phppgadmin-devel — For Developers Only

Re: [ppa-dev] Patches to roll to 9.3

[Karl O. P. <ko...@me...>]

Hi,

Since the 9.3 roll bits are in head the classifyhelp branch
no longer applies cleanly to head.  I've left
the classifyhelp, rollto93, and revhelp branches
alone, but rebased the classifyhelp branch onto
master and called this new branch "help".  This
new branch now contains the uncommitted portions
of the help system cleanup.

https://github.com/kpinc/phppgadmin/commits/help

Since the "import" branch was based off the rollto93
branch I've made a copy "import93" and rebased it
onto head.

And, since I'm rebasing, I rebased the sigenhance
branch into a sigenhance2 branch to make the patches
slightly prettier.  (And I fixed a minor bug.)

This seems a good time to summarize my branches:

  gitignore       Make ppa dev environment emacs friendly
  help            Fix help system
  import93        Import into views + replace on upload
  jquery18        Remove (some?) jQuery 1.8 dependencies
  multiport       Mutiple PPA instances on different webserver ports
  selenium-doc    Improve selenium docs (includes selenium-reset docs)
  selenium-reset  Menu item to reset selenium testdb state
  selenium-login  Delete cookies each selenium login
  sigenhance2     Control over what PG processes display  (Note:
                  branches off of 'jquery18'.)
  uselessplugin   A sample useless plugin that runs js

These should all apply on top of head and on top of each other.

Outstanding issues (I know of)

  selenium-reset  Includes il8n of selenium links.  Do we want il8n?
  selenium-login  ioguix questions clearing cookies each login.
  uselessplugin   Do we want a PPA examples folder?  We won't maintain
                  or endorse the code but someone may find it useful.
                  Skatox's CRUD generator probably falls in this
                  category.

"Old" branches:
  classifyhelp    Original "help" branch tip
  import          Original "import93" branch
  revhelp         Beginning of "help" branch
  rollto93        Some in HEAD, some in middle of "help"
  selenium-enabled  Original selenium-reset + remove links patch
  sigenhance      Orignial 'sigenhance2', w/o rebasing & w/o 1 fix.


On 05/18/2013 09:35:05 PM, Karl O. Pinc wrote:
> On 05/18/2013 06:02:50 PM, Robert Treat wrote:
> > The 9.3 bits are in official head now (which means I think you can
> > delete this branch now if you want to). 
> 
> Not yet, the first patch in the branch creates PostgresDoc93.php.
> 
> commit e4a39a2535976ae7e9073faaf52275737f8286b0
> Author: Karl O. Pinc <ko...@me...>
> Date:   Tue May 7 23:09:55 2013 -0500
> 
>     Roll the help system from 9.2 to 9.3.
> 
> 
> > I wanted to do this
> > independent of the help changes.
> 
> You must not have applied the above patch since it
> affects the help system, but that's what leaves
> HEAD without a 9.3 help system....



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

[ppa-dev] CRUD generator plugin for PPA stable

[Miguel U. <mig...@sk...>]

Hello guys,

Some years ago I did a CRUD generator for PPA, after long time and 3 
re-wrotes (due to PHP 5.3, Leonardo's plugin's architecture and official 
changes) I've released the plugin for the stable version of PPA.

You can check it at: https://github.com/Skatox/crudgen and you watch a 
video of how it works at: https://www.youtube.com/watch?v=ZjMyptlcYg4

Hope you check it and i wonder if it's possible to have it in the 
official software of ppa.

Have a nice day.

-- 
Miguel Ángel Useche Castro
mig...@sk...
@skatox
Visita http://www.skatox.com/  Blog de Informática, Linux, Desarrollo Web, Software Libre, Videojuegos y mas....

[ppa-dev] The future of the xtree code in ppa

[Karl O. P. <ko...@me...>]

Hi,

I've been looking at the xtree "database browser" code
and have begun some extensions of it to support the
ability to see views mixed in with tables, or not.

The last message from ioguix disturbed me and got me
thinking.  I don't want to work against the project's
direction and I've vague memories of discussions
regarding javascript and the db browser tree's future.

I want to discuss whether there is any thought that the xtree
code might be replaced before I put in more work.  I've looked
back through the mailing list and see no such discussion,
but for some reason I think this may have been
discussed on IRC, perhaps with a thought to using
a jQuery plugin like jstree -- in the context of
needing a newer jQuery.  (jstree requires
jQuery 1.9, I think.)  Since I've patches to remove
some dependencies on jQuery 1.8 this might be in
the offing.

I lean toward the "if it ain't broke don't fix it" side
of things, especially since it would be non-trivial
and I think we'd require extensions beyond a basic
"tree browser" to support the functionality we
currently have.

