[ppa-dev] Bug in Karl's honk'n patch

[Karl O. Pinc <kop@me...>]

Robert,

I'm seeing these errors in my logs from a ppa that
I've applied my patch to.  Have you found and
fixed these or should I do it?  If I fix it do you
want a whole new big patch or just a patch for
the fix?

Sorry I didn't catch this before.

Thanks.

(In all cases referrer is sqledit.php and outfmt=3Dundefined
in the URL's query string.)

  PHP Notice:  Undefined index:  outfmt in =20
/disk/srv/www/html/phpPgAdmin/sql.php on line 78

  PHP Notice:  Undefined index:  outfmt in =20
/disk/srv/www/html/phpPgAdmin/sql.php on line 283

  PHP Notice:  Undefined index:  outfmt in =20
/disk/srv/www/html/phpPgAdmin/sql.php on line 325

Karl <kop@...>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein

Re: [ppa-dev] ppa 4.1 release plan - 8.2 testing

[Javier Carlos <javier@ne...>]

   Robert,

   I'm sending you patches for the warnings I found when testing with
8.2beta, some of these were also present in 8.1.

   Cheers,

   Javier


> The following is my "hitlist" for getting to a 4.1 release.
>
> Gavin's slony patch & The autovacuum editing screens are now down.
>
> Chris role patch: still missing the critical file...  since Chris has not
> turned up with it, I am inclined to revert us back to the user/group
> screens...  I'd much prefer the file show up, or that someone step up and
> offer to code a new interface (we still have all the back-end
> functionality
> that we need afaik)
>
> John's create table w/ constraints patch:  i think we decided that the
> changes
> needed to be wrapped in a transaction if it would be accepted, I'd like to
> see that finished.  I'm not going to do it though, so unless someone steps
> up
> to do it, this patch will get dropped.
>
> Karl's wompin patch:  I still haven't looked at this in depth, but it is
> now
> at the top of my list.  Some issues have been passed on to me that I want
> to
> look at (things like not showing the correct rows affected when doing an
> insert or delete, and that we should show the download button for queries
> that tossed errors and more) which I will post feedback to the list about.
> This is probably the last patch I plan to look at before beta, and
> honestly
> if it gets too hairy I might drop this till 4.2.
>
> Safari issues:  I passed this info (see bugtracker) on to ioguix, who I am
> hoping is planning to look at them.  If anyone else wants to step up and
> review / test/ fix that please do.
>
> 8.2 testing.  I did some basic testing against 8.2 and the most serious
> thing
> seemed to be excessive warnings being generated for LIKE '\\' style
> queries.
> I fixed the common ones, but there are more that should be fixed and
> tested.
> I think Javier Carlos is looking into this, but since they are Warnings
> I'm
> not going to wait for it to get done.
>
> Other bugs from the bug tracker... well, we cant be any worse if we ship
> with
> existing bugs, so I'm not stopping for those either. That said fixes are
> welcome. :-)
>
> BTW, I'm currently without dedicated internet, so that is crimping my
> development efforts.  8.2 will probably ship in November, I'd like to ship
> 4.1 by then if we can.
>
> --
> Robert Treat
> Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share
> your
> opinions on IT & business topics through brief surveys -- and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> _______________________________________________
> phpPgAdmin-devel mailing list
> phpPgAdmin-devel@...
> https://lists.sourceforge.net/lists/listinfo/phppgadmin-devel
>


-------------
nediam.com.mx

Re: [ppa-dev] ppa 4.1 release plan

[Karl O. Pinc <kop@me...>]

On 10/19/2006 09:15:00 PM, Robert Treat wrote:
> On Tuesday 17 October 2006 11:09, Karl O. Pinc wrote:
> > On 10/01/2006 07:07:17 PM, Robert Treat wrote:

> I've reviewed the patch and have a list of about a dozen issues I'd
> like to
> have addressed. I am inclined to check in what you have sent so far
> but want
> to look at a few more things before I decide for sure, but thought I
> would
> give you a heads up. Many thanks to Javier Carlos for his feedback as
> well.

Thanks much for the heads-up and for looking at such a
honker.  Sorry for not having the self-discipline
to finish the refactoring without charging off
and coding in other directions.

Did you and Javier Carlos have your discussion
in some public place?  I'd have been curious
to have listened in.

Karl <kop@...>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein

Re: [ppa-dev] ppa 4.1 release plan

[Robert Treat <xzilla@us...>]

On Tuesday 17 October 2006 11:09, Karl O. Pinc wrote:
> On 10/01/2006 07:07:17 PM, Robert Treat wrote:
> > Karl's wompin patch:
>
>   Some issues have been passed on to me that I
>
> > want to
> > look at (things like not showing the correct rows affected when doing
> > an
> > insert or delete, and that we should show the download button for
> > queries
> > that tossed errors and more) which I will post feedback to the list
> > about.
>
> Please let me know if this is something you want me to charge ahead
> and fix.  I'd prefer to wait for feedback to make sure that
> you'd approve my approach.  (I'd even more prefer not to have
> to do the work at the moment.  I've fantasies about you
> reading the code, understanding the problem, and coding the
> fix right there.)  I'd also like feedback to ensure I'm thinking
> of the same problems you are.
>
> I'm very pressed for time at the moment but might be able
> to do something if I knew it'd make all the difference
> in getting the patch in.
>

I've reviewed the patch and have a list of about a dozen issues I'd like to 
have addressed. I am inclined to check in what you have sent so far but want 
to look at a few more things before I decide for sure, but thought I would 
give you a heads up. Many thanks to Javier Carlos for his feedback as well. 

-- 
Robert Treat
Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL

Re: [ppa-dev] About code cleaning...

