www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - D based BEEP library?

reply =?iso-8859-1?Q?Robert_M._M=FCnch?= <robert.muench saphirion.com> writes:
Hi, since I'm a big fan of BEEP has anyone tried or started to 
implement a BEEP lib in D?


For BEEP see:

http://en.wikipedia.org/wiki/BEEP
http://beepcore.org/
http://www.aspl.es/vortex/ (C based lib)

--
Robert M. Münch
Saphirion AG

http://www.saphirion.com
smarter | better | faster
Jan 04 2013
parent reply "Rob T" <rob ucora.com> writes:
On Friday, 4 January 2013 at 23:44:35 UTC, Robert M. Münch wrote:

 Hi, since I'm a big fan of BEEP has anyone tried or started to 
 implement a BEEP lib in D?


 For BEEP see:

 http://en.wikipedia.org/wiki/BEEP
 http://beepcore.org/
 http://www.aspl.es/vortex/ (C based lib)

 --
 Robert M. Münch
 Saphirion AG

 http://www.saphirion.com
 smarter | better | faster


For whatever reason BEEP never took off, so it's probably 
unlikely anyone will  write a D version of it. Google is now 
promoting SPDY as an alternative, and I would say it already has 
a lot more traction than BEEP does.

Since you are a fan of BEEP, you must know a lot more than I do 
about the subject. Your thoughts on BEEP vs SPDY?

--rt
Jan 04 2013
next sibling parent reply "Rob T" <rob ucora.com> writes:
I forgot to add that you can always create D bindings directly to 
the Vortex C library which will allow you to use BEEP with your D 
applications via Vortex. That's one of the cool things about D, 
it is fully compatible with the C ABI.

--rt
Jan 04 2013
parent =?iso-8859-1?Q?Robert_M._M=FCnch?= <robert.muench saphirion.com> writes:
On 2013-01-05 03:45:11 +0000, Rob T said:


 I forgot to add that you can always create D bindings directly to the 
 Vortex C library which will allow you to use BEEP with your D 
 applications via Vortex. That's one of the cool things about D, it is 
 fully compatible with the C ABI.


Yes, I know. Nevertheless, I think such an idea can make big use of D's 
features to create a very reliable library.

--
Robert M. Münch
Saphirion AG

http://www.saphirion.com
smarter | better | faster
Jan 11 2013
prev sibling parent reply =?iso-8859-1?Q?Robert_M._M=FCnch?= <robert.muench saphirion.com> writes:
On 2013-01-05 03:40:00 +0000, Rob T said:


 For whatever reason BEEP never took off, so it's probably unlikely 
 anyone will  write a D version of it.


I don't understand why it didn't took off. Maybe people didn't get 
what's the power behind it and how simple you can make your life for 
all network related things.


  Google is now promoting SPDY as an alternative, and I would say it 
 already has a lot more traction than BEEP does. Since you are a fan of 
 BEEP, you must know a lot more than I do about the subject. Your 
 thoughts on BEEP vs SPDY?


SPDY is more like an enhanced-HTML. So it's clearly trageted at browser 
and the web.

BEEP is a generic framework to desing all kind of application network 
protocolls. It's NOT a network protocol nor comes it with any specific 
protocol.

In a nut-shell you get 2^32-1 channels per connection, you can send 
arbitrary big messages on each channel in parallel, you can use 
REQ/RESP, P2P, streaming etc. model whatever fits.

The nice thing is, that you can support different version of your 
application protocol through different channels (profiles). So, a V2 
app can still have the V1 protocol to speak to a server A and at the 
same time use V2 protocol to speak to a server B.

It really saves you from 90% of all housekeeping code for network stuff.

BTW: Yes, the C binding is no big deal. And the vortex lib is very 
matured. IMO the code is a bit complex and sometimes hard to handle. I 
think using D will make the implementaiton more robust and simple. 
That's why I asked.

