From: | Decibel! <decibel(at)decibel(dot)org> |
---|---|
To: | Dmitry Turin <dmitry(dot)turin(at)belarusbank(dot)minsk(dot)by> |
Cc: | pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: SQL5 budget |
Date: | 2008-11-10 00:41:41 |
Message-ID: | 7DA18BDF-A92C-4B03-8420-217571397065@decibel.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
I think you're barking up the wrong tree here; the community can't
really do hacking for hire. If you want to pay for something to be
implemented (which is great!), you'll need to talk to companies that
do Postgres consulting. You can find examples on the website and
through google. You could also try posting to pgsql-jobs.
On Nov 5, 2008, at 6:33 AM, Dmitry Turin wrote:
> Hi, Pgsql-hackers.
>
> Full list of changes of two related projects (SQL5 & HTML6) at
> current moment
> (not reflected in documentation yet).
>
> ===terms
>
> master table is table, to primary key of which this (slave) table
> refers
> (terms "master" and "slave" are used only for MTD).
> if master table is not represented in the same html-document,
> where slave is (created by "select '<table name="A">'"),
> then new separate browser window will appear, and content of
> master table will shown in it.
> master and slave table must be bound by only one foreing key in
> database.
>
> communication - must be implemented via SQL5 over TCP
> (look at slides #27-28 of http://sql50.euro.ru/sql5.16.4.pdf)
> without closing TCP-connection after obtaining XML-document
>
> because saments are saved in database, but baments are not
> (we don't need to save temporary service information), instead of
> sament
> <_ name= event= arg= >
> the following bament must be sent
> <?mtd name= event= arg= /?>
>
> ===60
>
> in regard to
> http://html60.euro.ru/site/html60/en/author/looker_eng.htm
> http://sql50.euro.ru/site/sql50/en/author/mtd_eng.htm
>
> instead of
> <table name="M" action="./scriptM.cgi" portion="50" visual="20"
> id="id1">
> <table name="S" action="./scriptS.cgi" portion="40" visual="10"
> master="id1">
> the following construction must be implemented
> <table name="M">
> <table name="S">
> with CSS
> table[name=M] { online:yes; portion:50; visual:20 }
> table[name=S] { online:yes; portion:40; visual:10; master:M }
>
> as well as instead of
> <table name="M" action="./scriptM.cgi" portion="50" visual="20"
> id="id1">
> <table name="N" action="./scriptN.cgi" portion="50" visual="20"
> id="id2">
> <table name="P" action="./scriptP.cgi" portion="50" visual="20"
> id="id3">
> <table name="S" action="./scriptS.cgi" portion="40" visual="10"
> master="id1 id2 id3">
> the following construction must be implemented
> <table name="M">
> <table name="N">
> <table name="P">
> <table name="S">
> with CSS
> table[name=M] { online:yes; portion:50; visual:20 }
> table[name=N] { online:yes; portion:50; visual:20 }
> table[name=P] { online:yes; portion:50; visual:20 }
> table[name=S] { online:yes; portion:40; visual:10; master:M N P }
> or (if slave table refers to one master table by 3 own fields)
> table[name=M] { online:yes; portion:50; visual:20 }
> table[name=S] { online:yes; portion:40; visual:10; master:M/§s3 N/
> §s4 P/§s5 }
> where §s3, §s4, §s5 are fields of table "S"
>
>
> instead of sament
> <_ name= event= arg= >
> the following bament must be sent
> <?mtd name= event= arg= /?>
>
> as well as instead of
> <_ name= event= arg= >
> <a a1= a2= a3=>
> the following construction must be sent
> <?mtd name= event= arg= >
> <a a1= a2= a3=>
> </?mtd>
>
> as well as instead of
> <_ name="S" event="TableDown" arg="40">
> <m m1= m2= m3= >
> <n n1= n2= n3= >
> <p p1= p2= p3= >
> the following construction must be sent
> <?mtd name= event= arg= >
> <m m1= m2= m3= >
> <n n1= n2= n3= >
> <p p1= p2= p3= >
> </?mtd>
>
>
> instead of
