I love simple ideas :)

This is a thing of beauty. I've grown as a programmer seeing this brilliant algorithm.

The 'fromAscii85' tests are pretty broken in that they don't check size. So if you return a zero-byte result, it will 'pass.'

very nice solution. absolute beautiful.

Amazing!Collected it.

Language?

One of the test on test_encoding_with_multiple_zero_bytes_inside_a_binary

[61, -8, 22, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, -98, -109, 119, -85, -92, 23, -111, -112, 31, -108, -125, -55, -44, -73, 95, 79, -120, 76, -53, 39, 12, -16, 102, -61, 120, -110, 2, -59, 55, -74, -120]

The seconds 4 bytes blocks are all Zeros(null), whyit expects this expected:<<~4nocm[zzz]!!!"jPBX('(Qje$P_7*-...>, only 3 zzz instead of 5 zzzzz for that block?

apparently fixed

fixed

• added a call to srand() to the tests suite initialization
• fixed the example tests
• moved the reference solution out of the preloaded code section
• renamed some stuff in the initial code so it would make more sense (it was very awkward to call the binary strings "blocks" since "block" has a specific meaning in that kata)
• i wont dive more into the code because it's really hard to read

old issue with no details, closing

I don't understand :
all the test are good 100% and a "STDERR" append ??