Fork me on GitHub
#clojure-uk
<
2016-10-19
>
agile_geek06:10:25

@otfrom how an earth do you write Java in Emacs? I've used Emacs for Scala (ensime) but Java iDEs generate so much boilerplate and do some much refactoring that I can't do as fast in Emacs?

seancorfield06:10:24

^ that is the most searing indictment of Java for me... that it requires so much boilerplate that you need an IDE.

benedek06:10:24

recently jdee came to life again (ide like plugin for emacs for java) are you using that guys?

benedek06:10:46

I used it back in my java days and before java5 but old jdee never got the syntax update. but now it is actively developed/maintained again

korny06:10:03

I've tried jdee a few times, and never got it to work reliably. It might be OK if you built a java project from scratch in it? Or maybe something about how the projects I used it on were structured? Or maybe I just gave up too quickly.

markwoodhall06:10:37

@agile_geek: it's not Java but I'm spending most of the day writing C# in vim. It's not as bad as it sounds. Honest.

korny06:10:01

The main things I miss actually are navigation and really simple warnings. I want to be able to say “where is this method called?” and “where is this interface implemented?” and basic code navigation. And if I miss-type a method or class name, or get the type signature wrong, I want that problem syntax highlighted immediately. I don’t know how JDEE works these days, but last time I tried it it was not reliable at such things

benedek07:10:24

tbh I've not used the new JDEE only the old one which had these things

agile_geek07:10:16

@seancorfield I agree. However, I lean on an IDE more when the code base is a big ball of mud regardless of language as the refactoring support is generally better.

agile_geek07:10:12

@markwoodhall you may be the first C# dev I've ever met who doesn't use VisualStudio! I personally think VS is one of the best IDEs ever written and as it's almost default in the .NET world everyone's on a level playing field and pairing is easy.

glenjamin07:10:36

IDEs are sometimes a crutch which makes it easier to write messy stuff

glenjamin08:10:05

auto-import in Java is a good example of that IME - it’s so easy to use another class without thinking about coupling

agile_geek08:10:43

@glenjamin tbh most of the problems with Java code I deal with in the big corporates lies with the dev's not even knowing what OO is and writing very poorly structured procedural code.

agile_geek08:10:38

IDE's don't help or hinder when your dev team knows basic syntax but nothing else.

markwoodhall08:10:10

@agile_geek: there are quite a lot of C# devs trying to move to more lightweight editors, like VSCode. I wanted to move to something I could invest in and make use of the knowledge gained when switching platforms/languages.

thomas08:10:13

I’ve done all my Java in Eclipse. Mainly because my employer made Eclipse in the first place and all the “other” tooling here is Eclipse based.

thomas08:10:19

things like RTC etc.

thomas08:10:27

so kinda stuck in Eclipse.

otfrom12:10:53

anyone tried meghanada-mode for emacs Java editing yet? https://github.com/mopemope/meghanada-emacs

otfrom12:10:06

I used to use JDEE agile_geek I really liked beanshell before I had a good clojure repl

mccraigmccraig12:10:27

i used jdee for a while back in the day. i was eventually seduced away by intellij's refactorings though

agile_geek12:10:04

Looking at meghanada it's not got any refactorings and it's not clear to me if you can import other source code formating rules both of which I end up using heavily in most Java codebases.

agile_geek12:10:14

Having said that I may give it a punt

agile_geek12:10:14

Although as I'm currently writing Clojure/Clojurescript for a living I have no urgent need. Yay! 👏

jonpither12:10:20

Booking a bar for Monday night in Bratislava - first couple of rounds on JUXT

thomas12:10:10

dang….. going to miss EuroClojure this year 😞 going to sort out my new job though.

jonpither12:10:29

who is going in here?

agile_geek13:10:04

As the twisted bit of scraggy rope said...'fraid not

glenjamin15:10:48

hah, just had flashbacks to java autoformatting

glenjamin15:10:36

the IntelliJ and Eclipse users at a previous (recent) job couldn’t get the formatters to line up - so after 3 or 4 weeks they almost decided to force everyone onto the same editor

glenjamin15:10:47

but then decided instead to disable autoformatting

otfrom16:10:33

(or maybe it does, but not any better than iterate and test last)

glenjamin16:10:44

(in that context) 😛

glenjamin16:10:52

certainly still remains unproven

glenjamin16:10:19

>We verify the baseline study results, yet our results raises concerns regarding the selection of experimental objects, particularly with respect to their interaction with the order in which of treatments are applied. Can anyone parse that?

glenjamin16:10:00

oh, this is interesting

glenjamin16:10:26

> The task was divided into 13 user stories of incremental difficulty, each building on the results of the previous one. An example, in terms of input and expected output, accompanied the description of each user story

spike16:10:09

:thinking_face: the grammar on that is really not clear

spike16:10:14

(the first statement)

spike16:10:30

experimental objects = mars rover or bowling scoring kata

spike16:10:33

treaments = TLD or TDD

spike16:10:55

so it either means the way that each methodology interacts with the task order, or...

glenjamin16:10:05

RE the second quote - in my experience TDD helps me convert a spec from humans into a series of incremental tasks with example inputs an outputs

spike16:10:14

hmm nope they didn’t switch participants between groups so not that

glenjamin16:10:34

so if the spec is given as a series of incremental tasks with input and output, I would be surprised to see TDD make any difference

otfrom16:10:58

(btw, pretty much all "empirical" studies on this have huuuuuuge methodological problems)

spike16:10:19

oh here we go

spike16:10:21

>>> there is the possibility that the (fixed) sequence of the experimental tasks interacts with one (or both) treatment sequence—for example, the MRA task inherently suits TLD, while BSK tends to be TDD’s sweet-spot, or vice versa.

otfrom16:10:34

(and there are awkward anthropology loving people like me who think you'll never really be able to measure complex human activities like this in a repeatable way if they are actually interesting activities)

glenjamin16:10:44

it does make me wonder how accurate all the various business efficiency research studies are

otfrom16:10:02

but it is good to be able to argue methodology

glenjamin16:10:05

since they have to suffer from the same experimental difficulties

otfrom16:10:19

http://neverworkintheory.org is great for things like this