Esteban,

�� Could you explain line 8 ? why did you use a Base64MimeConverter?��

thanks.

line 8:

stream nextPutAll: 'ZnBase64Encoder '; nextPutAll: [ (Base64MimeConverter mimeEncode: ba readStream multiLine: false) ] bench printString; cr.

On Thu, Oct 29, 2015 at 3:43 PM, Esteban A. Maringolo <emaringolo@gmail.com> wrote:

2015-10-29 15:21 GMT-03:00 Peter Uhn��k <i.uhnak@gmail.com>:
>> >>> Option A) The printString of ByteArray, it is #[23 12 253 ...]
>>
>> >>> Option B) A Base64 encoded version of the ByteArray
>
>
> I like this one, because it is not Pharo-specific format. ByteArray is not
> pharo-specific, but the syntax is --- I can copy paste base64 from/to
> outside world.
> Plus base64 has better tool support in the wild.

I also prefer base64 even over native binary storage.

I ran a small comparison[1] of speed and size and these were my results.

ByteArray size: 19448860

Speed:
Base64MimeConverter: ''0.592 per second''
#hex: ''0.666 per second''
#asString: ''6.898 per second''
ZnBase64Encoder ''0.459 per second''

Size:
Base64MimeConverter: 26291980
#hex: 38897720
#asString: 19448860
ZnBase64Encoder 25931816

Esteban A. Maringolo

[1]: http://ws.stfx.eu/47PV2GUP9ITS

Bernardo E.C.

Sent from a cheap desktop computer in South America.