Not compatible with Cygwin?

[dmos62]

Can't seem to get it running under Cygwin. I shall try and illustrate by copy pasting commands from mintty:

$ ls -l boot.sh
-rwxr-xr-x+ 1 Domas None 7012598 Dec  7 21:23 boot.sh

$ boot.sh
Error: Unable to access jarfile /home/Domas/bin/boot.sh

$ java -jar boot.sh -V
#https://github.com/boot-clj/boot
#Mon Dec 08 10:27:04 EET 2014
BOOT_CLOJURE_VERSION=1.6.0
BOOT_VERSION=2.0.0-pre28

Now here comes the crash (?). I will run the repl, but it will not respond to any input, I will ultimately Ctrl-c:

Domas@DomoHP-Win81 ~/bin
$ java -jar boot.sh repl
nREPL server listening: 0.0.0.0:49542
REPL-y 0.3.4, nREPL 0.2.4
Clojure 1.6.0
Java HotSpot(TM) 64-Bit Server VM 1.7.0_71-b14
        Exit: Control+D or (exit) or (quit)
    Commands: (user/help)
        Docs: (doc function-name-here)
              (find-doc "part-of-name-here")
Find by Name: (find-name "part-of-name-here")
      Source: (source function-name-here)
     Javadoc: (javadoc java-object-or-class-here)
    Examples from clojuredocs.org: [clojuredocs or cdoc]
              (user/clojuredocs name-here)
              (user/clojuredocs "ns-here" "name-here")
boot.user=> (+ 2 2)

$ (+ 2 2)
-bash: +: command not found

Some additional information below. Notably I am using a Windows Java installation, which has not caused problems running sbt or leiningen by the way.

$ java -version
java version "1.7.0_71"
Java(TM) SE Runtime Environment (build 1.7.0_71-b14)
Java HotSpot(TM) 64-Bit Server VM (build 24.71-b01, mixed mode)

$ javac -version
javac 1.7.0_71

$ which java
/cygdrive/c/Program Files/Java/jdk1.7.0_71/bin/java

$ which javac
/cygdrive/c/Program Files/Java/jdk1.7.0_71/bin/javac

$ echo $TERM
xterm-256color

[micha]

Did you try the boot.exe windows executable instead of the boot.sh unix one?

[dmos62]

Running boot.exe repl in the Windows Command Prompt works, though some characters are badly displayed. Running boot.exe repl in Cygwin gives same results as boot.sh repl. It prompts for input, but does not respond.

[micha]

@dmos62 What is your system's locale set to?

[dmos62]

Are these the variables you have in mind? As you can see only LANG is set.

$ echo $LC_ALL

$ echo $LC_TYPE

$ echo $LANG
C.UTF-8

I have tried boot.sh repl on a fresh Cygwin install and it did not respond as well.

[micha]

Have you had any luck debugging this issue?

[dmos62]

No, I have not. Something else I have noticed is that repl -s returns immediately after printing nREPL server listening: 0.0.0.0:51144 without actually leaving behind a running REPL server. Same happens whether running boot.sh or boot.exe.

[alandipert]

@dmos62 re: repl -s - the repl task doesn't block by default. To get the behavior I think you want you can compose with another builtin task, wait, with e.g. boot repl -s wait

[dmos62]

Just to clarify again, boot repl -s wait works fine, can connect to it and if running boot.exe from Windows cmd it evals input properly (except for certain characters being badly displayed); it's boot repl -c that does not receive any input from Mintty.

[schmir]

I can confirm the issue.

It looks like cygwin's mintty has many problems with windows console applications:

nodejs/node-v0.x-archive#6459
https://code.google.com/p/mintty/issues/detail?id=56
https://code.google.com/p/mintty/issues/detail?id=244

Judging from the last link, the repl client may try to check if System/console is nil and refuse to start.

[micha]

@dmos62 Can you use the leiningen repl client in Cygwin? Or does that result in the same problem?

[dmos62]

lein repl works fine; however, interestingly, lein trampoline repl has the same problem. More specifically lein trampoline repl stopped responding after the first run, lein clean does not help.

[schmir]

see technomancy/leiningen#850

[dmos62]

Opened an issue technomancy/leiningen#1852

On Mon, Mar 9, 2015 at 9:12 PM, Ralf Schmitt notifications@github.com
wrote:

see technomancy/leiningen#850
technomancy/leiningen#850

—
Reply to this email directly or view it on GitHub
#30 (comment).

[micha]

@dmos62 Can you try the stty and system properties workarounds as described in technomancy/leiningen#850 and see if that will fix boot.sh in Cygwin? Thanks!

[dmos62]

The workaround works:

~/bin % cat boot
#!/bin/bash
stty -icanon min 1 -echo
java \
      -Djline.terminal=jline.UnixTerminal \
      -jar boot.sh "$@"
stty icanon echo

First(?) session on Cygwin:

~/bin % boot repl
nREPL server listening: 0.0.0.0:63406
REPL-y 0.3.4, nREPL 0.2.4
Clojure 1.6.0
Java HotSpot(TM) 64-Bit Server VM 1.7.0_71-b14
        Exit: Control+D or (exit) or (quit)
    Commands: (user/help)
        Docs: (doc function-name-here)
              (find-doc "part-of-name-here")
Find by Name: (find-name "part-of-name-here")
      Source: (source function-name-here)
    Javadoc: (javadoc java-object-or-class-here)
    Examples from clojuredocs.org: [clojuredocs or cdoc]
              (user/clojuredocs name-here)
              (user/clojuredocs "ns-here" "name-here")
boot.user=> (+ 2 2)
4
boot.user=> "hello"
"hello"
boot.user=>

[micha]

Ok awesome, I'll work something out for this. Hopefully it won't need to be too ugly. I'll push to a branch so you can test it before it gets merged into master, okay?

[dmos62]

Sounds good.

On Thu, Mar 12, 2015 at 6:08 PM, Micha Niskin notifications@github.com
wrote:

Ok awesome, I'll work something out for this. Hopefully it won't need to
be too ugly. I'll push to a branch so you can test it before it gets
merged into master, okay?

—
Reply to this email directly or view it on GitHub
#30 (comment).

[micha]

@dmos62 Can you please verify that this does in fact work? Thanks :)

[dmos62]

@micha I verify :) though an interesting echo effect appeared, but that's another story.

% bin/boot.sh repl                                      
nREPL server started on port 59805 on host 127.0.0.1 - nrepl://127.0.0.1:59805
REPL-y 0.3.5, nREPL 0.2.7
Clojure 1.6.0
Java HotSpot(TM) 64-Bit Server VM 1.7.0_71-b14
[omited]
boot.user=> (+ 2 2)
(+ 2 2)
4
boot.user=> "hello"
"hello"
"hello"
boot.user=>

[micha]

That looks like the stty -echo wasn't working for some reason?

[dmos62]

Yes. The shell calls seem to be exactly the same like in the workaround script, yet when I wrap the latest boot in that script the echo disappears.