#social 2015-06-24

2015-06-24 UTC
jasnell joined the channel
#
aaronpk
even twitter didn't start out with a good design and they quickly iterated and have a very scalable system now
tantek joined the channel
#
melvster
tantek: The way the Web spread was a piece at a time. So you could take html without taking http. So the failure of NEXT was a lesson, don’t try to sell it all at one time. Sell each piece on its own merits. Never insist that everybody take all. They will take all the pieces once they see how it fits together.
#
melvster
--Tim Berners-Lee
#
melvster
one of my favourite quotes
#
tantek
agreed - and yet, you now see people advocating "LDP all the things!"
#
tantek
the "piece at a time" is key, and what Tim also said was interoperability with what already exists (is deployed, in use, etc.)
#
melvster
tantek: LDP is just standardizing PUT for data in a file like way ... it's very close to the web already, and just scratching an itch for those that wanted to use data
#
melvster
LDP is the ultimate example of iterative work
#
tantek
melvster - the specs don't make that clear at all
#
melvster
tantek: yeah, you know how specs are ... design by committee
#
melvster
but it's a very simple idea
#
tantek
all the specs like that are very "big vision wholesale"
#
tantek
if it was a simple idea, it would be widely implemented
#
tantek
thus I reject that assertion
#
melvster
just take UNIX file system and webize it, webdav is too complex just allow simple reading of writing to files and directories
#
tantek
in contrast, indieweb building blocks really *are* simple, one page-ish specs each, each one has its own "Why" and "How"
#
tantek
and they build upon each other incrementally, each bit gives you a useful capabilities/features
#
melvster
tantek: i rejected the semantic web originally because of this, i found the specs too much work to get through, but when i finally set aside time to read them, I realized they were much simpler than I thought, and I was hooked
#
melvster
tantek: yes this is good, but you also need some of the features that give the web scale
#
tantek
yes I've pretty much given up on "specs too much work to get through" because it doesn't matter if I (one person) implement something, it matters if dozens, hundreds of people can
#
melvster
many systems build incrementally, most systems in the enterprise, for example, build incrementally, but many become monolithic
#
melvster
tantek: i think you're right, that's why I try and show people through demos, rather than, specs now, then people can ask "how did you did that"
#
tantek
melvster, most systems in enterprise are overbloated day 1
#
melvster
but the point in general, is that most systems built incrementally, dont scale
#
melvster
the web is a rare exception, and it helps to know WHY it is an exception
#
tantek
disagreed. all successfully scaled systems were built incrementally
#
tantek
e.g. Twitter, Facebook etc.
#
melvster
facebook is actually quite well engineered
#
melvster
because mark zuckerberg didnt try and do anything clever just K.I.S.S.
#
melvster
well anyway, the web scales because it is resource oriented
#
melvster
some physicists did a study on it
#
tantek
melvster - no he / they rewrote pieces as necessary
#
tantek
the key is they shipped, and iterated
#
melvster
the magic property of the web is namespaces which leads to URIs
#
tantek
rather than "architected for scale" at v1
#
tantek
nah, 99.999999% web cares zero about URI namespaces
#
tantek
HTML tags aren't namespaced, nor attributes
#
tantek
so it's not magic, nor necessary, nor in use
#
melvster
but hyperlinks are
#
melvster
hyperlinks are what gives the web its magic
#
tantek
sure, URL are useful, domain names are too
#
tantek
but they're just strings, not namespaces
#
melvster
they are both
#
tantek
no. they are only what they need to be. just strings and addresses to retrieve stuff.
#
tantek
nothing more
#
tantek
if you are assuming anything more, you're just making stuff up beyond their function in practice
#
melvster
they are unique
#
melvster
and also universal
#
melvster
what the ietf calls 'uniform'
#
melvster
even ted nelson says the greater invention of the web was the URI
#
melvster
that's what linked data is based on
#
melvster
bottom up design
#
tantek
no linked data is based on RDF, simply rebranded because RDF (model etc.) was overhyped for so long that it people became immune to it.
Loqi joined the channel
#
melvster
I see the confusion with HTTP
#
melvster
It's both a name / pointer AND a dereferencing protocol
#
melvster
it is a confusing thing
#
melvster
but understanding that helps to interoperate
#
melvster
the nature of the name is that it's hierarchical so can operate as a namespace
#
melvster
not just on the web, it's used this way in, for example, java libraries
#
melvster
what helped me understand this was when I learnt the web was designed to be like a piece of paper, the paper doesnt care about what is written on it
#
melvster
you could call it web neutrality
almereyda, bblfish, ben_thatmustbeme and rhiaro joined the channel
#
melvster
rhiaro: I sent a mail to the list regarding profiles
sandro, bblfish, cwebber2, AnnB, jasnell, jasnell_ and Lynn joined the channel
#
AnnB
hmm .. wondering if anyone is going to attend IG meeting
#
AnnB
starting now .. but it's only me
#
AnnB
and none of them are here in IRC either
#
Lynn
hi ann, i'm installing webex (sorry, not prepared)
#
AnnB
I see you!
#
AnnB
cannot hear you tho
#
ben_thatmustbeme
yeah, it takes a minutes to get webex going
#
ben_thatmustbeme
AnnB, to help others there should be a URL for the meeting you can share in there
#
AnnB
no worries
#
ben_thatmustbeme
s/there/here
#
ben_thatmustbeme
I'm not going to be on the call, but i'm always here in IRC
#
AnnB
there's one in the meeting notice
#
AnnB
thanks for doing that Ben
#
AnnB
for the record, only Lynn and I are attending IG meeting
#
AnnB
but we're having a great chat
#
AnnB
we think good roles for IG are:
#
AnnB
A) continue to develop vocabularies
#
AnnB
B) continue to develop / document use cases
#
AnnB
Lynn is working (with others) on Open Value Networks
#
Lynn
Open Value Networks are one form of "next economy" economic form. One example is Sensorica in Montreal, Canada - http://www.sensorica.co/. They produce open source hardware and operate as a network of peer to peer producers.
#
Lynn
(We write the software for them.)
#
AnnB
I see: Value Network description: https://en.wikipedia.org/wiki/Value_network
#
AnnB
"... a business analysis perspective that describes social and technical resources within and between businesses."...
#
Lynn
Here is one vocab effort, which is moving VERY slowly. https://github.com/openvocab
#
AnnB
so, the gist is, in order to have these types of business networks, there need to be common vocabularies
#
Lynn
ann i lost the call sound
#
AnnB
lynn, I hear you ..
#
Lynn
i hear you now
#
AnnB
ok ..
#
ben_thatmustbeme
pictures those old verizon commercials "Can you hear me now?"
#
Lynn
ha
tantek joined the channel
#
AnnB
cool.... Lynn and her partner Bob working on: http://mikorizal.org/
#
AnnB
(a play on the English word "mycorrhizal " .. hehe)
tantek, almereyda, jasnell, bblfish and oshepherd joined the channel