[Robert Treat <xzilla@us...>]

On Wednesday 18 October 2006 03:55, ioguix wrote:
> Robert Treat wrote:
> >> (OTOH, the same can be said
> >> about adding a database to an application, or
> >> a build system, or packaging, or anything else.
> >> There's complexity tradeoffs that eventually reach
> >> a tipping point once the application gets big enough.
> >> I don't feel expert enough in ppa or xml/xslt to
> >> guarentee that it's a good fit, but that's my
> >> inclination.  But only if somebody's excited about
> >> doing the work.  I'm also a big proponent of
> >> "If it ain't broke, don't fix it.")
>
> Well, nobody said we should use xml/xslt in ppa. I just said it was an
> interesting concept that I would give a try on my own...
>
> > Me too... do people really find the code that hard to work with?
>
> No ! ...Actually, I just had to learn more about the famous decorator
> patern ;)
>
> >  IMO the only bowels of the system now are the tree code and the adodb
> > bits...
>
> Well, I am *really* not an expert in js, but with some time and logic,
> it's pretty easy to deal with the tree code.
> About the adodb and stuffs comming with, my opinion is once you
> understood the object design of the beast, it's pretty easy to dive in.
>
> > I can understand the desire to try to make it more homogenious, but
> > adding something like xml/xslt will only push developers away...
>
> Ok, so what do you think about merging html generation classes and drop
> presentation (html) from the logical part (php) (to borrow karl's words) ?
> Moreover, it could be a way to answer this TODO list's point : "One day
> we should make it all XHTML"
>

I'm not against the idea *in theory*, but I've generally found that in 
practice doing these things is a lot harder than it looks (usually you find a 
couple little corner cases that are real hard to deal with).  Couple that 
with all of the the other items on the TODO list and it doesn't excite me all 
that much.  

> Again, not that I have lot of time to loose, but it's just to discuss
> and anderstand how ppa will grow. Finally, decisions belong to you Rob,
> you are driving ppa now ;)
>

Scary thought...   so in theory my management style would be that the best way 
to win a discussion is to send lot's of small, easily digestible patches. 
This shows that not only do you believe in the idea, but it's one that can be 
implemented, and other people can see it first hand.  Generally unless it's 
just a really bad idea, this convinces most people. 

-- 
Robert Treat
Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL

Re: [ppa-dev] Availability

[ioguix <ioguix@fr...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


Robert Treat wrote:
> 
> Hmm... maybe you should consider moving to Maryland, then you can work for my 
> company which does both PHP & PostgreSQL related stuff (among other things).  
> We've got a number of open positions too (http://omniti.com/people/jobs). :-)
> 
Mmmh, do you need some french gui with bad accent ??

:D

Aaaah, if only I would like to be in maryland and wasn't in position, it
could be serious...(I should first try to visit USA ;))

Cheers,
- --
ioguix
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFNeCIxWGfaAgowiIRAgMOAKCMsnE6iOFgTb0x8nir4EdDaNMWkQCgpWkn
Ek1M21It2giTUtMYzKz+CNc=
=gVye
-----END PGP SIGNATURE-----

Re: [ppa-dev] About code cleaning...

[ioguix <ioguix@fr...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1




Robert Treat wrote:
> 
>> (OTOH, the same can be said
>> about adding a database to an application, or
>> a build system, or packaging, or anything else.
>> There's complexity tradeoffs that eventually reach
>> a tipping point once the application gets big enough.
>> I don't feel expert enough in ppa or xml/xslt to
>> guarentee that it's a good fit, but that's my
>> inclination.  But only if somebody's excited about
>> doing the work.  I'm also a big proponent of
>> "If it ain't broke, don't fix it.")

Well, nobody said we should use xml/xslt in ppa. I just said it was an
interesting concept that I would give a try on my own...

> Me too... do people really find the code that hard to work with?

No ! ...Actually, I just had to learn more about the famous decorator
patern ;)

>  IMO the only bowels of the system now are the tree code and the adodb bits...

Well, I am *really* not an expert in js, but with some time and logic,
it's pretty easy to deal with the tree code.
About the adodb and stuffs comming with, my opinion is once you
understood the object design of the beast, it's pretty easy to dive in.

> I can understand the desire to try to make it more homogenious, but adding 
> something like xml/xslt will only push developers away... 

Ok, so what do you think about merging html generation classes and drop
presentation (html) from the logical part (php) (to borrow karl's words) ?
Moreover, it could be a way to answer this TODO list's point : "One day
we should make it all XHTML"

Again, not that I have lot of time to loose, but it's just to discuss
and anderstand how ppa will grow. Finally, decisions belong to you Rob,
you are driving ppa now ;)

Cheers,

- --
ioguix
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFNd3lxWGfaAgowiIRApHrAKCdGYw0xxF0ypkP711BXWJHwQonxgCeL90X
TjoINFPKj7hi/+XfGIaNl8E=
=bpmk
-----END PGP SIGNATURE-----

Re: [ppa-dev] About code cleaning...

[Karl O. Pinc <kop@me...>]

On 10/17/2006 07:29:14 PM, Robert Treat wrote:

> Me too... do people really find the code that hard to work with?  IMO
> the only
> bowels of the system now are the tree code and the adodb bits...

The adodb bits, uh, bit.  But the rest of the code was not
so hard.  Can't say that I've got an overview, but I was
able to go in and deal with a piece of it.  That's
all that can be expected.


Karl <kop@...>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein

Re: [ppa-dev] Availability

[Robert Treat <xzilla@us...>]

On Sunday 15 October 2006 00:19, Dan Wilson wrote:
> Chris,
>
> It's sad to see you go.  I'm sure you will still lurk on the mailing
> list as I do.
>
> I would just like to publicly thank Chris for his dedication and the
> many, many hours of coding he has done over the years.  As the
> "grandfather" of this project, I have never felt at all like it was
> not in good hands while Chris was at the helm.
>

Agreed.  I don't know where we would be without the rewrite for 7.3 which was 
certainly Chris vision. 

