binarycloud developer mailing list

Carlo Moneti | 1 Feb 2002 02:23

Re: Nested Sets

Thanks for your response. But I still had difficulty seeing the whole solution. So, I went to trusty B&N and
reviewed Joe Celko's "SQL for Smarties". There's a whole chapter on nested set based trees. The query is
complex; he recommends writing a procedure and includes an example.
I'm interested in all of this because I have an idea for a tool to allow rapid website creation (or layout)
from the point of view of an information architect. No more sketching flow diagrams, or creating and
editing visio files, or assembling mock prototypes in html, for either internal brainstorming and
visualization or external review. Just interactively type in a menu item in an input box at the bottom of a
web page and voila! it appears on the the page; click on that item and a blank page displays; enter a menu item
from the new page, it displays and is stored as a child of the previous item(page). The menu, navigation
trail, and default pages are created on the fly. All pages are defined in a hierarchy. Each page is
associated to a form where one can assign meta info, content file, template file, etc (i.e., a set of
directives for assembling that page), as well as reassign, reorder, and delete nodes. Obviously
content, modules, and templates are created separately; bu!

t they are simply flowed in as they are defined and built.
The only thing that is missing for use in large and sophisticated sites is a well thought-out application
platform. Enter BinaryCloud. I'm looking forward to seeing how my component might integrate with BC.
I'll be happy to contribute it to BC if people find it useful.

-Carlo

On 2002.01.29 01:39 alex black wrote:
> Question: How do you select just the immediate children of a node?

you know that child nodes are parallel because adjacent nodes have a
relationship:

L node1 R
R+1=L node2 R

(Continue reading)

headers

Alex Black | 1 Feb 2002 02:36

Re: Nested Sets

> Thanks for your response. But I still had difficulty seeing the whole
> solution. So, I went to trusty B&N and reviewed Joe Celko's "SQL for
> Smarties". There's a whole chapter on nested set based trees. The query is
> complex; he recommends writing a procedure and includes an example.
> I'm interested in all of this because I have an idea for a tool to allow rapid
> website creation (or layout) from the point of view of an information
> architect. No more sketching flow diagrams, or creating and editing visio
> files, or assembling mock prototypes in html, for either internal
> brainstorming and visualization or external review. Just interactively type in
> a menu item in an input box at the bottom of a web page and voila! it appears
> on the the page; click on that item and a blank page displays; enter a menu
> item from the new page, it displays and is stored as a child of the previous
> item(page). The menu, navigation trail, and default pages are created on the
> fly.

Much of that exists or will exist in binarycloud, however instead of using a
GUI initially we use XML. There will be an IDE which will allow people to
visually "build" XML definition files, including individual entities and
complete site maps.

> The only thing that is missing for use in large and sophisticated sites is a
> well thought-out application platform. Enter BinaryCloud. I'm looking forward
> to seeing how my component might integrate with BC. I'll be happy to
> contribute it to BC if people find it useful.

definitely, we were going to build it if nothing suitable was found :)

_a

headers

Carlo Moneti | 1 Feb 2002 03:15

Re: Nested Sets

Excellent!

-Carlo

On 2002.01.31 20:35 Alex Black wrote:
> Thanks for your response. But I still had difficulty seeing the whole
> solution. So, I went to trusty B&N and reviewed Joe Celko's "SQL for
> Smarties". There's a whole chapter on nested set based trees. The query is
> complex; he recommends writing a procedure and includes an example.
> I'm interested in all of this because I have an idea for a tool to allow rapid
> website creation (or layout) from the point of view of an information
> architect. No more sketching flow diagrams, or creating and editing visio
> files, or assembling mock prototypes in html, for either internal
> brainstorming and visualization or external review. Just interactively type in
> a menu item in an input box at the bottom of a web page and voila! it appears
> on the the page; click on that item and a blank page displays; enter a menu
> item from the new page, it displays and is stored as a child of the previous
> item(page). The menu, navigation trail, and default pages are created on the
> fly.

Much of that exists or will exist in binarycloud, however instead of using a
GUI initially we use XML. There will be an IDE which will allow people to
visually "build" XML definition files, including individual entities and
complete site maps.

> The only thing that is missing for use in large and sophisticated sites is a
> well thought-out application platform. Enter BinaryCloud. I'm looking forward
> to seeing how my component might integrate with BC. I'll be happy to
> contribute it to BC if people find it useful.

(Continue reading)

headers

Andris Spruds | 1 Feb 2002 09:33

OT: specifications

Hi everyone,
Can someone send me the latest specs.pdf? They have currenly disabled
everything but email access for my user account here... :) I will get to
a normal connection, but that's going to be late this night or
tomorrow...

Thnx,
Andris

headers

Peter Bowyer | 1 Feb 2002 15:49

Re: OT: specifications

Maybe this is a good time to change the Reply-to header to the original 
sender and let list members use the reply to all button in the mail 
client?  I used to prefer lists configured like binarycloud-dev, but after 
joining many more mailing lists using the reply-to-all button makes more 
sense, and prevents messages like this from choking everyone's connection.

Peter.