--
Robert M. Münch
Saphirion AG

http://www.saphirion.com
smarter | better | faster
Jan 05 2013
next sibling parent reply Nick Sabalausky <SeeWebsiteToContactMe semitwist.com> writes:
On Sat, 5 Jan 2013 10:56:00 +0100
Robert M. M=FCnch <robert.muench saphirion.com> wrote:


 On 2013-01-05 03:40:00 +0000, Rob T said:
=20

 For whatever reason BEEP never took off, so it's probably unlikely=20
 anyone will  write a D version of it.


=20
 I don't understand why it didn't took off.


This is the first I've heard of BEEP, but my guess so far is that the
main website for it:

http://beepcore.org/

Provides no explanation for how to use it beyond linking to a series of
long-winded and poorly-formatted RFCs, plus the site doesn't offer a
clear link to any ready-to-use lib. Either of those problems alone is
enough to turn away most people.

In other words, bad marketing. Unfortunate, since it sounds like a good
idea upon my first glance of it (aside from its choice to use XML for
certain things, which IMO is too much of an _unnecessary_ baggage for
something as low level as BEEP.)


 Maybe people didn't get=20
 what's the power behind it and how simple you can make your life for=20
 all network related things.
=20


It may very well do that, but unfortunately, figuring out how to get up
and running with it doesn't appear to be simple at all, at least
if you're looking at beepcore.org. That would certainly hinder its
ability to hit critical mass and really take off.

I don't really get why some software engineers seem to think that in
20xx they can write up a series of code-numbered legalese-esque
documents (and with no formatting, and with baked-in page-breaks
despite being in electronic format), and expect that people will pay
attention to it.

It's kinda like how academic folk will write overly-convoluted (almost
patent-like) explanations, employ other forms of obfuscation such as
calling a summary or intro an "abstract" (just because some outdated
standard tells them to), stick it all into a multi-column PDF, and then
wonder why the non-academic side never bothers to pay any attention.
Jan 05 2013
next sibling parent reply "Joakim" <joakim airpost.net> writes:
On Saturday, 5 January 2013 at 19:54:11 UTC, Nick Sabalausky 
wrote:

 In other words, bad marketing. Unfortunate, since it sounds 
 like a good
 idea upon my first glance of it (aside from its choice to use 
 XML for
 certain things, which IMO is too much of an _unnecessary_ 
 baggage for
 something as low level as BEEP.)


I lost interest when I saw the XML mentions.


 I don't really get why some software engineers seem to think 
 that in
 20xx they can write up a series of code-numbered legalese-esque
 documents (and with no formatting, and with baked-in page-breaks
 despite being in electronic format), and expect that people 
 will pay
 attention to it.


The horrific formatting of those RFCs surprised me also.


 It's kinda like how academic folk will write overly-convoluted 
 (almost
 patent-like) explanations, employ other forms of obfuscation 
 such as
 calling a summary or intro an "abstract" (just because some 
 outdated
 standard tells them to), stick it all into a multi-column PDF, 
 and then
 wonder why the non-academic side never bothers to pay any 
 attention.


You assume they want someone to pay attention.  The real goal is 
slip under the radar, keep collecting the free government money 
for doing nothing of any import.
Jan 05 2013
parent Joseph Rushton Wakeling <joseph.wakeling webdrake.net> writes:
On 01/05/2013 09:07 PM, Joakim wrote:

 You assume they want someone to pay attention.  The real goal is slip under the
 radar, keep collecting the free government money for doing nothing of any
import.


Most academics would love attention (which usually gets you more money) and 
would love to be doing something of import.  The problem is that if you don't 
get sufficient publications in the right journals or conferences, with their 
mandated publication formats, you don't get the career brownie points required 
to keep getting funding (government or otherwise).

It's lock-in via the requirements of the funding organizations, unfortunately.
:-(
Jan 05 2013
prev sibling next sibling parent reply "Rob T" <rob ucora.com> writes:
On Saturday, 5 January 2013 at 19:54:11 UTC, Nick Sabalausky 
wrote:

 Provides no explanation for how to use it beyond linking to a 
 series of
 long-winded and poorly-formatted RFCs, plus the site doesn't 
 offer a
 clear link to any ready-to-use lib. Either of those problems 
 alone is
 enough to turn away most people.


