Fork me on GitHub
#hoplon
<
2015-11-15
>
leontalbot01:11:13

Hi guys. It has been two years since I have touch Hoplon for the last time. Seems like Hoplon is back in my radar. Has the general performance improved?

micha01:11:08

i think so, what was the trouble you were experiencing?

leontalbot01:11:07

Hi @micha ! Some lag in minimal todo app example.

micha01:11:27

oh, this was in the very early days, yes?

leontalbot01:11:42

Yes I think so

micha01:11:56

we quickly got all that stuff squared away

leontalbot01:11:58

So it is better? :-)

alandipert02:11:13

also, computers are faster simple_smile

leontalbot02:11:44

Ok thanks @micha and @alandipert . Might give it a new spin :-)

leontalbot02:11:13

The simplicity of Hoplon for Web development is really appealing :-)

onetom06:11:14

@micha: i was also thinking about getting that curved monitor. i just saw it 3 weeks ago live and it was rather impressive. i found that problems with taller displays is that they r putting a strain on the neck when u r looking up and down, but looking left and right is less of an pain. for example a 27" 2k display vertically - while super convenient because u hardly ever have to scroll your source files vertically - is tiring to scan up n down.

onetom07:11:28

@micha: re: flux-challange and cancelling (i assume xhr) requests. we just had to do that again and we found that the xhr.abort method "gets lost" along the way, because the .fail and .done calls on the original xhr promise will return new promises, but just plain vanilla jq.Deferred ones, which dont have .abort method.

onetom07:11:34

the need for this arose from the fact that currently our backend have a huge variation in response time and sometimes earlier requests overwrite the response of a newer request once the earlier request finally returns.

onetom07:11:56

think about search-as-u-type functionality for example

thedavidmeister08:11:50

@onetom: +1 on tackling that

onetom08:11:37

@thedavidmeister: why, did u have a similar need?

thedavidmeister08:11:59

i’ve had that problem a bunch of times

thedavidmeister08:11:04

i’m sure i’ll run into it again simple_smile

thedavidmeister08:11:27

autocompletes, checkboxes for search filters, etc...

onetom08:11:30

well, we shall bake it into castra eventually

onetom08:11:52

the function created by mkremote is already returning a promise, which indeed do have an xhr attribute, it's just "not the right one"...

thedavidmeister08:11:01

pretty much anything you’ve got sending more than one ajax request should only allow the most recent one sent to update its state, or be prepared to show a log of everything that’s coming back and be careful about the order of what was sent vs. returned

thedavidmeister08:11:45

off the top of my head, i can’t really imagine anything where you’d want older ajax to override newer ones

onetom08:11:44

what im also missing is the debuggability. the loading cell containing the vector of requests in flight is a great idea, but currently i can't inspect it to see what's in flight

onetom08:11:26

because currently it has these deferred objects or what

thedavidmeister08:11:59

hmm, other than debugging, is there a benefit to keeping track of canceled ajax?

onetom08:11:19

no, i didn't mean to track cancelled ajax calls

onetom08:11:44

i was just saying it's a nice concept to have a list of in-flight ones

onetom08:11:16

because then u can see which ones are are stuck and in which order do they return

thedavidmeister08:11:25

you know, i’m going to just assume that whatever i did in the past isn’t directly applicable here 😛

onetom08:11:25

in our case for example i would expect to see something like:

[get-config get-current-user]
[get-current-user]
[get-apps]
[]

onetom08:11:14

though i would probably use a different loading cell for stuff which is only relevant for the logged-in states

thedavidmeister08:11:16

aaaah, so a list of “unique” ajaxes

thedavidmeister08:11:28

not a list for each of the ajaxes

onetom08:11:59

no, i didn't think about uniq now. why do u think thats useful?

thedavidmeister08:11:29

nope, i just thought you were doing something else

onetom08:11:41

ok, see u!

micha21:11:56

haha what's the bubbles demo?

onetom23:11:51

u need rebol2/view to run it (which needs 32bit compatibility libs under linux unfortunately)

alandipert23:11:34

this Pair! datatype

alandipert23:11:07

were you a rebol user before red?

alandipert23:11:24

ah, it would appear so

onetom23:11:49

since 2001 or so

onetom23:11:27

just noticed the audio track on the video... sorry for that. i was sitting on the 7am lamma ferry simple_smile the kid next to me was sucking his nose...

alandipert23:11:03

haha no problem

alandipert23:11:08

the rebol "ide" looks interesting

onetom23:11:09

Its a super simple editor BUT its part of the 1MB rebol/view exe so it's cross platform and u can count on it... U can also inspect and modify its source...

onetom23:11:11

The thing is w rebol that u have to type a lot less so u dont need a super-duper editor for it... Which is not the case w lisps, though w lisp u dont need a generic editor. But in a lisp project u might still want to edit regular txt files...

onetom23:11:26

Nenad's vision is to have a full collaborative ide with chat linked to the author of every lines of code within a 1MB single file exe

alandipert23:11:54

that's a pretty great vision