> And going forward, I feel completely comfortable with Rob as the lead.
>  I'm just glad the project didn't fall on it's face when I hit the
> point that Chris just hit (not that I ever thought it would in his care).
>

Thanks for the vote of confidence.... I wish I felt as comfortable. :-)    

> Just a big thanks for all your hard work!
>
> -=Dan
>
> PS.  As a side note (to anyone), my company is hiring [1].  Anyone
> willing to move to Orem, UT and work at Omniture?  We really need some
> really sharp minds and we do most of our stuff in PHP.  If you haven't
> heard of us, we are one of (if not THE) leading web analytics (think
> webalizer/awstats on some serious steroids) companies in the world.
> We just went public in June. Big drawback is we use MySQL
> exclusively... not that I don't complain about that on a daily basis.
>  If you are interested, send me your resume.
>
> [1] http://www.omniture.com/career/148
>

Hmm... maybe you should consider moving to Maryland, then you can work for my 
company which does both PHP & PostgreSQL related stuff (among other things).  
We've got a number of open positions too (http://omniti.com/people/jobs). :-)

-- 
Robert Treat
Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL

Re: [ppa-dev] About code cleaning...

[Robert Treat <xzilla@us...>]

On Tuesday 17 October 2006 12:19, Karl O. Pinc wrote:
> On 10/17/2006 11:04:12 AM, ioguix wrote:
> > Maybe I should restrained my tendency to give in complexe coding
>
> Well, nothing like adding an xml/xslt layer to add
> complexity either! :-)
>

That's an understatement. 

> (OTOH, the same can be said
> about adding a database to an application, or
> a build system, or packaging, or anything else.
> There's complexity tradeoffs that eventually reach
> a tipping point once the application gets big enough.
> I don't feel expert enough in ppa or xml/xslt to
> guarentee that it's a good fit, but that's my
> inclination.  But only if somebody's excited about
> doing the work.  I'm also a big proponent of
> "If it ain't broke, don't fix it.")
>

Me too... do people really find the code that hard to work with?  IMO the only 
bowels of the system now are the tree code and the adodb bits... I can 
understand the desire to try to make it more homogenious, but adding 
something like xml/xslt will only push developers away... 

> There might also be cross platform issues with xslt if
> ppa is supposed to run under IIS.  Couldn't say.
>

It's suppose to, and I have heard of people getting it to work. 

-- 
Robert Treat
Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL

Re: [ppa-dev] About code cleaning...

[Karl O. Pinc <kop@me...>]

On 10/17/2006 11:04:12 AM, ioguix wrote:

> Maybe I should restrained my tendency to give in complexe coding

Well, nothing like adding an xml/xslt layer to add
complexity either! :-)

(OTOH, the same can be said
about adding a database to an application, or
a build system, or packaging, or anything else.
There's complexity tradeoffs that eventually reach
a tipping point once the application gets big enough.
I don't feel expert enough in ppa or xml/xslt to
guarentee that it's a good fit, but that's my
inclination.  But only if somebody's excited about
doing the work.  I'm also a big proponent of
"If it ain't broke, don't fix it.")

There might also be cross platform issues with xslt if
ppa is supposed to run under IIS.  Couldn't say.

Karl <kop@...>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein

Re: [ppa-dev] About code cleaning...

[ioguix <ioguix@fr...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Karl O. Pinc wrote:
> 
> On 10/16/2006 09:44:44 AM, ioguix wrote:
>> Hello,
>>
>> I would like to have your point of view about code cleaning.
>>
>> In a first step, let's talk about html generation.
> 
> My inclination when it comes to separation of
> presentation (html) from logic (php) is to not
> re-invent the wheel and to use xslt to generate
> html from xml files, filled with (xml) data generated
> by php.

Well, let's talk about xml generation :D ...sorry, too easy...

> You can see the page layout in the
> xml, and read the logic in the php without
> distracting imbedded html.  And you don't have
> to write a whole html generation system that
> nobody but the ppa developers will be familiar with.

Ok, it sounds like a pretty good architecture. I never worked with xslt
before, but I guess, it would needs to write a dtd and xslt processor...

> However the truth of the matter is that when
> it comes to this sort of rework there's
> no point in making up projects for other people
> to do.  What will happen in the real world is that
> whoever does the work will decide on the
> design.

My opinion is that with a good architecture, guidelines and coding
conventions, we can spend more time on what's really matter (here, to
not distract with html). Project will become more homogeneous, readable
& easiest to maintain as contributors would not always maintain their codes.

On the other hand, it open the door to attract more code contributors...

Maybe I should restrained my tendency to give in complexe coding (The
Cathedral and the Bazaar, Eric Steven Raymond)...

Well, I just wanted to have some opinions, and so far, I got one with in
bonus a great concept to study on my own :)

Thanks :)
- --
+-----------------------+
| IoGuiX                |
| ioguix@...        |
| GPG/PGP ID : 0828C222 |
+-----------------------+
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFNP78xWGfaAgowiIRArhwAJ94b/zEh73y8NEbE9tj+tmT/dFGyACgivsZ
/hRVZrQT/FO0Wpdz3hclGDs=
=KeGM
-----END PGP SIGNATURE-----

Re: [ppa-dev] ppa 4.1 release plan

[Karl O. Pinc <kop@me...>]

On 10/01/2006 07:07:17 PM, Robert Treat wrote:

> Karl's wompin patch:

  Some issues have been passed on to me that I
> want to
> look at (things like not showing the correct rows affected when doing
> an
> insert or delete, and that we should show the download button for
> queries
> that tossed errors and more) which I will post feedback to the list
> about.

Please let me know if this is something you want me to charge ahead
and fix.  I'd prefer to wait for feedback to make sure that
you'd approve my approach.  (I'd even more prefer not to have
to do the work at the moment.  I've fantasies about you
reading the code, understanding the problem, and coding the
fix right there.)  I'd also like feedback to ensure I'm thinking
of the same problems you are.

