearn from scratch

there is still touchstone.stikipad.com 

but it is way out of date unfortunatly

And posts like this on the dev blog http://community.particls.com/blogs/circle-of-stones/archive/2007/03/12/Making-Subscribing-Easier-_2D00_-Touchstone-Subscription-Helpers.aspx

Just set up http://particls.pbwiki.com/FrontPage

I am still getting failed to scan 'because cannot find Application Data\Particls\nlp.tdat'

what is that file ?

That is a file that helps with the indexer that scans your PC for your interest

interests*

we are looking into it for you now 

is Particls downloading again all files ? as I found some in "\Local Settings\Temp"

How do you mean downloading all files? 

saw some .zip and .exe that I downloaded in the Temp directory

Sorry still not clear what you mean...

for example, I have in temp dir a tmp84F.tmp.exe, a divx web player that I dowloaded today

Ok not sure about that - will investigate 

what is P workflow ? is it gets the history list from Firefox and redownload and checks for rss / text ?

That is basically what the Autosubscriber does (put simply)

that is 'Autodetect my interes', right ?

It is one aspect of "Automatically Find information for me"

It does other things as well

The NLP.TDAT error should go away over the next hour or so now - fix has been applied

should I redownload or it autoupdates itself ?

the patch was applied server-side

so should be automatic

nice

what shortly means in "Feed Adapter will try to subcribe to URL shortly" ?

It means that it has detected you visit that site fairly regularly and it will try to subscribe to it for you so you can be updated of changes

it will get more accurate as time goes on and it gets smarter by learning more

actionextension are for newsticker items right click ? as I dont have any item now I cannot check

yep 

and the alerts as well

they are another way of extending Particls without .NET

u can digg right from inside Particls - just right click and away u go 

to create like Growl notification system we need to make an adapter right ? using subplugins is not possible ?

depends - what do u want to notify

if u can reach it from a URL + XML return - u can do it with a subscription plugin

if it is more complex than that you will need to write an input adapter right now

for example long downloads

Need more info

most of downloaders have antivir check, so we can call a proggie that injects an item in P stream

oh yeah - for client-side applications you will need to do a .NET input adapter

but I would like someone to make a .NET input adapter that makes it easy to inject items into it - perhaps one that just looks for some XML files in a directory 

so any app can just leave a bit of XML in a directory and the adapter will pick it up and inject it in

if that makes any sense

that will be nifty.

pitty that it must be .net

tryin to get Google Reader's shared RSS feed in P - no luck. just empty.

Im thinking we start with something simple 

where we display the headlines and descriptions down the page in a cool way, but also factor in the rank tag to vary the size of the posts

and add the words that decided the inclusion

so, we go in the input or output adapter first ?

well that information is not available in the RSS feed yet 

the newspaper would technically be considered a 'visualization of the pebbles rss feed'

output adapter if anything

how about making the Pebbles URL without the email in the plain, like particls.com/pebbles/rss/ABCDEF

where ABCDEF is a hash of the email

dont like to have my email address all around

yeah - we will change the pebbles RSS feed URL soon 

yep - I agree with that one

been on the plans for a while

just make it possible for the URL to be variable for now - sand we will change the pebbles RSS feed URL Convention ASAP

what I dont like at this approach is that the stream is goin from my computer to particls.com than back for processing

double the data sent

how about sending it to an .xml that we can monitor and work on that

well the newspaper needs to be a hosted solution - not local

why not start with something that we can do local ?

because part of the reason to do this is to create a newspaper that is accessable from work generated based on data from your Particls at home

or vice versa

yeah, but that means to have one computer running all the time to fetch the stream

to create the stream u mean? 

yes

Dont worry we have solutions for that coming

I have several computers running at home all the time

but i am not your average user

ok, Particls will have all my visiting behaviour

visiting behaviour? 

what URL i visit, my white/black list, ....