> <a a1= a2= a3="tiger.jpg">
> <file name="a3" filename="tiger.jpg">Y29udGVudHMgb2YgZmlsZTEudHh0</
> file>
> the following construction must be implemented
> <a a1= a2= a3="tiger.jpg">
> <?file attr="a3" value="tiger.jpg" xml:size=
> >Y29udGVudHMgb2YgZmlsZTEudHh0</file>
>
>
> instead of
> <table name="A" action="./script.cgi" portion="50" visual="20">
> <select name="B" action="./script2.cgi" portion="40"
> visual="10">a2</select>
> </table>
> the following construction must be implemented
> <table name="A"/>
> with CSS
> table [name=A] { online:yes; portion:50; visual:20 }
> table [name=A] { §a2: B } /* 'portion',
> 'visual' are inherited */
>
> in addition to mentioned, statement
> table [name=A] { §a2: B C }
> means, that "C" is master table for "B", "B" is master table for
> "§a2",
> and that browser represents content "C" before content "B" to
> choose sub-set of "B"
>
>
> in addition to mentioned in paper, each 1-st click on column header
> (e.g. concerning to xml-attribute @a2) sends bament
> <?mtd name= event=asc arg=a2/?>
> each 2-nd click sends bament
> <?mtd name= event=dsc arg=a2/?>
> each 3-rd click sends bament
> <?mtd name= event=none arg=a2/?>
>
>
> in addition to mentioned in paper, master table rise appearance
> of new separete browser window
> and displaying content of slave table in it, if
> *) additional virtual column is specified in CSS for master table
> (slave field is specified too, if several fields of slave
> table refer to one master table)
> table[name=M] §m20::after { slave:S/§s5 }
> *) user make double-click on this virtual column
> Relation "master-slave" is served by the same saments, as
> relation "slave-master";
> but refering key will be specified
> <?mtd name= event= arg= >
> <m/§s3 m1= m2= m3= >
> </?mtd>
>
> ---
>
> in regard to
> http://html60.euro.ru/site/html60/en/author/combo_eng.htm
> http://sql50.euro.ru/site/sql50/en/author/mtd_eng.htm
>
> instead of
> <select name="C" action="./script2.cgi" portion="40">
> the following element must be implemented
> <select name="C">
> with CSS
> ul[name=A] { online:yes; portion:40 }
>
>
> instead of sament
> <_ name= event= arg= >
> the following bament must be sent
> <?mtd name= event= arg= /?>
>
> as well as instead of
> <_ name= event= arg= >
> <option id= value=>word40<option>
> the following construction must be sent
> <?mtd name= event= arg= >
> <option id= value=>word40<option>
> </?mtd>
>
>
> ---
>
> in regard to
> http://html60.euro.ru/site/html60/en/author/tree_eng.htm
> http://sql50.euro.ru/site/sql50/en/author/mtd_eng.htm
>
> instead of
> <ul name="A" action="./script.cgi">
> the following element must be implemented
> <ul name="A">
> with CSS
> ul[name=A] { online:yes }
>
> instead of sament
> <_ name= event= arg= >
> the following bament must be sent
> <?mtd name= event= arg= /?>
>
> ---
>
> in regard to
> http://html60.euro.ru/site/html60/en/author/looker_eng.htm
> http://html60.euro.ru/site/html60/en/author/tree_eng.htm
> http://sql50.euro.ru/site/sql50/en/author/mtd_eng.htm
>
> in addition to mentioned in papers, tree as master element for
> table is possible
> ul [name=A] { online:yes }
> table[name=S] { online:yes; portion:40; visual:10; master:A }
> and browser sends at user actions
> <?mtd name=S event= arg= >
> <li id=1>
> </?mtd>
> it's supposed, that foreign key of database table "S" refers to
> primary key of database table "A"
>
> database table, corresponding to HTML-tree, consists of only two
> fields -
> of primary key (extracted as @id), and of any other field with
> any name
> (extracted as content of element LI)
>
> ===other HTML6 additions
>
> pressing browser botton "back" not only display previos XML-
> document, but also sends bament
> <?back/?>
> (which rollback transaction in DBMS to previos 'select ...')