I'm very pressed for time at the moment but might be able
to do something if I knew it'd make all the difference
in getting the patch in.

Karl <kop@...>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein

Re: [ppa-dev] About code cleaning...

[Karl O. Pinc <kop@me...>]

On 10/16/2006 09:44:44 AM, ioguix wrote:
> Hello,
>=20
> I would like to have your point of view about code cleaning.
>=20
> In a first step, let's talk about html generation.

My inclination when it comes to separation of
presentation (html) from logic (php) is to not
re-invent the wheel and to use xslt to generate
html from xml files, filled with (xml) data generated
by php.  You can see the page layout in the
xml, and read the logic in the php without
distracting imbedded html.  And you don't have
to write a whole html generation system that
nobody but the ppa developers will be familiar with.

Standard is better than better.

However the truth of the matter is that when
it comes to this sort of rework there's
no point in making up projects for other people
to do.  What will happen in the real world is that
whoever does the work will decide on the
design.

Karl <kop@...>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein

[ppa-dev] About code cleaning...

[ioguix <ioguix@fr...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

I would like to have your point of view about code cleaning.

In a first step, let's talk about html generation. I've found two small
classes about HTML generation : Gui.php and (with just one printCombo
method ) and class.select.php. However, we have 3 coding style in ppa
php pages : using Gui.php, using XHTMLElement or putting html directly
in the code.

I suggest to merge Gui.php and class.select.php in one single file and
drop html from the php files. This way, we will have some better code
guideline...

Obviously, this refactoring could be done after next release pg & ppa...

your thoughts ?

(yes, it's a long work...)
- --
+-----------------------+
| IoGuiX                |
| ioguix@...        |
| GPG/PGP ID : 0828C222 |
+-----------------------+
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFM5rcxWGfaAgowiIRAlSwAJ4mL98jxMf0UdmQC6nvztLuJ+WXCwCgo2Ew
rnIaFpkOZYgE5HP4GrX43Kw=
=8cFO
-----END PGP SIGNATURE-----

Re: [ppa-dev] Availability

[Christopher Kings-Lynne <chris.kingslynne@gm...>]

Thanks guys :)

On 10/15/06, Dan Wilson <killroyboy@...> wrote:
> Chris,
>
> It's sad to see you go.  I'm sure you will still lurk on the mailing
> list as I do.
>
> I would just like to publicly thank Chris for his dedication and the
> many, many hours of coding he has done over the years.  As the
> "grandfather" of this project, I have never felt at all like it was
> not in good hands while Chris was at the helm.
>
> And going forward, I feel completely comfortable with Rob as the lead.
>  I'm just glad the project didn't fall on it's face when I hit the
> point that Chris just hit (not that I ever thought it would in his care).
>
> Just a big thanks for all your hard work!
>
> -=Dan
>
> PS.  As a side note (to anyone), my company is hiring [1].  Anyone
> willing to move to Orem, UT and work at Omniture?  We really need some
> really sharp minds and we do most of our stuff in PHP.  If you haven't
> heard of us, we are one of (if not THE) leading web analytics (think
> webalizer/awstats on some serious steroids) companies in the world.
> We just went public in June. Big drawback is we use MySQL
> exclusively... not that I don't complain about that on a daily basis.
>  If you are interested, send me your resume.
>
> [1] http://www.omniture.com/career/148
>
>
> Christopher Kings-Lynne wrote:
> > Hi All,
> >
> > I've recently started my own company, and have (predictably) lost 100%
> > of my time to it, and everything else in my life :(
> >
> > I'm also now working in areas completely unrelated to PHP and PostgreSQL.
> >
> > So, I really don't know when I'll next be active.  I think
> > realistically I'll have to entrust the entirety of project management
> > to Rob Treat.
> >
> > I don't even find time to read the mailing lists these days, so if you
> > need to contact me, mail me personally at:
> >
> > chris.kingslynne@...
> >
> > Thanks guys.
> >
> > Chris
> >
> > ps. I will probably need to help out Rob with his first release :)
>


-- 
Chris Kings-Lynne
BSc (hons)

Director
KKL Pty. Ltd.
10 Orana Crescent
City Beach, Australia
6015

Biz: +61 8 6364 0484
Mob: +61 (0)409 294078

Re: [ppa-dev] Availability

[Dan Wilson <killroyboy@us...>]

Chris,

It's sad to see you go.  I'm sure you will still lurk on the mailing
list as I do.

I would just like to publicly thank Chris for his dedication and the
many, many hours of coding he has done over the years.  As the
"grandfather" of this project, I have never felt at all like it was
not in good hands while Chris was at the helm.

And going forward, I feel completely comfortable with Rob as the lead.
 I'm just glad the project didn't fall on it's face when I hit the
point that Chris just hit (not that I ever thought it would in his care).

Just a big thanks for all your hard work!

-=Dan

PS.  As a side note (to anyone), my company is hiring [1].  Anyone
willing to move to Orem, UT and work at Omniture?  We really need some
really sharp minds and we do most of our stuff in PHP.  If you haven't
heard of us, we are one of (if not THE) leading web analytics (think
webalizer/awstats on some serious steroids) companies in the world.
We just went public in June. Big drawback is we use MySQL
exclusively... not that I don't complain about that on a daily basis.
 If you are interested, send me your resume.

[1] http://www.omniture.com/career/148


Christopher Kings-Lynne wrote:
> Hi All,
> 
> I've recently started my own company, and have (predictably) lost 100%
> of my time to it, and everything else in my life :(
> 
> I'm also now working in areas completely unrelated to PHP and PostgreSQL.
> 
> So, I really don't know when I'll next be active.  I think
> realistically I'll have to entrust the entirety of project management
> to Rob Treat.
> 
> I don't even find time to read the mailing lists these days, so if you
> need to contact me, mail me personally at:
> 
> chris.kingslynne@...
> 
> Thanks guys.
> 
> Chris
> 
> ps. I will probably need to help out Rob with his first release :)

Re: [ppa-dev] Availability

[ioguix <ioguix@fr...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

It's sad you have to leave ppa, but anyway, this project is a very good
piece of php thanks to you and others.

