stebalien changed the topic of #ipfs to: Heads Up: To talk, you need to register your nick! Announcements: go-ipfs 0.4.22 and js-ipfs 0.35 are out! Get them from dist.ipfs.io and npm respectively! | Also: #libp2p #ipfs-cluster #filecoin #ipfs-dev | IPFS, the InterPlanetary FileSystem: https://github.com/ipfs/ipfs | Logs: https://view.matrix.org/room/!yhqiEdqNjyPbxtUjzm:matrix.org/ | Forums: https://discuss.ipfs.io | Code of
thomasanderson has joined #ipfs
soju has quit [Ping timeout: 268 seconds]
thomasanderson has quit [Ping timeout: 246 seconds]
lordcirth has quit [Ping timeout: 245 seconds]
soju has joined #ipfs
gnufr33d0m has quit [Quit: gnufr33d0m]
mowcat has quit [Remote host closed the connection]
soju has quit [Remote host closed the connection]
_whitelogger has joined #ipfs
soju has joined #ipfs
pecastro_ has quit [Ping timeout: 276 seconds]
tec__ has quit [Remote host closed the connection]
woss_io has quit [Ping timeout: 240 seconds]
RamRanRa has quit [Read error: Connection reset by peer]
<postablesDiscord>
The main problem is that the organizations you pointed out are already deeply invested in IPFS. Outside of those already invested in IPFS either directly or indirectly, the feelings around IPFS are not that great.
<postablesDiscord>
There is very little incentive for those not invested in IPFS to pick it up simply due to the ginormous headaches present when using IPFS, especially so at scale
i9zO5AP has quit [Ping timeout: 265 seconds]
<postablesDiscord>
For example I went to a conference in August. Out of the 6 people I met that had tried IPFS in the last few months not 1 had a good thing to say, not 1 was still using it
<momack2[m]>
hmm, hard to diagnose without knowing the context - maybe they wanted something that IPFS isn't right now, or weren't able to understand how to apply IPFS to their problem effectively.
<momack2[m]>
I haven't had that experience at conferences at all - but have talked to frustrated people who get sold on IPFS as magical free storage, or who get lost in our documentation and end up giving up
<momack2[m]>
I think there's a lot of room for improvement there too - IPFS needs to set good expectations for what it can do, and then make it really easy for everyone to meet those needs (ex better golden paths / how to guides / etc)
<momack2[m]>
feel free to DM me any specific stories / groups you talked to so I can help follow up with them or at least try to glean more takeaways from their feedback
rendar has joined #ipfs
<momack2[m]>
and if you have ideas on how to put that golden path front and center, check out the documentation revamp we're doing and help us prevent the misconceptions/frustrations from happening in the first place! --> https://github.com/ipfs/docs/
<PhoenixDiscord[m>
I feel like if IPFS could just reliably resolve IPNS names in a few seconds, IPFS would be so easy to use for so many things. Developers can get around a lot of overly complicated systems if they work. Look at docker! I still don't know how to use docker, i ask a friend to help me whenever i need to change how a docker image is working. But it's really hard right now when IPFS doesn't have any internal reliable way to
<PhoenixDiscord[m>
share a mutable folder as something like a URL.
<momack2[m]>
ack - actually some IPNS improvements coming in the next release (both via improving DHT lookup times, and adding IPNS over pubsub)
<PhoenixDiscord[m>
DNSLink can get the job done, but it honestly sucks, because to actually make use of it, you have to figure out how to get your nameservers to talk to a script that builds your IPFS data, and since so many nameservers have different APIs for a lot of users if you just use ones provided by your domain registrar, there's no way to even write a simple clear guide on how to publish mutable data on IPFS reliably, because
<PhoenixDiscord[m>
there are so many different interfaces to get a DNSLink thing updated
<PhoenixDiscord[m>
I'm a bit confused about how IPNS over PubSub works? what does that improve? Will there be new interfaces so I can subscribe in my app and be notified when an IPNS address is assigned to a different CID?
<postablesDiscord>
@momack2 I'm about to head to bed right now, however I believe you have my email. Send me a message and we can schet a call
<postablesDiscord>
* @momack2 I'm about to head to bed right now, however I believe you have my email. Send me a message and we can schedule
<momack2[m]>
Phoenix (Discord): aschmahmann is the expert, but AFAIU yes - effectively you can "subscribe" to subsequent updates to an IPNS record and not have to do a DHT query for them
<momack2[m]>
postables (Discord): SG! will send you a note
reit has quit [Quit: Leaving]
<PhoenixDiscord[m>
Oh that's really great. I look forward to it! Substantially changes my thinking of how to implement my toy app
<ShokuninDiscord[>
To echo on the above: there is an overwhelming sense that people simply do not understand IPFS. It's concerning to see people post _daily_ about how it it a magical cloud/dropbox solution. I feel like a lot of the communication around future-proofing access to knowledge ("Today's web can't preserve humanity's history") are to blame.
<ShokuninDiscord[>
* To echo on the above: there is an overwhelming sense that people simply do not understand IPFS. It's concerning to see people post _daily_ about how it is a magical cloud/dropbox solution. I feel like a lot of the communication around future-proofing access to knowledge ("Today's web can't preserve humanity's history") are to blame.