Funny you mentioned this, because a few years back when I thought 
BEEP was a great idea, that's exactly what happened to me. I 
could not easily get a grasp on how it worked, there were not 
even any examples. What really killed it for me was the only BEEP 
library I could find was broken, and I really did not want to try 
patching it up when I did not even fully understand what it was 
supposed to be doing.


 In other words, bad marketing. Unfortunate, since it sounds 
 like a good


Try searching Google for BEEP, bad choice of name. But then we 
use "D"!


 idea upon my first glance of it (aside from its choice to use 
 XML for
 certain things, which IMO is too much of an _unnecessary_ 
 baggage for
 something as low level as BEEP.)


Yeah, XML is always a turn off for me too. JSON is better, but 
even still ...


 Maybe people didn't get what's the power behind it and how 
 simple you can make your life for all network related things.
 


 It may very well do that, but unfortunately, figuring out how 
 to get up
 and running with it doesn't appear to be simple at all, at least
 if you're looking at beepcore.org. That would certainly hinder 
 its
 ability to hit critical mass and really take off.

 I don't really get why some software engineers seem to think 
 that in
 20xx they can write up a series of code-numbered legalese-esque
 documents (and with no formatting, and with baked-in page-breaks
 despite being in electronic format), and expect that people 
 will pay
 attention to it.

 It's kinda like how academic folk will write overly-convoluted 
 (almost
 patent-like) explanations, employ other forms of obfuscation 
 such as
 calling a summary or intro an "abstract" (just because some 
 outdated
 standard tells them to), stick it all into a multi-column PDF, 
 and then
 wonder why the non-academic side never bothers to pay any 
 attention.


We should though look into the mirror wrt to D. I'm not 
suggesting that D is anywhere near as dysfunctional, it is not, 
but there's plenty of room for improvements. The worse we can do 
is not think so.

--rt
Jan 05 2013
parent reply =?iso-8859-1?Q?Robert_M._M=FCnch?= <robert.muench saphirion.com> writes:
On 2013-01-05 22:33:45 +0000, Rob T said:


 Funny you mentioned this, because a few years back when I thought BEEP 
 was a great idea, that's exactly what happened to me. I could not 
 easily get a grasp on how it worked, there were not even any examples. 
 What really killed it for me was the only BEEP library I could find was 
 broken, and I really did not want to try patching it up when I did not 
 even fully understand what it was supposed to be doing.


The Vortex library is a good one (if not the only really matured one), 
actively developed and maintained. It's a no-brainer to get it 
compiled, has a lot of examples, test-suite, good docs etc. all there.


 Try searching Google for BEEP, bad choice of name. But then we use "D"!


:-) Yep.

--
Robert M. Münch
Saphirion AG

http://www.saphirion.com
smarter | better | faster
Jan 11 2013
parent reply "Rob T" <alanb ucora.com> writes:
On Friday, 11 January 2013 at 09:13:56 UTC, Robert M. Münch wrote:

 The Vortex library is a good one (if not the only really 
 matured one), actively developed and maintained. It's a 
 no-brainer to get it compiled, has a lot of examples, 
 test-suite, good docs etc. all there.


Back in the day when I first tried BEEP, All I knew about was the 
beepcore-c library, but I see it's now dead, and it may have been 
dead already when I was attempting to use it. I was not as 
knowledgeable back then.

Ah crap, now you got me reconsidering BEEP again ...

BTW my understanding of the XML is that it's only used minimally 
for control and configuration reasons. However given perhaps 
better alternatives these days, a re-imagining of BEEP may be a 
good idea.


 Try searching Google for BEEP, bad choice of name. But then we 
 use "D"!


 :-) Yep.


... and I could go on pointing out many other failings with D 
that are in-line with doing the opposite of what you should be 
doing in terms of marketing, but it's all fixable, one step at a 
time, just needs the right people to make it so.