Anyhow, I wish you to success in your new challenge !

++
- --
+-----------------------+
| IoGuiX                |
| ioguix@...        |
| GPG/PGP ID : 0828C222 |
+-----------------------+


Christopher Kings-Lynne wrote:
> Hi All,
> 
> I've recently started my own company, and have (predictably) lost 100%
> of my time to it, and everything else in my life :(
> 
> I'm also now working in areas completely unrelated to PHP and PostgreSQL.
> 
> So, I really don't know when I'll next be active.  I think
> realistically I'll have to entrust the entirety of project management
> to Rob Treat.
> 
> I don't even find time to read the mailing lists these days, so if you
> need to contact me, mail me personally at:
> 
> chris.kingslynne@...
> 
> Thanks guys.
> 
> Chris
> 
> ps. I will probably need to help out Rob with his first release :)
> 
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> phpPgAdmin-devel mailing list
> phpPgAdmin-devel@...
> https://lists.sourceforge.net/lists/listinfo/phppgadmin-devel
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFLzknxWGfaAgowiIRAq43AJ9v6qEpsPaj9DNWBjB8ZTJxlVG+uACfYESE
lRzvjpmQKLnZZiXljvkjbJo=
=dCH+
-----END PGP SIGNATURE-----

[ppa-dev] Availability

[Christopher Kings-Lynne <chris.kingslynne@gm...>]

Hi All,

I've recently started my own company, and have (predictably) lost 100%
of my time to it, and everything else in my life :(

I'm also now working in areas completely unrelated to PHP and PostgreSQL.

So, I really don't know when I'll next be active.  I think
realistically I'll have to entrust the entirety of project management
to Rob Treat.

I don't even find time to read the mailing lists these days, so if you
need to contact me, mail me personally at:

chris.kingslynne@...

Thanks guys.

Chris

ps. I will probably need to help out Rob with his first release :)

[ppa-dev] Explain and Explain Analyze buttons for user-submitted sql

[Karl O. Pinc <kop@me...>]

On 10/01/2006 07:07:17 PM, Robert Treat wrote:

> Karl's wompin patch:  I still haven't looked at this in depth, but it
> is now
> at the top of my list.  Some issues have been passed on to me that I
> want to
> look at (things like not showing the correct rows affected when doing
> an
> insert or delete, and that we should show the download button for
> queries
> that tossed errors and more) which I will post feedback to the list
> about.

Sounds like this could be a good time to discuss my comments
on the Explain and Explain Analyze buttons that I included with
my patch submission.  My proposal is that these buttons
be removed. I reproduce the comments here:

Explain & Explain analyze:

These buttons started out broken and I've not fixed them.
  - They only modify the first sql statement in the script.
  - They'll happily modify, and break, something that's not
    a SELECT.
  - They modify the the SQL, so if, after examining
    the results, you re-edit the SQL to see it (via supplied link)
    and then re-execute with explain again, then you get
    _another_ explain inserted before the first and the
    query is broken.
IMO they should be removed.  Anybody who knows enough to
interpret the EXPLAIN output
knows enough to type the extra words in front of their queries.
Making it work right, and work in a manner that's both
transparent to the user _and_ clear as to what's happening,
seems somewhat complicated and error prone.  I'm more
concerned about confusing the user's perceptions of what's
happening than I am of actually making the code do _something_.
In other words, the right way to do this is not clear.


Karl <kop@...>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein

Re: [ppa-dev] ppa 4.1 release plan

[ioguix <ioguix@fr...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

I finally closed this bug. You will find my patch in attachment to the
bug ticket.

However, I don't know if I should have close this ticket before asking
you, admins ?

By the way, since the tree is working with Safari I discovered a bug in
nodes' icons when expanding them. I will open a new ticket soon.

++
- --
+-----------------------+
| IoGuiX                |
| ioguix@...        |
| GPG/PGP ID : 0828C222 |
+-----------------------+


ioguix wrote:
> Hello,
> 
> I received your bug tracker about safari bugs. I hadn't time to work on
> it pasts days, but at least, I will try to work on it this week or next one.
> 
> Anyhow, I would have work (hope finish) on this before november ! BTW, I
> broke my linux partition on my ibook this w.end, so I will have to work
> under MacOSX...
> 
> ++
> --
> +-----------------------+
> | IoGuiX                |
> | ioguix@...        |
> | GPG/PGP ID : 0828C222 |
> +-----------------------+
> 
> 
> Robert Treat wrote:
>>> The following is my "hitlist" for getting to a 4.1 release.  
>>>
>>> Gavin's slony patch & The autovacuum editing screens are now down.  
>>>
>>> Chris role patch: still missing the critical file...  since Chris has not 
>>> turned up with it, I am inclined to revert us back to the user/group 
>>> screens...  I'd much prefer the file show up, or that someone step up and 
>>> offer to code a new interface (we still have all the back-end functionality 
>>> that we need afaik)
>>>
>>> John's create table w/ constraints patch:  i think we decided that the changes 
>>> needed to be wrapped in a transaction if it would be accepted, I'd like to 
>>> see that finished.  I'm not going to do it though, so unless someone steps up 
>>> to do it, this patch will get dropped. 
>>>
>>> Karl's wompin patch:  I still haven't looked at this in depth, but it is now 
>>> at the top of my list.  Some issues have been passed on to me that I want to 
>>> look at (things like not showing the correct rows affected when doing an 
>>> insert or delete, and that we should show the download button for queries 
>>> that tossed errors and more) which I will post feedback to the list about. 
>>> This is probably the last patch I plan to look at before beta, and honestly 
>>> if it gets too hairy I might drop this till 4.2. 
>>>
>>> Safari issues:  I passed this info (see bugtracker) on to ioguix, who I am 
>>> hoping is planning to look at them.  If anyone else wants to step up and 
>>> review / test/ fix that please do.
>>>
>>> 8.2 testing.  I did some basic testing against 8.2 and the most serious thing 
>>> seemed to be excessive warnings being generated for LIKE '\\' style queries. 
>>> I fixed the common ones, but there are more that should be fixed and tested. 
>>> I think Javier Carlos is looking into this, but since they are Warnings I'm 
>>> not going to wait for it to get done. 
>>>
>>> Other bugs from the bug tracker... well, we cant be any worse if we ship with 
>>> existing bugs, so I'm not stopping for those either. That said fixes are 
>>> welcome. :-)
>>>
>>> BTW, I'm currently without dedicated internet, so that is crimping my 
>>> development efforts.  8.2 will probably ship in November, I'd like to ship 
>>> 4.1 by then if we can.  
>>>