<PhoenixDiscord[m>
I actually think a really useful service someone could make, which I would adore, wouldn't really involve IPFS at all. It'd be a subdomain provider, it would have a simple API to register a "username" at one of a few domains on offer, and perhaps with a proof of work, would allow any user to claim a subdomain, and point it to whichever gateway and DNSLink they like, and update it as needed with cryptographically signed
<PhoenixDiscord[m>
update notifications pushed to the centralised server
<PhoenixDiscord[m>
one of the few areas where "make a website on IPFS" isn't free right now, is getting a domain name. but it wouldn't be very costly to run a service like that, I think?
charley_ has quit [Remote host closed the connection]
<vaultec81[m]>
Phoenix (Discord): Another thing that you forgot to mention, there is no mutability, or history tracking for current day IPFS websites. The only thing you can even come close to is IPNS.
ygrek has joined #ipfs
<vaultec81[m]>
There needs to be a good standard for mutable content in IPFS that actually suites the needs of websites.
<aschmahmann[m]>
The changes make IPNS over PubSub work without needing a DHT (e.g. using MDNS on a local network), and in the event that it utilizes a DHT to find out about peers it is **much** faster than how IPNS currently uses the DHT. The end result should be IPNS resolution times that are just as fast as IPFS resolution times instead of much slower.
mapetik has quit [Remote host closed the connection]
mapetik has joined #ipfs
arch-nemesis has joined #ipfs
cxl000 has joined #ipfs
cxl000 has quit [Ping timeout: 276 seconds]
aldenp has joined #ipfs
hqdruxn08_ has joined #ipfs
<aldenp>
hello; when files are uploaded, they are broken up into chunks, represented as a DAG, and put into the DHT right? does the DHT track peers who have a given block of the DAG, or does it track the payload itself?
hqdruxn08 has quit [Ping timeout: 246 seconds]
<aldenp>
what I'm getting at is A) how is data lost in IPFS, since if the data itself is stored in the DHT then any data every uploaded should be being served by k nodes at any given time B) how are nodes which have a copy of the blocks through pinning rather than being (pseudo)randomly selected by the DHT found when fetching a block?
<aldenp>
sorry if what I'm asking doesn't make much sense
<aschmahmann[m]>
aldenp: The DHT a map of `hash(data) -> list of peer addresses the advertise they have data`. IPFS calls this list of peer addresses provider records.
<aschmahmann[m]>
This means that A) random nodes on the internet don't have to provide huge free storage B) data will go offline when nobody is advertising it any more
<postablesDiscord>
Data is lost in IPFS when the nodes storing it are offline. The data isn't stored on the DHT. The DHT is basically used to store pointers to where the data is actually stored
<aldenp>
ah, thanks, that makes sense
<postablesDiscord>
Whenever you add something to your ipfs node, with the default settings your node will create a provider record that indicates that your node provides some content. This record is then published to the DHT
Jybz has quit [Excess Flood]
Jybz has joined #ipfs
robcat has joined #ipfs
noaht[m] has joined #ipfs
easyKL has joined #ipfs
arch-nemesis has quit [Ping timeout: 246 seconds]
null1337 has joined #ipfs
blablok[m] has quit [*.net *.split]
opt[m] has quit [*.net *.split]
wzwldn[m] has quit [*.net *.split]
silotis has quit [*.net *.split]
wchargin[m] has quit [*.net *.split]
eshohetDiscord[m has quit [*.net *.split]
nofwayyDiscord[m has quit [*.net *.split]
cesarosumDiscord has quit [*.net *.split]
jryans has quit [*.net *.split]
Technoquake has quit [*.net *.split]
Shmultz9776[m] has quit [*.net *.split]
modul8[m] has quit [*.net *.split]
freddor has quit [*.net *.split]
test123Discord[m has quit [*.net *.split]
leoalvarezhDisco has quit [*.net *.split]
bekoDiscord[m] has quit [*.net *.split]
Markus72[m] has quit [*.net *.split]
peddie has quit [*.net *.split]
kbc[m] has quit [*.net *.split]
itsundef[m] has quit [*.net *.split]
Myrth has quit [*.net *.split]
blablok[m] has joined #ipfs
opt[m] has joined #ipfs
kbc[m] has joined #ipfs
Technoquake has joined #ipfs
Shmultz9776[m] has joined #ipfs
freddor has joined #ipfs
Markus72[m] has joined #ipfs
wzwldn[m] has joined #ipfs
wchargin[m] has joined #ipfs
itsundef[m] has joined #ipfs
nofwayyDiscord[m has joined #ipfs
eshohetDiscord[m has joined #ipfs
cesarosumDiscord has joined #ipfs
leoalvarezhDisco has joined #ipfs
test123Discord[m has joined #ipfs
peddie has joined #ipfs
jryans has joined #ipfs
silotis has joined #ipfs
bekoDiscord[m] has joined #ipfs
Myrth has joined #ipfs
modul8[m] has joined #ipfs
bekoDiscord[m] has quit [Changing host]
modul8[m] has quit [Changing host]
modul8[m] has joined #ipfs
bekoDiscord[m] has joined #ipfs
freedom has quit [Quit: freedom]
jaller94[m] has left #ipfs ["User left"]
mowcat has quit [Remote host closed the connection]