Particls the client knows what URLs you visit and uses that to work out what you care about (unless you turn that feature off)

But Particls the company/business has no idea 

it is not sent to our servers

We do measure clicks to items via redirects - but only if anon usage data is turned on in the settings - and it is anon (as the name suggests) 

ok

how can I test the Pebbles - to have items in the stream so I can have the rss feed ?

just turn it on

in the advanced tab turn it on (click the pause icon to change it to the play icon) 

and it should soon have data in it

and maybe lower the threshold to make sure it gets items

"lower the threshold" ?

if u click on Pebbles - under the heading it has  a pulldown saying 'only very important items'

lower it to important or most

just so u get plenty of items to test with

put send all items - nothing

it will

restart to help speed up the proccess - and it will still take about 5 mins to push the items to the server 

it does it every 1-5 minutes

i have to crash mate - looong day - it is 3am here

catch u in about 8 hours or so

good night

20:00 here

oops. I change to send all items to Pebbles and gets reversed to Send only very important

bugger

anybody can confirm this bug ?

can u recreate it ?

yeah. everytime when I change. restarted Particles 3 times

same thing

testing it here as wll

hmm, now it stays on all items. must wait to push the stream on the site. .zZ

its working fine for me here -

set to all items - close particls - open particls - still set to all items

I am going to have to crash aswell I am in the same timezone as Chris so its 3:00 am here

ttyl

cyaz

back to development page

let me know when u need to know those 3 commands (opml import, add watch word, mark item as read) 

opml import <filename>, add watch word <word>, mark item as read <URL>

yep 

more info about GR API - http://code.google.com/p/pyrfeed/wiki/GoogleReaderAPI

seems it is possible to mark as read items in GR too

bummer :/ 

we should call them out on it when we announce this tool 

yeah. be right back. must go baby food shopping

haha ok cya mate

back

wb

was thinkin about GR integration. why not get the fresh RSS feed, enter the items in the stream and if they are read in PS mark them as read in GR

fresh RSS feed?

And i thought you can't mark as read in GR? 

GR fresh = " When a new item of one of your feeds arrive, it's labeled as fresh."

I said that we can makr as read in GR

oh you CAN mark as read in GR? sweet! 

oh yeah

check this out - http://paul.totterman.name/darcs/maemo-google-reader/maemo-google-reader

yes that is perfect as well

the only way to tell when an item is read in PS though is to write an output adapter that traps the 'mark as read' event 

you can make a GR input adapter in Ps in notime

but we are hoping to extend pebbles to generate an RSS feed of read items

hmm, kinda complicated. why not make an adapter that syncs with GR based on the unread items from GR

up to you 

Im not 100% technical btw - for that you will have to talk to Ashley my business partner and CTO of Faraday/Particls

but i get the conceptual stuff as u know

But I can get ash in here when/if you need

I added him as contact in Tangler - no confirmation yet

yeah he is super swamped atm

he jokes that we keep him in a box

the idea is not have another library for HTTP get if you already have it in Ps

yep thats cool - well u can catch mark as read events via an output adapter

as long as u can work out how to make one with ruby

maybe bundle ruby runtime library and allow plugins written in ruby ?

what I also like is Rebol - designed for net access, all is included in ~300kb

found http://google-reader-notifier-windows.googlecode.com/svn/trunk/GoogleReaderNotifier/GoogleReaderNotifier.ReaderAPI/ - maybe you can use it somehow

it is .NET

ready a test version that gets the stream of unread items from GoogleReader (14kb unoptimized exe, no other .dll needed)

Cool - its nearly midnight here and all the developers have gone to bed

Hi TBD - sounds awesome - what can I do to test?

I will make today a command line version that takes the user/pass and outputs the xml of the unread items from GR

the user/pass for GR? 

yes. otherwise how to access the GR ?