Please give me some guidance here.  I don't want
to put work into xtree that won't get committed
because xtree isn't wanted.  (You may not like
the code for other reasons, but that's another
story.)

Regards,

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

Re: [ppa-dev] Enhancments to the process page

[Karl O. P. <ko...@me...>]

On 05/30/2013 09:19:01 AM, Karl O. Pinc wrote:
> > On 05/30/2013 03:18:46 AM, Karl O. Pinc wrote:
> 
> > > I've a spiffy new branch on github: sigenhance

I've pushed 3 more patches to this branch.  2 are cosmetic-ish,
but the third merits consideration.

Fixed an image placement bug due to images not being fully loaded.

To solve this I added the MIT licensed imagesloaded library
into PPA.  This seems a bit much to solve this one problem
but it seems like it will be useful if we're to do
any amount of js that frobs the user interface.

I hope I really am done on this branch now, excepting any
sort of rebasing you want done to reorder the patches
to make history prettier since I first announced
the jquery18 and sigenhance branches.

> > > 
> > > https://github.com/kpinc/phppgadmin/commits/sigenhance
> > > 
> > > 
> > > This branches off the tip of jquery18.
> > > 
> > > This branch adds 2 improvements to the
> > > process page:
> > > 
> > > - A checkbox that controls display of the
> > > PG process PPA uses to find processes.
> 
> > > - A checkbox (for 9.2+) that controls display
> > > of inactive connections.  

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

Re: [ppa-dev] Enhancments to the process page

[Karl O. P. <ko...@me...>]

Hi,

I've pushed 4 more patches to this branch.

I should be done now.

On 05/30/2013 04:01:48 AM, Karl O. Pinc wrote:
> Hi,
> 
> I've done a bit more testing.  This branch
> still suffers from (I think) the same error
> trapping issues as the upstream code.

I've looked at this and I introduced the
problem with error reporting when the webserver
goes away, so I've added another patch to the
branch to fix this.  (Should probably
be rebased back to 6c911a36 where the bug was 
introduced.  This note is in the commit message.)

> When the webserver or PG goes away
> the failure mode remains lame.  Oh well.
> I didn't pay attention to this aspect of
> the code and I need to move on.

Things remain lame when PG goes away.

The other new patches:

Clean up old jQuery code.  (Might be
rebased back into the jquery18 branch patch
e92014774 that rewrote the code.  This is
not in the commit message.)

Improve the error text when the webserver disappears.

Better positions the animated loading graphic.
(Might also be rebased back to e92014774 where
the jquery code was rewritten.  This is not
in the commit message.)

> 
> On 05/30/2013 03:18:46 AM, Karl O. Pinc wrote:

> > I've a spiffy new branch on github: sigenhance
> > 
> > https://github.com/kpinc/phppgadmin/commits/sigenhance
> > 
> > 
> > This branches off the tip of jquery18.
> > 
> > This branch adds 2 improvements to the
> > process page:
> > 
> > - A checkbox that controls display of the
> > PG process PPA uses to find processes.

> > - A checkbox (for 9.2+) that controls display
> > of inactive connections.  


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

Re: [ppa-dev] Enhancments to the process page

[Karl O. P. <ko...@me...>]

Hi,

I've done a bit more testing.  This branch
still suffers from (I think) the same error
trapping issues as the upstream code.
When the webserver or PG goes away
the failure mode remains lame.  Oh well.
I didn't pay attention to this aspect of
the code and I need to move on.

On 05/30/2013 03:18:46 AM, Karl O. Pinc wrote:
> Hi,
> 
> I've a spiffy new branch on github: sigenhance
> 
> https://github.com/kpinc/phppgadmin/commits/sigenhance
> 
> 
> This branches off the tip of jquery18.
> 
> This branch adds 2 improvements to the
> process page:
> 
> - A checkbox that controls display of the
> PG process PPA uses to find processes.
> 
> By default, do not display this process.
> 
> - A checkbox (for 9.2+) that controls display
> of inactive connections.  To look at it
> another way, that suppresses display
> of entirely idle connections.  (Those
> connections that are idle but inside
> a transaction display regardless.)
> 
> By default, do not display entirely
> idle connections.
> 
> - Since there are checkboxes, the non-js
> interface changes the refresh control
> from a link to a button.
> 
> 
> I didn't test on IE.  I think it works,
> I'm doing what PPA does elsewhere, but
> I've forgotten all the stupid I.E.
> required workarounds. 
> 
> 
> Now that the process page is so spiffy
> the next step is to add it as a tab on the
> server page so that all the processes
> on the server can be seen at once.
> But that's not on my immediate list
> of things to do.
> 
> Regards,
> 
> Karl <ko...@me...>
> Free Software:  "You don't pay back, you pay forward."
>                  -- Robert A. Heinlein
> 
> 
> ------------------------------------------------------------------------------
> Introducing AppDynamics Lite, a free troubleshooting tool for
> Java/.NET
> Get 100% visibility into your production application - at no cost.
> Code-level diagnostics for performance bottlenecks with <2% overhead
> Download for free and get started troubleshooting in minutes.
> http://p.sf.net/sfu/appdyn_d2d_ap1
> _______________________________________________
> phpPgAdmin-devel mailing list
> php...@li...
> https://lists.sourceforge.net/lists/listinfo/phppgadmin-devel
> 