--rt
Jan 11 2013
parent =?iso-8859-1?Q?Robert_M._M=FCnch?= <robert.muench saphirion.com> writes:
On 2013-01-11 17:54:33 +0000, Rob T said:


 Back in the day when I first tried BEEP, All I knew about was the 
 beepcore-c library, but I see it's now dead, and it may have been dead 
 already when I was attempting to use it. I was not as knowledgeable 
 back then.


Yes, take a look at Vortex. It plays in a total different league.


 Ah crap, now you got me reconsidering BEEP again ...


:-) You are not alone. So, let me know or PM me.


 BTW my understanding of the XML is that it's only used minimally for 
 control and configuration reasons. 


Right. It's really no big deal. Vortex converts the XML config stuff 
into a compile unit and links it directly into your app. So no config 
files necessary.

And as long as you don't need generic servers that can handle protocol 
announcements etc. you just make a channel and your peers know what to 
do with it.


 However given perhaps better alternatives these days, a re-imagining of 
 BEEP may be a good idea.


IMO it is. There is a JavaScript, Lua and Python version too. Since 
mobile apps are en vogue and these all need network stuff, IMO BEEP is 
a great thing. I make my life simple and avoid wasting time with the 
basics of good protocol design and implementation.

-- 
Robert M. Münch
Saphirion AG

http://www.saphirion.com
smarter | better | faster
Jan 12 2013
prev sibling parent =?iso-8859-1?Q?Robert_M._M=FCnch?= <robert.muench saphirion.com> writes:
On 2013-01-05 19:54:11 +0000, Nick Sabalausky said:


 This is the first I've heard of BEEP, but my guess so far is that the
 main website for it:
 
 http://beepcore.org/
 
 Provides no explanation for how to use it beyond linking to a series of
 long-winded and poorly-formatted RFCs, plus the site doesn't offer a
 clear link to any ready-to-use lib. Either of those problems alone is
 enough to turn away most people.


Sure, it's not marketed professional. There is a lot of stuff in the IT 
sector that suffers this. I learned not to decided on this but take a 
look, at least spend some time to get an understanding. Most of the 
time it paid off.


 In other words, bad marketing. Unfortunate, since it sounds like a good
 idea upon my first glance of it


It is.


 (aside from its choice to use XML for
 certain things, which IMO is too much of an _unnecessary_ baggage for
 something as low level as BEEP.)


Well, I don't like XML too, but it's not a showstopper in this case. 
And the Vortex lib doesn't rely on it.

--
Robert M. Münch
Saphirion AG

http://www.saphirion.com
smarter | better | faster
Jan 11 2013
prev sibling parent reply "Joakim" <joakim airpost.net> writes:
On Saturday, 5 January 2013 at 09:56:01 UTC, Robert M. Münch 
wrote:

 SPDY is more like an enhanced-HTML. So it's clearly trageted at 
 browser and the web.


I think you mean "enhanced HTTP," as SPDY will be the first draft 
for HTTP 2.0:

http://lists.w3.org/Archives/Public/ietf-http-wg/2012OctDec/0447.html

Considering HTTP is now used by all kinds of apps that have 
nothing to do with browsers or the web- although in large part to 
sneak around firewalls- it certainly won't be limited to web 
browsers.
Jan 05 2013
parent =?iso-8859-1?Q?Robert_M._M=FCnch?= <robert.muench saphirion.com> writes:
On 2013-01-05 20:15:31 +0000, Joakim said:


 I think you mean "enhanced HTTP," as SPDY will be the first draft for HTTP 2.0:
 
 http://lists.w3.org/Archives/Public/ietf-http-wg/2012OctDec/0447.html
 
 Considering HTTP is now used by all kinds of apps that have nothing to 
 do with browsers or the web- although in large part to sneak around 
 firewalls- it certainly won't be limited to web browsers.


Well, yes, but I don't think HTTP is useful nor the way to go for 
everything. Same as web-apps, I just don't like them.

But anyway, it's a taste of style and there is no right or wrong.

I'm only interested in being more productive than all the others. So, 
using BEEP I'm just there when others are still fiddling around with a 
plentora of technologies.

-- 

--
Robert M. Münch
Saphirion AG

http://www.saphirion.com
smarter | better | faster
Jan 11 2013