Fork me on GitHub
#clojure-uk
<
2018-03-12
>
thomas08:03:20

wow... people didn't clojure over the weekend......

otfrom08:03:43

@thomas I was on my belly underneath a yew bush attacking a bramble with a pair of secateurs

otfrom08:03:57

and seeing Black Panther (which is great, and better than bramble hacking)

maleghast09:03:16

'ello evrywun ๐Ÿ˜‰

maleghast09:03:24

Greetings from Londinium

maleghast09:03:24

@thomas - I was trying to not be ill, going to a school fundraiser (that was surprisingly fun) and then celebrating Mother's Day yesterday, before hopping on a plane... No time for Clojure until very late last night when I arrived at my digs and realised that I needed to have made more progress than I had... ๐Ÿ˜‰

maleghast09:03:44

I didn't bother with Slack - I assumed you'd all be sleeping at midnight / 1 o'clock

thomas09:03:44

and yes I should have said that people didn't bother with slack.... they might as well have done some Clojure

maleghast09:03:45

How was your weekend, @thomas?

thomas09:03:44

good, we went away for Saturday night and saw a play about Andy and Terry's treehouse, which was very funny

maleghast09:03:31

Cool ๐Ÿ™‚

yogidevbear10:03:16

Morning all ๐Ÿ™‚

yogidevbear10:03:45

@thomas, I did a little Clojure-ing over the weekend

guy10:03:11

@thomas i wasnโ€™t Clojure-ing but i did try out https://cryptozombies.io/en/course/ which was quite fun

guy10:03:38

Quite a funny intro into developing on Ethereum

guy10:03:42

Whats crazy about it @maleghast ๐Ÿ˜ฎ

yogidevbear10:03:35

Sounds like a fun idea imo

guy11:03:03

Yeah i found it quite fun haha

maleghast11:03:15

I'm sorry @guy, I am just experiencing extreme blockchain fatigue at the moment and the idea of building a game using BC tech of any stripe just makes me think "WHY? WHY FOR THE LOVE OF GOD?"

maleghast11:03:23

I am sure that it was fun ๐Ÿ™‚

guy11:03:05

Oh its just a tutorial thing @maleghast

guy11:03:10

so dont worry man ๐Ÿ™‚

guy11:03:31

I think itโ€™s a cute way to learn about a new technology with some hand holding

guy11:03:44

then if ur interested later, u can delve deeper

guy11:03:03

and yes blockchain fever is upon the tech world ๐Ÿ˜‚

guy11:03:23

I saw that a recruiter firm is making their own blockchain which is er interesting

maleghast11:03:04

There is so much pointless bullshit about blockchain - the simple truth is: 1. It's not all that new an idea 2. There are almost no use-cases where a central authority is not preferable to a distributed public ledger 3. The people who think it is going to "solve the unbanked problem" or "guarantee the provenance of food commodities" are absolutely delusional

guy11:03:58

๐Ÿ˜‚

guy11:03:11

and yet people are getting paid silly money to work on it

guy11:03:24

I think from a tech perspective its quite interesting

guy11:03:32

But i agree with your points tbh

guy11:03:54

Itโ€™s hard for me to think of an idea to use blockchain actually for

guy11:03:16

so from a pet project perspective itโ€™s hard to even get started

maleghast11:03:34

brb - need to reboot

guy11:03:43

๐Ÿ˜ฎ robot confirmed

guy11:03:55

๐Ÿ˜„ just made it sound like you were rebooting

guy11:03:05

Donโ€™t worry, was a silly joke

maleghast11:03:11

My trackpad was playing up - couldn't get it to snap out of a weird behaviour, so rebooted... Fixed now

maleghast11:03:21

Oh ok - sorry, not much sleep last night...

guy11:03:31

no worries man, Its monday!

guy11:03:38

The second hardest day of the week ๐Ÿ˜„

otfrom12:03:10

I'd be happy with a distributed ledger that didn't require so much electricity to operate, but then I'm thinking more about edge computing and distributed consensus with patchy networking atm (thus my interests in JSON CRDTs too)

maleghast13:03:30

Sounds interesting...

maleghast13:03:56

So, anyone got a good suggestion as to how to slow down a go-loop..? I have a go-loop that is consuming a channel of http API calls, and I am losing data from the API 'cos the calls are happening too close together - the API is throttled for no more than 5 requests / second

tmulvaney13:03:56

You could put a (<! (timeout some-amount-fo-time)) in the loop

tmulvaney13:03:26

where some-amount-of-time is a number of milliseconds

maleghast13:03:58

Yeah, actually I was wrong above ^^ the multiple http requests are getting stacked up inside a doseq block...

maleghast13:03:41

So all the consumer stuff that uses channels can go as fast as it likes (so long as the DB and ES cluster can keep up), it's just the first part that needs to slow down.

maleghast13:03:22

The API I am hitting is limited to 5 requests / second, so I want to just force the doseq loop to do one iteration per second.

maleghast13:03:01

Thanks @mccraigmccraig - I will take a look

maleghast15:03:07

Looks exactly what I was looking for, thanks @mccraigmccraig ๐Ÿ™‚

mccraigmccraig16:03:26

it's worth having a browse through the code @maleghast - it's pretty simple, and a great example of the advantages of making processing pipelines explicit (with buffers, backpressure etc)

maleghast16:03:44

I will do that later on, thanks ๐Ÿ™‚