>
> --
>
> bottons to append and remove table row must be implemented in
> menu of browser window
> (implementation in way of WebForms2 is optional)
>
> --
>
> botton to commit transaction must be implemented in menu of
> browser window, pressing on which will send
> <?commit/?>
>
> --
>
> property 'save' means, that at pressing ^S, html/xml-document
> will be saved with entered values
> input { save:yes }
>
> ===50
>
> in regard to
> http://sql50.euro.ru/sql5.16.4.pdf (slide #31,etc)
>
> in addition to mentioned in pdf-document, statement
> SELECT ... COMMENTING.
> extracts additional record from table as first xml-element,
> attributes of which contain comments, specified for each field
> earlier by statement
> COMMENT tab/@fld BY "column title";
> this first xml-element contains additional service xml-attribute
> '_=comment'.
> usually comments are specified for html-tables, instead of html-
> trees
>
> --
>
> in regard to
> http://sql50.euro.ru/sql5.16.4.pdf (slide #18 )
> http://html60.euro.ru/site/html60/en/author/looker_eng.htm
> http://html60.euro.ru/site/html60/en/author/tabfile_eng.htm
> (Sending of form)
> http://html60.euro.ru/site/html60/en/author/cube_eng.htm
> (Sending of form)
>
> in addition to mentioned in pdf-document, obtaining single xml-
> element (instead of xml-tree)
> with primary key (PK) for DBMS table (specified in some xml-
> attribute) means the following:
> *) if this PK is already in DBMS table, then DBMS *updates* fields
> of record with this PK
> *) if this PK is not in table (this usually means, that record with
> this PK was already deleted),
> then DBMS *inserts* this xml-element under new PK (obtained from
> sequence for this PK), i.e. PK will be changed during insertion.
>
> --
>
> in regard to
> http://sql50.euro.ru/sql5.16.4.pdf (slide #31-32, 39(Another
> output), 40)
> http://html60.euro.ru/site/html60/en/author/looker_eng.htm (for
> "select '<table name="A">'", i.e. extracting HTML )
> http://html60.euro.ru/site/html60/en/author/combo_eng.htm
> http://html60.euro.ru/site/html60/en/author/tree_eng.htm
> http://html60.euro.ru/site/html60/en/author/forxml_eng.htm (for
> "select * from A", i.e. extracting XML)
> http://html60.euro.ru/site/html60/en/author/
> quest_eng.htm#webforms_data (<data id="records">...</data>)
>
> each 'select ...' extracts data into LAN or WAN
> (instead of 'insert ... select ...' and 'select ... into ...',
> coping data into internal DBMS table).
> each 'select ... union ... select ...' interpreted as single
> extraction of xml into LAN or WAN.
>
> in addition to mentioned in pdf-document,
> each extraction of data into LAN or WAN is automatically:
> *) anticipated by content of field 'beginning' of system table
> 'SysXmlFormatting'
> *) ended by content of field 'end' of system table
> 'SysXmlFormatting' and
> *) by bament <?new/?> (to inform client, e.g. browser, about end of
> parcel [sent by MTD]
> and end of document [sent by "select * from A"] )
> P.S.
> Other field of system table 'SysXmlFormatting' is URL
> (which specifies browser URL; 'null' means all URL and is applied,
> only if 'SysXmlFormatting' does not contain record with exact URL).
>
> --
>
> in regard to
> http://sql50.euro.ru/sql5.16.4.pdf (slide #173)
>
> authentification (obtaining username and password, slide #173)
> must be implemented by baments
> (look at slides #26-27) and is described nowhere.
>
> ---
>
> in regard to
> http://sql50.euro.ru/site/sql50/en/author/mtd_eng.htm
> http://html60.euro.ru/site/html60/en/author/looker_eng.htm
>
> in addition to mentioned in paper,
> following bament forces to resent data, having appended ascendant
> sorting by field @a2
> <?mtd name= event=asc arg=a2/?>
> following bament forces to resent data, having appended
> descendant sorting by field @a2
> <?mtd name= event=dsc arg=a2/?>
> and following bament forces to resent data, having subtracted
> sorting by field @a2
> <?mtd name= event=none arg=a2/?>
>
> ===other SQL5 additions
>
> in addition to mentioned in pdf-document, each
> 'select ...' (extracting data into LAN/WAN) is savepoint.