At 08:17 AM 2/1/02 -0500, Albert Lash wrote:

--
Maple Design - http://www.mapledesign.co.uk

headers

Fredrik Svensson | 1 Feb 2002 15:58

RFC 7 progress ?!

> Yes, you should only change hits from user/*

Ok,
user/testsite/ent/RevisedDef.php.xml:   <create>0</create>
user/testsite/ent/furbee_type.ent.xml:  <create>0</create>
user/testsite/ent/furbees.ent.xml:      <create>0</create>
user/testsite/mod/tests/schemas/QueryExample.xml:  <create>0</create>

Should be done in 2 sec. Simples thing would be to remove the furbee
example since another bigger, better is comming.

> Actually that syntax should be removed entirely, as <create> will no longer
> be a part of the Entity definition format.

Might be out on deep water here..

I have rewritten the xsl stylesheet (ent2schema). It works for furbee ..
I think :)

Using the recursive template style instead of the iterative for-each.
"The only way to learn recursion is to learn recursion"

Where, how do I place it ?

Had also some thoughts about making a schema definition .xsd
(the DTD of XML) but that takes more time.. and that I don't have :)
What are the plans for schemas (see note) ?
Right now the ent has duplicate tag types for <sequence> for example
one that takes subtags like <name>, <start>, <on>, <table> and <field>
and another that holds the value true. (looking at furbee.. )

(Continue reading)

headers

Andris Spruds | 1 Feb 2002 17:03

Re: Nested Sets

> The reason this is better than the "standard" flat way of doing it (id and
> parent id in the same table) is that you can manipulate infinitely large
> trees and maintain their integrity. If you remove a parent the crappy
model,
> all of the children are immediately orphans. This method allows you to add
> and delete nodes anywhere in the tree.

If you had an infinitely large tree and wanted to insert a item just after
the first item of the tree, you would wait for some 3 hours or more while
MySQL or whatever you use updates almost all records in the table. And if
you used indexes (index on both R&L) it would be even longer...

That's why I dont like the method - it may be fine for static trees, but for
dynamic ones I prefer the old, good id-parent_id way. The biggest problem
here is (with deep trees) when you want to get all the parents of an item.
Usually this means N queries where N is the depth of the item in the tree.
However, I believe, that it's possible to introduce a footprint table where
footprints (footprint=the path to the top of the tree) for each item would
be stored, thus introducing mulpiple queries when adding the item, but
eliminating the need to run multiple queries when receiving the data.

Andris

headers

Andris Spruds | 1 Feb 2002 16:38

Re: query format & loading

If we will have SO many queries, I think I will have no other choice...

Andris
----- Original Message ----- 
From: "Alex Black" <enigma@...>
To: "binarycloud-dev" <dev@...>
Sent: Monday, January 28, 2002 11:40 PM
Subject: [binarycloud-dev] query format & loading

> hey andris,
> 
> I had a look at the query format, what do you think about splitting each
> query into separate files?
> 
> for example:
> 
> user/qry/ent_name/add/*.qry
>                             get/*.qry
>                             set/*.qry
>                             delete/*.qry
> 
> the reason I'm suggesting that is because I think we're going to have a
> large number of queries available by default.
> 
> _a
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...

(Continue reading)

headers

Andris Spruds | 1 Feb 2002 17:22

Re: Movie Database

We are going to need EDFs for movie database? Will you create them?

Andris

headers

alex black | 1 Feb 2002 18:30

Re: RFC 7 progress ?!

> I have rewritten the xsl stylesheet (ent2schema). It works for furbee ..
> I think :)

I'll have a look.

> I had to patch my xsl to handle this.. by using match="database//secuence"

Yep, that's fine, need context anyway.

> Note : The name XML schema is a W3 standard, while schema as BC
>        uses it is a Database schema expressed in XML.
>        The ent and schema xml should maybe both have a schema file
>        sometimes in the future..

Probably xml schemas, not DTDs, as they are annoying to write :) Certainly
all of the XML in the system will have schema built, I want to wait until
we're release to do that.

_a

Mon	Tue	Wed	Thu	Fri	Sat	Sun

				1	2	3
4	5	6	7	8	9	10
11	12	13	14	15	16	17
18	19	20	21	22	23	24
25	26	27	28

2	January 2006
26	December 2005
45	November 2005
5	October 2005
6	September 2005
31	August 2005
34	July 2005
28	June 2005
59	May 2005
18	April 2005
148	March 2005
534	February 2005
96	January 2005
102	December 2004
70	November 2004
66	October 2004
326	September 2004
647	August 2004
695	July 2004
246	June 2004
289	May 2004
193	April 2004
281	March 2004
166	February 2004
143	January 2004
227	December 2003
457	November 2003
338	October 2003
112	September 2003
99	August 2003
414	July 2003
340	June 2003
394	May 2003
642	April 2003
291	March 2003
370	February 2003
297	January 2003
465	December 2002
365	November 2002
617	October 2002
874	September 2002
435	August 2002
728	July 2002
1355	June 2002
875	May 2002
600	April 2002
1192	March 2002
1112	February 2002
507	January 2002
661	December 2001
865	November 2001
5	July 2000