IRC channel logs

2017-11-14.log

back to list of logs

***hydraz is now known as sigint
***sigint is now known as hydraz
<happy_gnu[m]>kristofer: thanks!!
<happy_gnu[m]>amz3`: thanks too!!
<daviid>anyone good at math can help me with this small coordinate translation problem: given two bounding boxes, one for an object in IMAGE and the other for another object in SEEDS, for example '(78 0 200 47) in IMAGE and '(62 29 79 35) in SEEDS, I know they share an area calculated as '(78 29 79 35)
<daviid>now, for some reason can not explain (to long), I actually need to convert these in terms of relative to the object of IMAGE, so given by '(78 0 200 47)
<daviid>these coordinate are (left top right bottom) in a coordinate system where the Y axis points down
<daviid>in this particular example, I can calculate that the coordinates of the 'shared' object in the 'sub image' object are 0 29 121 12
<daviid>nope, wrong already!
<daviid>these would be 0 29 1 35
<daviid>fine, but I failed to come up with a formula which work in all cases, a bit tired, decided to see if anyoe is more awake then myself :):)
<ArneBab>sneek: later tell daviid: do you mean you need to transform all three areas (IMAGE, SEEDS, shared) into coordinates relative to IMAGE?
<sneek>Will do.
<ArneBab>sneek: later tell daviid: or rather: each of the edges relative to the corresponding edge in IMAGE?
<sneek>Got it.
<ArneBab>sneek: botsnack
<sneek>:)
<amz3>hélll all!
<amz3>héllo all :)
<ArneBab>hi ☺
<rekado_>like amz3 reported a few days ago, I cannot seem to get (web client) to work with gnutls.
<rekado_>ERROR: In procedure get-bytevector-n:
<rekado_>ERROR: Throw to key `gnutls-error' with args `(#<gnutls-error-enum The TLS connection was non-properly terminated.> read_from_session_record_port)'.
<rekado_>Do you know how I can debug this?
<amz3>I use the following as a workaround: https://github.com/a-guile-mind/culturia.one/blob/master/src/web/helpers.scm#L68
<amz3>I don't know how to debug
<rekado_>Oh, wrapping curl is nasty :)
<rekado_>I think I’ll just snarf the working code from Guix’s (guix build download)
<civodul>rekado_: would be good to see why (web client) fails
<civodul>it is/used to be the same code as in (guix build download)
<rekado_>it looks rather simpler than what is in (guix build download)
<civodul>maybe it lacks the server authentication code, no?
<rekado_>(guix build download) catches 'gnutls-error whereas (web client) does not
<rekado_>but that’s probably not important here
<civodul>right, it's even a good thing that (web client) doesn't catch it
<rekado_>tls-wrap in (web client) uses a custom binary port
<rekado_>“make-custom-binary-input/output-port”
<lloda>looking for an Emacs spell to indent define-values properly and not as if it was a function definition
<amz3`>héllo #guile
<nilg>I'm pretty sure there is a function to a load a file without compiling, but I can't find (duckduckgo didn't find it)
<bavier>nilg: "load"?
<nilg>Apparently not, according to https://www.gnu.org/software/guile/manual/html_node/Loading.html#Loading this is equivalent to compiling, then loading
<nilg>I know one can call guile with option `--no-auto-compile` but that is not what I'm looking for, rather I'm looking for a way to disable compilation on some files
<rekado>that error in (web client) is the only thing that keeps me from making a proper release of guile-debbugs.
<amz3`>rekado: you made a http client library for debbugs?
<OrangeShark>hello everyone
<amz3`>héllo OrangeShark
<OrangeShark>how are you amz3`?
<OrangeShark>What are you working on now?
<rekado_>amz3`: it’s a SOAP client library, yes.
<OrangeShark>rekado_: debbugs uses SOAP?
<rekado_>OrangeShark: it provides a SOAP service, yes. The Emacs interface connects to the SOAP service as well.
<amz3`>OrangeShark: nothing, I am looking for a new job..