Show Idle (>14 d.) Chans


← 2021-07-08 | 2021-07-11 →
whaack: billymg: Are you planning on making some link functionality? I.e. let's say i'm looking at a txn and there's an address, do you want the user to be able to click the address to query for the information for that address?
whaack: the cost of this is you need a separate "raw" command for users who want to use this service as an API
billymg: whaack: do you mean being able to click a txn hash or an address and then navigate to the page for that txt/addr?
whaack: billymg: yes
billymg: yeah, i was imagining it would have that ability
billymg: what do you mean about the extra cost for the "raw" version?
whaack: billymg: If you include the links and all that jazz that means that you are injecting html into the response of trbexplorer
whaack: i guess it's not really an extra cost, you can add a query param ?raw=true or something for users who want to see the data without html
billymg: whaack: i'm still not sure if i'm following 100% -- i see it as two components: 1) trbexplorer, your piece which does the heavy lifting and responds to requests for raw data, and 2) the bitdash www, which simply makes requests to trbexplorer, and then formats the raw data into html/css
billymg: so the trbexplorer component should only ever be serving raw
whaack: billymg: trbexplorer currently displays the information in a format that attempts to work nicely with awk
whaack: it's a human-readable but kinda-parseable tab format
billymg: ah, hrm
billymg: i mean that could also be read on the python web end
whaack: maybe for an mvp we can just display the current output
billymg: so you were thinking of users who want the tab-delimited format?
whaack: i have to add libraries and maybe even a decent number of LOC in order to have a json option
whaack: billymg: Yes, I decided to do that after speaking with jfw
billymg: ah, in that case i'd say skip the json altogether on your end
whaack: alright
billymg: the python www can eat the tab-delimited format
billymg: i only suggested json because i'm most familiar with that format
billymg: how does nesting work in that format?
whaack: tbh i'm also not familiar with the format and i may have included lots of stupid design decisions
billymg: e.g. how do you represent { 'txhash': 'c04b3c', 'outputs': ['output1', 'output2',...] } ?
whaack: billymg: afaik it requires understanding of the data you're working with
billymg: couldn't that be said of json too?
whaack: well json has a common structure to define a list and a dictionary
billymg: sure, but still need to know when to use which, when to nest vs. keep flat, which items nest under which other items, etc. et.c
whaack: hm i thought json specified nested lists/maps
billymg: json is just a key/value spec that supports strings, numbers, arrays, and dictionaries
billymg: can mix and match however you like, however nested and however flat
billymg: i think keys must be strings, but values can be any of the above (forgot booleans)
whaack: ok, well in any case, i don't know how tab and new line delimeted output specifies nested lists , my only experience working with this kind of structure has been parsing various command line outputs and for parsing inputs given for coding competitions (a la usaco.org etc.)
billymg: do you have an endpoint up now that someone could hit with curl and receive the tab-delimited response?
billymg: if so i could start playing around with that just to see what it involves on the www side
whaack: billymg: I don't have that up yet but I'll try to get it online maybe by end of tomorrow
whaack: let me try and put my bot in the channel ...then at the very least you can run a command and hit the paste.deedbot endpoint
billymg: ok, not a super rush or anything, i still probably have about another week's worth of work left for the crawler www before it can be left alone for a while
billymg: whaack: ah, yeah, that works too
← 2021-07-08 | 2021-07-11 →