- -------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
phpPgAdmin-devel mailing list
phpPgAdmin-devel@...
https://lists.sourceforge.net/lists/listinfo/phppgadmin-devel
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFJM5bxWGfaAgowiIRAgLEAJ0TGUohGhaNNhjzD7wKQoW77DpAOQCfRnSM
x1XBS7SiyPYawg/SEmKYdvA=
=2azl
-----END PGP SIGNATURE-----

[ppa-dev] Bugs in Karl's whompin patch

[Karl O. Pinc <kop@me...>]

On 10/01/2006 07:07:17 PM, Robert Treat wrote:

> Karl's wompin patch:  I still haven't looked at this in depth, but it
> is now
> at the top of my list.  Some issues have been passed on to me that I
> want to
> look at (things like not showing the correct rows affected when doing
> an
> insert or delete, and that we should show the download button for
> queries
> that tossed errors and more) which I will post feedback to the list
> about.

I haven't looked at it in a while, but the problem with not
showing the correct number of rows affected when doing an
insert or delete, as well as the problems with the
"success-ness" of creating tables, is, I beleive, in the
first hunk of the patch to:
webdb/libraries/adodb/drivers/adodb-postgres64.inc.php

(It's a one-line hunk.  I haven't tried, could be that
it's not enough to just test for other pg_result_status()
values on that line.  But for some reason I've the feeling
at the moment that that would do it.)

There are comments in the Bugs section of Script.php
the class SqlQuery->ExecuteQuery() function that basically
say the same thing -- that adodb's Execute() function is
designed to execute selects, not inserts, updates, deletes,
table creation, and so forth.

If I recall correctly,
adodb does not recognize the result when you do inserts
or deletes or create tables with Execute().  I patched it so that it
recognizes the COPY FROM results, because that was the
point of the old "execute sql uploaded in a file"
code but pretty much otherwise did not want to overly
mess with adodo.  Really, I think I only changed maybe
3 or 4 lines in adodb-postgres64.inc.php, but the
patch is bigger than that because I added an "if"
and that made some existing code indent.  (I did try
to make sure that I did not break any of the existing
functionality, but merging the COPY FROM file upload
code into the rest of the generic user-supplied SQL
handling ment going deep into the guts and I guess I
was just too timid in not making things generic enough.
I didn't actually discover a problem until the
patch had been sitting on the list for a while.
At that point I just added comments documenting the
problem and figured there'd be discussion, or everybody'd
be happy and want to ignore the problem for now
in the interest of moving forward. :-)

So, the direction I'd take is to have adodb recoginze
most postgresql result codes in that little section,
but I don't really want to go making more adodb changes
without talking to the list first.

I could enumerate all the possible directions I can
think of to resolve the issue, but don't want to
spend the energy if everybody's happy with the obivious
answer.

This would be the tricky bug.  Everything else, like
not having a 'download' button when there's errors,
should be straightforward coding.

Regards,

Karl <kop@...>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein

P.S.
Robert,
Sorry to hear about your Internet connectivity woes.  :-P

Re: [ppa-dev] User-supplied SQL rework patch -- Was: Re: When to start showing patches

[Karl O. Pinc <kop@me...>]

Updated to apply against the latest CVS.
No other changes.

Karl <kop@...>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein

---------------------<snip>-----------------
Patching instructions:

Make a new directory.
cd into it.
Use cvs to checkout the 2 PPA directories:
   webdb
   www-site
(cvs  -z3 =20
-d:pserver:anonymous@...:/cvsroot/phppgadmin =20
checkout -P webdb www-site)

gunzip -c usersql.patch.gz | patch -p1

Save icon:
webdb/images/themes/default/History.png

Read webdb/INSTALL and update webdb/conf/config.inc.php
to webdb/conf/config.inc.php-dist.

Changes since last patch:
   Applied against CVS as of Oct 3.

Changes since last patch:
   Additional ADODB related comments added to Script.php in
   Query->ExecuteQuery() and Script->ExecuteQuery bugs sections.

Changes since last patch:
   Applied against CVS as of Aug 22.  Should apply anyway but...

Changes since last patch:
   Applied against Aug 13 cvs updates.  Just some line number changes in
   english.php that probably wouldn't bother patch anyway.

Changes since last patch:
   Bumped config.inc.php-dist serial number.
   Applied against most recent cvs -- constraints at table creation =20
time.
   Patch now includes webdb/lang/recoded/ so no need to run make in
     webdb/lang/.

Changes since last patch:
   Futher instructions in dot_htaccess-dist explaining how to get
     apache configured.  Details for various distros.
   When there's no history, make the "no history" a
     "message" css class.
   Get rid of fancy alignment of textual column headings for floating
     point numbers and other datatypes containing decimal points.

Changes since last patch:
   Show php module settings (multibyte stuff was missing.)

