On Fri, Jul 18, 2014 at 2:45 PM, [email protected]
<[email protected]> wrote:
> Laruence:
>
> Thank you for your reply and suggestion.
>
> You are right. In fact, most of the self-description serialize protocols are
> similar.
>
> The very reason we made BINPACK is that we want a simple enough and fast
> enough serialize protocol.
>
> Where I work, we are using a RPC protocol which is based on BINPACK to swap
> data between languages. Because of the huge traffic of requests, we want to
> make the serialize protocol as fast as possible.
>
> In PHP, BINPACK is much faster than msgpack:
> https://github.com/binpack/binpack-php/wiki/Performance  . Using BINPACK,
> every RPC request will take less request time and use less CPU time of the
> Server.
>
> Msgpack does more things both in JAVA and PHP. In PHP, it can also handle
> something else, the session, for example.
>
> BINPACK is totally like JOSN, which only supports basic data types. All the
> implementation of BINPACK in different languages only simply do one job,
> encode / decode.
>
> So let BINPACK do the simple work and it will do better.
Hey:
    maybe we could be compatible in the serialized bin data? that will
be great...

thanks
>
> ________________________________
> Huqiu Liao
>
> http://liaohuqiu.net
>
>
> From: Laruence
> Date: 2014-07-18 12:22
> To: 廖祜秋
> CC: PECL Announce; Ferenc Kovacs
> Subject: Re: [PECL-DEV] Request to Publish in PECL
> Hey:
>
>
> On Fri, Jul 18, 2014 at 11:11 AM, Laruence <[email protected]> wrote:
>> Hey:
>>
>>
>> On Mon, Jul 7, 2014 at 11:27 PM,  <[email protected]> wrote:
>>> Hi all,
>>>
>>>     I am a developer currently with Alibaba inc.  I want to host my
>>> extension in php.net. A few weeks ago,  I emailed this mail list and was
>>> told that my codes needed some improvements because of my non-C89 code style
>>> and lacking test case.
>>>
>>>     Since then I have made improvements accordingly and am requesting to
>>> publish the extension again. This extension is called BIN PACK. Here is the
>>> source code: https://github.com/binpack/binpack-php
>>>
>>>     It’s a binary serialize format. Like JSON, It can be used to transfer
>>> data between languages.
>>>     It’s quite simple. Like JSON, it supports the basic data types: bool
>>> / null / int / double / string / blob / list / map.
>>>     It’s also fast. It may be the fastest serialize implementation in
>>> php. It is 25+ % faster than msgpack.
>>>
>>>     This serialize data format, BIN PACK, is very suitable for RPC. In
>>> our company,  there are lots of services written in Java and  backend API
>>> and web page in PHP. BIN PACK is used to exchange data between JAVA and php.
>>> It’s simpler and  faster than msgpack ;  It is self-descritptive, which is
>>> quite different from other compression serialize formats such as
>>> ProtoBuffer.
>>>
>>> Listed below are some sites for your information:
>>>     Here’s my Github: https://github.com/liaohuqiu.
>>>     StackOverflow: http://stackoverflow.com/users/2446397/srain
>>>     Blog: http://www.liaohuqiu.net/
>>>
>>> Thank you for your time.
>>
>> sorry for delay, I was quite busy recently :<
>>
>> from what I can see,  actually bin pack and msgpack are very similar.
>>
>> and the performance are also very similar (msgpack did more logics).
>> but msgpack pack supports more languages and have more usage already.
>>
>> so, to me, I prefer make bin pack and msgpack computable rather than
>> make a new "bin formart serialize protocol"...
>>
>> what do your think?
> I am not subjecting this to be list in PECL,  that is my 2 cents
>
> Anyway, the codes seems clear,
>
> one suggest, if you don't need MINIT/RINIT , then remove them..
>
> Tyrael, please have a look of this :)
>
> thanks
>>
>> thanks
>>>
>>> Regards,
>>>
>>> Huqiu
>>>
>>> [email protected]
>>
>>
>>
>> --
>> Laruence  Xinchen Hui
>> http://www.laruence.com/
>
>
>
> --
> Laruence  Xinchen Hui
> http://www.laruence.com/



-- 
Laruence  Xinchen Hui
http://www.laruence.com/

--
PECL development discussion Mailing List (http://pecl.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php

Reply via email to