Yep thats cool

And outputting to XML on the client-side I guess

yes. GR gives an atom feed that I can think it can be mixed in Ps stream

well to actually send 'mark as read' commands to P it will need to be a series of shell commands

e.g. particls.exe mark_as_read [parmalink] 

I think it needs to be other way round - what is read in P to get flagged as read in GR

as with my proggie you get the stream of unread items from GR

oh i thought u were doing both ways

think is possible to do that also - but how ? you check all items from GR for read state and then send that data to P. seems overkill

sorry the Atom feed you are getting from GR has what in it? Fresh Items?

fresh unread items

ok but for that you don't really need a special app do you? You can just put that atom feed into the normal feed reader in Particls 

In my mind I am thinking you can do it like this:

1. Plug the Fresh Items Atom Feed into the Particls Feedreader (by itself)

2. Write an app that gets a stream of items you newly READ in Google Reader and Marks as them read in Particls (via command line thing we have)

3. Write an app (maybe the same app) that gets a stream of items you mark as in Particls and sends the Mark As Read command to GR (hopefully this wont create a feedback loop - Particls should NOT resend 'mark as read' events for items that are already read)

That's complete syncronization 

1) you need to make the auth to read the GR feed

if it's a normal authenticated feed - particls should be able to handle that

2) how to get the newly read items ? based on a time frame ?

you need to go to https ServiceLogin, send the pass/email and then you get back, must get the user ID to be able to send commands

HTTPS should not be a problem for the reader - if there is a problem that's a bug

You get newly read items from Google Reader - it has an API call for that I thought?

http://www.google.com/reader/atom/user/-/state/com.google/reading-list?xt=user/-/state/com.google/read

you need to be logged in, otherwise /user/-/ is not working

ok - well there u go

so that's where the challenge is - point 2 and 3 - getting the 'mark as read' commands from GR and Particls and sending them back and forth 

But fresh items in an Atom file should be handled by Particls itself 

yes

just need to mark as read in GR to not have duplicate items

yep 

then from time to time get the unread items from GR - the others items that are not there are read in GR

that would be sweet

so mark them as read in P as well

for that you need to add in db a field that the item comes from GR adapter

why?

just send the mark as read comand to particls - it looks after the rest

ok. how do I know then that an item is read in my adapter ?

to send it back to GR?

to send to P

lol what!

are we talking about point 2 or 3

2

Ok.. so you use the GR API to get a stream of items that have been marked as read in GR?

And for each item - you send a 'MarkasRead [permalink]' command to Particls 

possible

that's what I thought u were doing in the first place lol

nah, I was getting the unread items from GR, then to send to P

ok - but I think particls can do that already 

the hardest part is making it easy but we can probably do that with a SubscriptionPlugin - whch is easy bit of XML

getting the feed with Google auth

btw, the feed of read items is like this - http://plugins.ro/read_items.xml

LOL at your last read item

the Can't get Google Earth in Sudan? Try Tor ?

released today - http://code.google.com/apis/ajaxfeeds/documentation/ - maybe it will help someone for getting feeds

<div class="author">Jaykul</div><div class="msgAttribs"> Today, 02:07 PM</div>

By the way, you really need to like ... auto-respond when people say they want C or Ruby:

C++ is a first-class .Net language, you've always been able to write
.Net apps in C++ (yes, C++, not C, but hey, the API is object oriented,
so that's that).

Ruby has several "bridge" projects and there's even a Ruby Compiler for .Net available from Gardens Point:

• The compiler can be used to statically compile a Ruby source
  file into a verifiable .NET v2.0 assembly or it can be used to directly execute
  a Ruby source file (compile, load and execute). The implementation is not yet fully
  complete, but it does pass all 871 tests in the samples/test.rb installation
  test suite of Ruby 1.8.2.

It's coming along nicely.

saw it. already know about them

but we need docs from Particls and some examples to start

Yes i agree - I have started cleaning up the wiki today - and asked the developers to join in

Can somebody help me with this Del.icio.us subscription plugin?

I tried coding and testing but nothing happens.

Here's what I have so far:

<?xml version="1.0"?>
<particls-subscription-plugin
    unique-id="Delicious"
    name="Delicious"
    description="Delicious Search">
  <variable key="tag" name="Tag" type="string"/>

<feed id="Items" url="http://del.icio.us/tag/${tag}" />
</particls-subscription-plugin>

Thanks