Changes since last patch:
   Moved .htaccess file to dot_htaccess-dist.  Updated INSTALL
     and FAQ.
   Added php_settings.php page, and 'show_php_settings' config
     parameter to aid in PHP configuration of multibyte character
     encoding or other PHP configuration.  Update INSTALL and
     FAQ.
   Add to FAQ should there be problems with output buffering.

Changes since last patch:
   Added Apache custom error page to web site and direct traffic
     there when new security restrictions are triggered.
   Adjust FAQ and INSTALL document for error redirection.

Changes since last patch:
   Changed config.inc.php-dist not config.inc.php.
   Added comments to .htaccess file.
   Added instructions to INSTALL file.
   Added to website FAQ.

Changes since the last patch:
   History "tab" now hilights when selected.
   File upload undefined function when PHP has mb support is fixed.
   Explain buttons now properly do nothing when the old
     pagination routine is called.
   Tabify and correct indent levels throughout code.
   Misc code cleanups.

Changes:
Enhanced error reporting for larger scripts.
Textual output (<pre></pre>) support:
   Supports data
     containing end-of-line characters imbedded in the text.
   Speeds large query result display in browser.
   Available with | char as column separators and without.
Multibyte character support (in user-supplied SQL and
   in the column widths of textual output -- SQL file upload
   did not have multibyte support.)
Unification of code the executes SQL uploaded in a file
and code that is supplied in a textbox:
    Textbox code now displays output from all row-returning
      sql statements. (E.g:
      (begin; select * from foo; delete foo where id =3D1;
       select * from foo; rollback;)
      shows the output from both selects.)
    Textbox code now does COPY foo FROM STDIN; so it
     will run pasted pg_dump output.
Re-working of interface to ADODB to support COPY FROM command;
    all code works (as much as possible) though ADODB.
Elimination of callback for SQL file upload.
Download from user-supplied SQL without viewing results.
User-supplied per-database SQL command history.  Both in
   a popup window and in the database frame.

Note that the multibyte character support is going to have
to be tested by somebody who's familiar with one of those
character sets, preferably something complicated like
the various Japanese encodings.  It requres php
multibyte support be ./configure-d on and the default server
encoding properly set and a browser that properly reports
encodings to the server.  It is virtually untested and
I don't know how to test it because I've no clue how to
even enter multibyte characters on the keyboard much less
know whether they come out the right width when they're
displayed back to me.

What must be tested to be sure that
backward compatibilty does not break is multibyte chars
entered into the SQL textbox.  The new code parses this
whereas the old code passed it straight to the server.
(The old SQL upload code parsed it, and did
not parse multibyte encoded SQL correctly.)  The
parser is virtually unchanged, but uses php multibyte
support when available.  Because the original parser code
is based on the psql code which _is_ multibyte
capable, it _should_ work.  But....


Bugs:

Known bugs are noted in comments heading files and
functions.

http://validator.w3.org fails the code containing
javascript.  (Output of database.php, sqledit.php,
and history.php.)  But, it did this before I started so
I figure at least it's consistent.  I don't know
if this is really a problem with the code or just
a problem with the w3 sgml parser vis javascript
or maybe because I'm using Firefox to save the
frame before validating the code and it's interpreting
the javascript when it saves it.  Firefox bug?

The themes/default/global.css file does not validate
at http://jigsaw.w3.org/css-validator/.  Again,
it didn't before this patch and I didn't add any
new errors.


Notes on specific files:

images/themes/default/History.png

This is from http://tango.freedesktop.org/Tango_Icon_Library
tango-icon-theme-0.7.2/16x16/apps/accessories-text-editor.png
Distributed under:
Creative Commons Attribution-ShareAlike 2.5 License Agreement
AUTHORS content (for attribution) is:
Steven Garrity <sgarrity@...>
Lapo Calamandrei <calamandrei@...>
Ryan Collier <rcollier@...>
Rodney Dawes <dobey@...>
Andreas Nilsson <nisses.mail@...>
Tuomas Kuosmanen <tigert@...>
Garrett LeSage <garrett@...>
Jakub Steiner <jimmac@...>


./.htaccess

Heads-up -- hidden file.  Alternative is an Apache 2
/etc/httpd/conf.d/phpPgAdmin.conf file, but that's
best done with a Makefile that installs.

I once sent a patch for the README in regards the security issues
of sending usernames and passwords in the clear
over the network and got a reply that it was accecpted, but it's not
there.  Anyhow, I've included security by default
but have not messed with the README where there should
be a note.  A spiffier rewrite rule that automatically
converted the current URL to https might be in the
original patch.

Output buffering must be on for full pagination support
for user-supplied SQL results.


classes/Misc.php

The first changes here should really be submitted as a
separate patch.  It includes a full set of
entity references -- required in the case when the
client does not have them built-in, especially when
there is an alternate character encoding on the
client.  I could be wrong here, but in any case
I can't see how it hurts to include entity definitions.
I do use entity references in my main patch so...



libraries/adodb/drivers/adodb-postgres64.inc.php

  Enhanced to support COPY foo FROM STDIN


classes/database/Connection.php

  Querying at random times is not so good.  It breaks COPY FROM
and might do other odd things in the middle of a complex user
submitted SQL script.  Yes, this makes things slower, but
not really with newer postgres and php implimentations.


classes/database/Postgres.php

  Parser code moved into Script.php.


Other comments to do with the unpatched build:

Explain & Explain analyze:

These buttons started out broken and I've not fixed them.
  - They only modify the first sql statement in the script.
  - They'll happily modify, and break, something that's not
    a SELECT.
  - They modify the the SQL, so if, after examining
    the results, you re-edit the SQL to see it (via supplied link)
    and then re-execute with explain again, then you get
    _another_ explain inserted before the first and the
    query is broken.
IMO they should be removed.  Anybody who knows enough to
interpret the EXPLAIN output
knows enough to type the extra words in front of their queries.
Making it work right, and work in a manner that's both
transparent to the user _and_ clear as to what's happening,
seems somewhat complicated and error prone.  I'm more
concerned about confusing the user's perceptions of what's
happening than I am of actually making the code do _something_.
In other words, the right way to do this is not clear.

