mafintosh: mostly we talked about having multilpe servers that also publish to each other
mafintosh
clkao: ah yea i remember that. i did some research though and i dont think we need that
clkao: because bt trackers don't need to that to support bt discovery and we'll probably have less peers than bt
ogd: sry was offline. reading your comments
ogd: ah .key there is a typo. it should just be .content
ogd: .content is null if the archive only has empty files (edge case)
ogd: re, have messages. the indexes are block numbers. end defaults to start + 1 or bitfield.length if that is present.
ogd: yep you can supply a higher start offset. its hardcoded to 0 for now, just so i could ship it. you'd use an offset if your bitfield was super big (tb's of data)
iml_ has quit
juul1 is now known as juul
toddself_ joined the channel
toddself_ has quit
toddself_ joined the channel
toddself_ has quit
ungoldman has quit
toddself_ joined the channel
toddself_ has quit
null_radix has quit
null_radix joined the channel
ddem-bot has quit
clkao
mafintosh: for bt trackers the client connects to multiple of them right? we'll want to do the same for multiple signalhub instances as well then?
ddem-bot joined the channel
m-i_ joined the channel
m-i has quit
mafintosh
clkao: yea but they dont share any state. we should do the same using signalhubs
also i am reviewing the issues we closed: http://hackfoldr.org/datteam/ and will be archiving the done items in ~8hrs
toddself_ joined the channel
toddself_ has quit
null_radix joined the channel
m-i_ has quit
m-i joined the channel
m-i_ joined the channel
toddself_ joined the channel
toddself_ has quit
m-i_ has quit
m-i joined the channel
m-i has quit
m-i joined the channel
m-i has quit
m-i joined the channel
m-i has quit
toddself_ joined the channel
toddself_ has quit
m-i joined the channel
m-i_ joined the channel
m-i has quit
m-i_ has quit
m-i joined the channel
m-i has quit
iml_ joined the channel
toddself_ joined the channel
m-i joined the channel
m-i has quit
toddself_ has quit
m-i joined the channel
m-i has quit
m-i joined the channel
m-i joined the channel
dat-gitter-bot
(Compy) Good morning all
G-Ray has quit
mafintosh
morning
dat-gitter-bot
(Compy) Hey mafintosh! So I'm looking at all the options for my peer assisted delivery JS library, and I'm a bit torn as to whether or not dat or webtorrent would be ideal for this particular use case (distributed CDN).
mafintosh
Compy: i'm taking off in 10. both are good options. depends on the trade-offs / use cases you wanna support
dat-gitter-bot
(Compy) Yeah, no worries mafintosh! I appreciate the response. I like the versioning capabilities that dat offers.
mafintosh
Compy: the best thing you can do imo is build a decoupled module that allows you to build both versions trivially
Compy: then you can pick the one you need based on use case :)
jhand: we tried to do more elaborate JS APIs before but it doesn't really make sense and is too hard to maintain. So that's my only qualm is to keep it super clean and minimal like the cli
Jhand the cli could call the dat API and pass the logger as an option which is by default turned off, perhaps
jhand
karissa: ya that seems like a good middle path. And it is mostly done now =).