?
line modes
+monospace *bold _italic )right justify =centered text >quote ~sampel-palnet !https://..image.jpg <http://link.com
the first character of a line sets its mode.
a mode applies to the whole line.

[view hawk code]


~migrev^dolseg
welcome to hawk
a programmable interface
the manx machine
[hawk] consists of that
[userspace] software which admits of
[ubiquitous html5 clients].
- urbit precepts on “real software”
this forum is made with hawk
is this a chat or a forum?
current line count: 235
don't you get it? hoon won forever.
features:
- line-based formatting
- persistent reading position
- eauth identities
bookmark it, i dare ya

~migrev-dolseg
i'm considering pausing 499 work
for a week or so
and just implementing
a simple build cache in 500.
because it turns out these little
manx apps are pretty ergonomic to
write. the only issues are:
- the speed
- the unimplemented features (like fine-grained access control)
both are solvable without 499
---
i bet in under ~450 lines,
this can become a useful forum:
defined as a single file.
---
have been meaning to write up
a post fully explaining my
thoughts on software distribution.
the tldr is:
most apps fit in a single file
if your programming environment
is doing what it should.
once this is true, the optimal
installation experience is to
simply paste the code into the
place where you want to install
the apps.
this gives you a relationship with
the code. if you ever want to change
it, you know where it is.
this is the thing that
"suckless" software
wanted to be
quality software with a focus
on simplicity, clarity,
and frugality.
- the suckless.org philosophy page
still working on the frugality...
there's a need for speed.

~nordus-mocwyl
what's this than?
'ello
'ello

~sarlev
I still think that there needs to be an idiot-proof (idiots can't copy paste) version of software distribution. (Read: it needs a button)

~migrev^dolseg
the hawk ui is not simple enough
until even the idiots can copy/paste
into it.
hawk is just a worse version of
traditional software until the
user understands that the code in
the code tab controls the display
of the page.
should have just done markdown
but nooooooooo i Had To Invent

~mismet_dolseg
oh hey anon

~namtun_dolseg
yo yo yo. let us see what this mention feature does:
~malwer-fotbel

~malwer-fotbel
so this code:
[%'~' rest=*]
::
=/ mention (slaw %p line)
?~ mention ;div:(-p)
;div.mono.o8.bold
=data-mention +:(scow %p u.mention)
;- (scow %p u.mention)
==
...just turns it bold currently? but also labels it as a "mention" in such a way that may be used for more (such as link to their hawk page or send them a notification) at some later time?
whoops, here
... but will, you also have a small whoops in your instructions, which led to my whoops.
bold is done with *bold
monospace is done with +monospace

~migrev^dolseg
indeed, thank you
~malwer-fotbel

~ripwyd_dolseg
Very cool!

~malwer-fotbel
no thank you for making this!
persistent reading position is interesting, as re-loading causes it to gradually scroll down.
what was at the bottom becomes what's at the top (roughly).
took me a few reloads to figure out what it was doing.
actually, on further testing, it goes down, down, down until ~sarlev's "I still think that there needs to be an idiot-proof" post.
then it stays there, stationary, on all subsequent reloads.
presumably that is related to him always being right.
perhaps the software is interpreting that as "he should get the last word".
that would make sense.
~ripwyd_dolseg, eh?
a long-lost cousin?
it is fascinating that each message gets it's own @p!
the above e.g. is #ranmeg-binlug
those can't be actual comet names, right? since they all end in dolseg?
what's going on there?
ok, further troubleshooting (can't help myself?):
I zoomed in the browser magnification to 150% and determined it does not, in fact, respect
~sarlev
properly. at 150% it stops moving after seven reloads.
other magnifications are different.
actually at 100% it seems to continue all the way to the end. so i must have had it zoomed in previously.

~migrev-dolseg
the persistent reading position just means:
when you reload,
(or come back to the page after having closed it)
you should end up near the thing you were previously
looking at.
this means if you scroll to the bottom each time
you read the forum, when you get to the bottom,
you know you've read everything that's new.
i think comet ids with eauth are based on the host ship somehow.

~migrev^dolseg
i tweaked the positioning logic
to avoid the
"slowly scrolls down
page on successive reloads"
issue

~malwer-fotbel
can confirm it is repaired.
amazing turn-around time!
this makes reporting bugs a joy. you report them, boom they're gone. yes!

~migrev^dolseg
test again
after creating a new post,
you should now always scroll to the bottom.
fixing bugs is 10x faster when
you have no "deploy" step.
i just come to this page
and start changing a few lines
often even from my phone
we're merging with the machine
---
the obvious 2 next "line modes":
clickable links
image urls
all images are grayscale-d
cause we're just cool and aloof like that
i'll leave clickable links
for another day

~rinlur_dolseg
Cool. Now I can chat in %hawk.
One more step towards %hawk dominance.
This feels like a guestbook and I like it.

~pasnec^salmyr
Test. Is this thing on?

~mastyr-bottec
Hypermedia event stream reactive webapp no spa maximalism?

~migrev^dolseg
"thin client" maximalism
"hypermedia" maximalism
"manx" maximalism
welcome to the manx machine
they don't tell you this
in the 6 week react bootcamp
but you can just create
new mediums and new shapes

~malwer-fotbel
for another day
beat you to it!
<https://malwer-fotbel.tlon.network/~~/talk

~migrev^dolseg
< is a good choice

~rinlur_dolseg
This is crazy. Wth?

~milset
OK I'm in

~filbyl_dolseg
does this work?

~waldys_dolseg
you should call this wall as an homage to you know who. ;)

~malwer-fotbel
i actually dont know who. zuckerberg?