Never ending webpage downloads:

Something in the user-submitted query processing is broken,
maybe.  Firefox (Win, Mac, and Unix) and seamonkey
never seems to see the
connection close after obtaining results from user-submitted
SQL.  That is, the little "processing" icon
is always rolling (or whatever) and the little progress
bar reaches the right side but never goes away, etc.
There's nothing going on on the server, no CPU, no disk,
no network.  And both the server and the client seem
to have no network socket open.

My theory is that there's some continuously running javascript
that happens when the left hand frame is reloaded.   ??
(See:
http://www.velocityreviews.com/forums/t161353-javascript-for-dreamweaver-=
mx-rollover-buttons-and-firefox.html
http://www.thescripts.com/forum/thread163969.html)

Safari broken:

Tried phppgadmin with Safari on Tiger and the left hand frame is
broken.  Clicking on it to login gets the message (from memory, could
be wrong) "no server".  After that trying to do things like expand the
tree just don't work.  Perhaps this relates to the firefox "forever
download" problem.

Re: [ppa-dev] ppa 4.1 release plan

[ioguix <ioguix@fr...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

I received your bug tracker about safari bugs. I hadn't time to work on
it pasts days, but at least, I will try to work on it this week or next one.

Anyhow, I would have work (hope finish) on this before november ! BTW, I
broke my linux partition on my ibook this w.end, so I will have to work
under MacOSX...

++
- --
+-----------------------+
| IoGuiX                |
| ioguix@...        |
| GPG/PGP ID : 0828C222 |
+-----------------------+


Robert Treat wrote:
> The following is my "hitlist" for getting to a 4.1 release.  
> 
> Gavin's slony patch & The autovacuum editing screens are now down.  
> 
> Chris role patch: still missing the critical file...  since Chris has not 
> turned up with it, I am inclined to revert us back to the user/group 
> screens...  I'd much prefer the file show up, or that someone step up and 
> offer to code a new interface (we still have all the back-end functionality 
> that we need afaik)
> 
> John's create table w/ constraints patch:  i think we decided that the changes 
> needed to be wrapped in a transaction if it would be accepted, I'd like to 
> see that finished.  I'm not going to do it though, so unless someone steps up 
> to do it, this patch will get dropped. 
> 
> Karl's wompin patch:  I still haven't looked at this in depth, but it is now 
> at the top of my list.  Some issues have been passed on to me that I want to 
> look at (things like not showing the correct rows affected when doing an 
> insert or delete, and that we should show the download button for queries 
> that tossed errors and more) which I will post feedback to the list about. 
> This is probably the last patch I plan to look at before beta, and honestly 
> if it gets too hairy I might drop this till 4.2. 
> 
> Safari issues:  I passed this info (see bugtracker) on to ioguix, who I am 
> hoping is planning to look at them.  If anyone else wants to step up and 
> review / test/ fix that please do.
> 
> 8.2 testing.  I did some basic testing against 8.2 and the most serious thing 
> seemed to be excessive warnings being generated for LIKE '\\' style queries. 
> I fixed the common ones, but there are more that should be fixed and tested. 
> I think Javier Carlos is looking into this, but since they are Warnings I'm 
> not going to wait for it to get done. 
> 
> Other bugs from the bug tracker... well, we cant be any worse if we ship with 
> existing bugs, so I'm not stopping for those either. That said fixes are 
> welcome. :-)
> 
> BTW, I'm currently without dedicated internet, so that is crimping my 
> development efforts.  8.2 will probably ship in November, I'd like to ship 
> 4.1 by then if we can.  
> 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFIL/NxWGfaAgowiIRAk81AJ4z5CE2T3BhACFt6HPIN+RNW0KcCQCggX9S
UR5jZhjNm/CmjXnzxXROlUw=
=E89O
-----END PGP SIGNATURE-----

[ppa-dev] ppa 4.1 release plan

[Robert Treat <xzilla@us...>]

The following is my "hitlist" for getting to a 4.1 release.  

Gavin's slony patch & The autovacuum editing screens are now down.  

Chris role patch: still missing the critical file...  since Chris has not 
turned up with it, I am inclined to revert us back to the user/group 
screens...  I'd much prefer the file show up, or that someone step up and 
offer to code a new interface (we still have all the back-end functionality 
that we need afaik)

John's create table w/ constraints patch:  i think we decided that the changes 
needed to be wrapped in a transaction if it would be accepted, I'd like to 
see that finished.  I'm not going to do it though, so unless someone steps up 
to do it, this patch will get dropped. 

Karl's wompin patch:  I still haven't looked at this in depth, but it is now 
at the top of my list.  Some issues have been passed on to me that I want to 
look at (things like not showing the correct rows affected when doing an 
insert or delete, and that we should show the download button for queries 
that tossed errors and more) which I will post feedback to the list about. 
This is probably the last patch I plan to look at before beta, and honestly 
if it gets too hairy I might drop this till 4.2. 

Safari issues:  I passed this info (see bugtracker) on to ioguix, who I am 
hoping is planning to look at them.  If anyone else wants to step up and 
review / test/ fix that please do.

8.2 testing.  I did some basic testing against 8.2 and the most serious thing 
seemed to be excessive warnings being generated for LIKE '\\' style queries. 
I fixed the common ones, but there are more that should be fixed and tested. 
I think Javier Carlos is looking into this, but since they are Warnings I'm 
not going to wait for it to get done. 

Other bugs from the bug tracker... well, we cant be any worse if we ship with 
existing bugs, so I'm not stopping for those either. That said fixes are 
welcome. :-)

BTW, I'm currently without dedicated internet, so that is crimping my 
development efforts.  8.2 will probably ship in November, I'd like to ship 
4.1 by then if we can.  

-- 
Robert Treat
Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL