What's stopping Nim from going mainstream? (And how to fix it?)

[giaco]

added nimcr to the party and I can confirm that nimcr + tcc takes the lead. I 
now see your point, thanks. Didn't know about nimcr before.

I'd like to know why `nim r` is so much slower than `nimcr`, considering the 
fact that it skips compilation if source is not changed, too.
    
    
    #!/usr/bin/env nimcr
    
    var c = 0
    for i in 0 ..< 1_000_000:
      c += 1
    echo c
    
    
    Run
    
    
    rm -rf ~/.cache/nim/count_* && rm .count_nimcr && hyperfine --runs 10 'nim 
r -d:debug count.nim' 'nim r -d:release count.nim' 'nim r -f -d:debug 
count.nim' 'nim r -f -d:release count.nim' './count_nimcr.nim' 'nim e 
count.nim' 'python count.py' 'bash count.bash' 'ruby count.rb'
    
    
    Run
    
    
    --threads:off
    --cc:gcc
    
    
    Summary
      'ruby count.rb' ran
        1.30 ± 4.03 times faster than './count_nimcr.nim'
        1.81 ± 0.14 times faster than 'python count.py'
        2.56 ± 1.50 times faster than 'nim r -d:debug count.nim'
        3.23 ± 3.43 times faster than 'nim r -d:release count.nim'
        4.17 ± 0.25 times faster than 'nim e count.nim'
        6.49 ± 0.33 times faster than 'nim r -f -d:debug count.nim'
       13.78 ± 1.43 times faster than 'nim r -f -d:release count.nim'
       21.52 ± 1.74 times faster than 'bash count.bash'
    
    
    Run
    
    
    --threads:off
    --cc:tcc
    --passl:"-ldl -lm"
    
    
    Summary
      './count_nimcr.nim' ran
        1.95 ± 4.73 times faster than 'ruby count.rb'
        3.37 ± 8.18 times faster than 'python count.py'
        4.78 ± 11.67 times faster than 'nim r -d:debug count.nim'
        5.08 ± 12.38 times faster than 'nim r -d:release count.nim'
        8.08 ± 19.60 times faster than 'nim e count.nim'
        8.50 ± 20.61 times faster than 'nim r -f -d:debug count.nim'
        8.80 ± 21.34 times faster than 'nim r -f -d:release count.nim'
       45.27 ± 109.84 times faster than 'bash count.bash'
    
    
    
    Run