> DBMS rollbacks to this savepoint, having obtained bament
> <?back/?>
>
> --
>
> accepting bament
> <?commit/?>
> DBMS commits current transaction to point of previous 'select ...'
>
> --
>
> in addition to mentioned in pdf-document, if table "b" below is
> created
> --------------------------
> | b1 | b2 | b3 |
> |------------------------|
> | | a1 | a2 | a3 |
> |------------------------|
> | |
> | |
> --------------------------
> then expression
> IMPLY b/@b3 as b/@b3/@a2;
> influence so, that any 'SELECT @b3 FROM b' will extract only one
> elementary field (@a2)
> instead of all branch elementary fields (@a1, @a2, @a3),
> i.e. mentioned statement will be equivalent to 'SELECT @b3/@a2
> FROM b'.
> it's for multi-language applications (@a1, @a2, @a3 contain notes
> in different languages)
>
> --
>
> in addition to mentioned in pdf-document, permissions for field
> are controlled:
> GRANT/REVOKE SELECT/INSERT/UPDATE ON tab/@fld FOR user1;
>
> --
>
> in regard to
> http://html60.euro.ru/site/html60/en/author/forxml_eng.htm
> http://www.whatwg.org/specs/web-forms/current-work/#the-output
> http://lists.w3.org/Archives/Public/public-html/2007May/0173.html
>
> DBMS creates separete CSS-file "username.css" for each user
> "username" in local (for DBMS) directory.
> table fields are specified as ¶fieldname in it instead of
> §fieldname, because DBMS does not know,
> what fields are service (like @colspan, @rowspan) or
> are values of properties (like in http://html60.euro.ru/site/
> html60/en/author/chart_eng.htm),
> and what fields contain data for visualization - so mark ¶ acts
> like comment for this xml-attribute
>
> all master tables are specified for each slave table in it,
> i.e. if database table "S" refers to database tables "M", "N", "P",
> then the following lines are in CSS
> table[name=M] { online:yes }
> table[name=N] { online:yes }
> table[name=P] { online:yes }
> table[name=S] { online:yes; master:M N P }
> all slave tables are specified for each master table after last
> column
> table[name=M] ¶m20::after { online:yes; slave:S1; }
> table[name=M] ¶m20::after::after { online:yes; slave:S2; }
> table[name=M] ¶m20::after::after::after { online:yes; slave:S3; }
> table[name=M][_] ¶m20::after { content:"s1_fk_comment" }
> table[name=M][_] ¶m20::after::after { content:"s2_fk_comment" }
> table[name=M][_] ¶m20::after::after::after { content:"s3_fk_comment" }
>
> all accessable database fields (for user "username") are listed
> in this file for
> select * from tab;
> except fields, which are primary keys in database, and which are
> always specified (and never ignored) as invisible
> ¶pk { display: none}
> if user can update database field, than corresponding xml-
> attribute will be specified as re-writable
> t { display: table-row }
> ¶t1, ¶t2, ¶t3 { display: table-cell input} /* read-write */
> if user cannot update database field
> REVOKE update ON t FOR user1;
> then corresponding xml-attribute will be specified as read-only
> t { display: table-row }
> ¶t1, ¶t2, ¶t3 { display: table-cell } /* read only */
> fields, calculated upon other fields and don't saved really,
> specified in CSS as calculated
> via @onforminput, implemented in browser as property (pay
> attention at "§" inside braces)
> ¶t1 { onforminput="value=§t2.value*§t3.value" }
>
> CSS-files are available for editing for DBMS administrator.
> DBMS does not re-write corrections, made manually.
>
>
>
> Dmitry (SQL50, HTML60)
>
>
>
> --
> Sent via pgsql-hackers mailing list (pgsql-hackers(at)postgresql(dot)org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-hackers
>
--
Decibel!, aka Jim C. Nasby, Database Architect decibel(at)decibel(dot)org
Give your computer some brain candy! www.distributed.net Team #1828
From | Date | Subject | |
---|---|---|---|
Next Message | Decibel! | 2008-11-10 00:55:25 | Re: WIP: Page space reservation (pgupgrade) |
Previous Message | Hitoshi Harada | 2008-11-10 00:38:58 | Re: Windowing Function Patch Review -> ROW_NUMBER without ORDER BY |