Fork me on GitHub
#clojure-uk
<
2017-07-19
>
chrjs06:07:10

EuroClojure day! :airplane_departure:

seancorfield07:07:12

Are we going to get live Slack-ing of the whole event?

chrjs07:07:10

Would you like it?

dominicm07:07:55

@seancorfield how do you turn a namespace into a dependency?

agile_geek10:07:28

Have fun everyone whose at EuroClojure.

agile_geek10:07:14

Question of the day: Given that our industry made it possible to collaborate remotely, why do so few companies allow developer/architects/designers/testers etc. to do so? (Don't give me the "co-location is best" argument as it's been demonstrated as false if team is all remote. Personally and cynically I think it's a 'trust' issue)

Rachel Westmacott10:07:33

I wouldn’t be surprised if VR is enough to cross that boundary for most people

minimal10:07:56

I think it’s harder to move an existing company in that direction. If you start a company with remote as a first class thing then it is easier to make it work and trust etc

Rachel Westmacott10:07:04

although if VR conferencing ends up like video conferencing now, then I take that back

Rachel Westmacott10:07:09

also, most people are idiots (myself included) so they don’t always see that you need to trust your employees anyway

glenjamin10:07:34

because people believe that co-location is best

glenjamin10:07:40

and are rarely challenged on it

minimal10:07:51

yep an ex-ceo of mine said something like “as marissa mayer has said, remote doesn’t work”

glenjamin10:07:52

co-located communication is very high bandwidth, but getting everyone into a room is high latency. People who get to decide these things tend to focus on the former and not the latter in my experience

glenjamin10:07:18

all else being equal, co-located is better. all else is never equal

agile_geek10:07:30

@minimal I think you're right. I think remote works best with an all remote team as all conversations go on on-line. I've seen really well disciplined teams with some in office and some remote work if there are enough remote (Mastodon C is a great example, kudos to @otfrom and Fran). The worst examples are the typical large corporates with split teams across massive time zone differences.

minimal10:07:13

@agile_geek yeah the more people working remote at one time, the less the local team can dominate the conversation and have local (secret) chats

minimal10:07:02

I know companies that have dedicated wfh days like tuesday and thursday where everyone is remote so there is no friction

agile_geek10:07:40

@glenjamin I take your point but my experience is co-located communication is often unfocused with a high noise to signal ratio.

geek-draven10:07:28

I'm currently working with a dev in the Philippines on a few things - he works UK hours and all the communication is via email and slack, and it works fine. We've even done some peer programming - something I never did with the dev I sat next to for 2 years in my old job 🙂

agile_geek10:07:58

@geek-draven Working UK hours from Philippines is hard on individual tho! Used to have to work with dev team in China and PO in New York...7am meetings with team in China and 7pm meetings with NY..and I had to commute to my office everyday to do so despite everyone else in office not being in my team!

geek-draven11:07:21

@agile_geek I know, I have to keep telling him to go to bed when we're still working on something 7pm UK time 😄 It usually means I wake up to a screen of Slack messages at 6am though

jasonbell11:07:14

I get more done remote than I ever do in an office. Plus I now have added complications of mobility so I prefer not to commute too much. Also my commute to the office is rather expensive 🙂

agile_geek11:07:25

The team in China used to finish 6pm for them (either 10 or 11am UK) except team lead who used to be still around at 4pm UK (11pm or 12am for him) unless I sent him home!

agile_geek11:07:02

It seems ridiculous to me that London based companies will pay me an extra £2000+ a month just to commute to London....but they do!

agile_geek11:07:35

@jasonbell that jokes got a bit Jaded!

jasonbell11:07:40

On the minus side I probably miss the conference side of things and the office banter, though I do try to create me own in Slack….. walking stick air guitar during keynotes must be made a thing,

geek-draven11:07:24

I used to have to commute into London to work for a team that was in a different timezone and location to me - some companies are weird

yogidevbear12:07:43

Jade! How are you doing?

yogidevbear12:07:54

@agile_geek I'll keep championing for remote working. Don't know if it ever have an impact that will help you out, but I'll do my best 😉

maleghast13:07:14

Hello Everyone! Hope you are having fun at EuroClojure if you are there. I am not *jealous*, no not at all…

maleghast13:07:41

Side note, don’t Clojure apps, particularly ones that use Datomic in memory need quite a lot of RAM… 😞

agile_geek13:07:14

👍 👍 🆒

agile_geek13:07:22

Datomic peers use a lot of ram as dB copied/cached locally... Datomic clients not so much

jasonbell13:07:56

@yogidevbear Jade’s good, Jase is even better 🙂

maleghast13:07:14

This made me sad…

maleghast13:07:27

I am going to have to kill the instance that I am running and re-provision the instance with MOAR memory. This is also going to mean that I need to figure out a way to automate wake-up and go-to-bed as the instance is going to cost too much to just leave it running… 😞

maleghast13:07:58

Of course AWS does make it really easy to do these things in truth… 🙂

agile_geek13:07:41

That looks like your query is constructing a huge results vector or your queries are hanging on to data for a while as by default Datomic caches on LRU and should release memory. Having said that at one point at previous client we were using 20Gb! Bit of refactoring got it down to about 6Gb but still huge

mccraigmccraig17:07:39

@maleghast clojure apps do not necessarily use a lot of ram... all my clj processes run happily in 512M - probably less, but i haven't tried to trim it

bronsa17:07:26

@maleghast are you sure you're not leaking memory/causing head retention?

mccraigmccraig17:07:45

ah, remembering now... i had to bump my api process memory allocation, solely because of some 2D graphics ops on the large images our users were unkind enough to submit, but that seems fair enough... otherwise general api and db ops use very little

seancorfield17:07:10

Re: remote work -- I've been fortunate enough to work almost entirely remotely for a decade now. That started with architecture-level consulting and my clients were all over the world and no one wanted to pay to have me travel to them 🙂 then a local client hired me but allowed one or days remote working (then they imploded -- as Bay Area startups tend to do!) and I was a remote consultant again, until another client hired me and I just stayed remote.

seancorfield17:07:39

It amazes me how few companies allow remote work tho'...

otfrom21:07:01

it was @bru and @elise_huard who made us good at being remote (w/people in a head office)

otfrom21:07:18

& I can't pair except over a screen share now. I get a sore neck otherwise