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

[ppa-dev] Enhancments to the process page

[Karl O. P. <ko...@me...>]

Hi,

I've a spiffy new branch on github: sigenhance

https://github.com/kpinc/phppgadmin/commits/sigenhance


This branches off the tip of jquery18.

This branch adds 2 improvements to the
process page:

- A checkbox that controls display of the
PG process PPA uses to find processes.

By default, do not display this process.

- A checkbox (for 9.2+) that controls display
of inactive connections.  To look at it
another way, that suppresses display
of entirely idle connections.  (Those
connections that are idle but inside
a transaction display regardless.)

By default, do not display entirely
idle connections.

- Since there are checkboxes, the non-js
interface changes the refresh control
from a link to a button.


I didn't test on IE.  I think it works,
I'm doing what PPA does elsewhere, but
I've forgotten all the stupid I.E.
required workarounds. 


Now that the process page is so spiffy
the next step is to add it as a tab on the
server page so that all the processes
on the server can be seen at once.
But that's not on my immediate list
of things to do.

Regards,

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

Re: [ppa-dev] Fixes to js/database.js

[Karl O. P. <ko...@me...>]

Hi,

Just added a 5th patch to the branch to set the
correct cursor graphic.  (Should really probably
be patch 3, but I'm not rebaseing and rewriting
history.)

When examining this patch and caa5cff9 (patch 2)
it's worth thinking about whether js created
hot-spots should have their own css selector,
as with foreign keys.  I've not thought about it
much and it's not obvious to me that it's useful,
nor obvious that it's useless, so I opted for
simplicity.

See also:
https://developer.mozilla.org/en-US/docs/Web/CSS/
cursor#Browser_compatibility

On 05/29/2013 04:41:57 PM, Karl O. Pinc wrote:
> Hi,
> 
> I've a new branch "jquery18":
> 
> https://github.com/kpinc/phppgadmin/commits/jquery18
> 
> Uh, I branched this off HEAD, not 5.1.  There are
> fixes, there are enhancements....
> 
> I did not test with multiple browsers.  It's all jQuery
> so it's supposed to work but it wouldn't hurt to
> get more testing before committing.  Tested with FF 3.5.
> 
> 
> There are 4 patches:
> 
> 1) Remove jQuery 1.8 .toggle() dependency from js/database.js
> 
> This turned out to be really annoying.  See the commit
> message.  
> 
> I strongly suspect that this fixes a bug
> that causes js to constantly run, forever, even
> when the refresh seems to be stopped.  Often
> this seems to manifest itself in my browser as
> an alert in the morning that says a script is not
> responding.  On the other hand, maybe some other
> random web page is doing that to me.
> 
> This also re-orders when display is done and
> when the ajax requests are started with an
> eye toward making things less racy.  It probably
> does not matter, but if you're interested
> a 2nd set of eyes on this wouldn't hurt.
> 
> 
> 2) Add style to the links made active by js.
> 
> 
> 3) Have refresh happen instantly when "start"
> is clicked.  
> 
> The window.setInterval() docs
> seem to deliberately leave unspecified
> whether the function is called instantly
> or the first call happens after a delay.
> Regardless, this does no harm.  Right?
> It makes my browser more snappy.
> 
> 
> 4) Fix bug where clicking on "stop" would
> display an error message.
> 
> This is not a particularly elegant solution,
> but it works.
> 
> Regards,
> 
> Karl <ko...@me...>
> Free Software:  "You don't pay back, you pay forward."
>                  -- Robert A. Heinlein
> 
> 
> ------------------------------------------------------------------------------
> Introducing AppDynamics Lite, a free troubleshooting tool for
> Java/.NET
> Get 100% visibility into your production application - at no cost.
> Code-level diagnostics for performance bottlenecks with <2% overhead
> Download for free and get started troubleshooting in minutes.
> http://p.sf.net/sfu/appdyn_d2d_ap1
> _______________________________________________
> phpPgAdmin-devel mailing list
> php...@li...
> https://lists.sourceforge.net/lists/listinfo/phppgadmin-devel
> 
> 




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

[ppa-dev] Fixes to js/database.js

[Karl O. P. <ko...@me...>]

Hi,

I've a new branch "jquery18":

https://github.com/kpinc/phppgadmin/commits/jquery18

Uh, I branched this off HEAD, not 5.1.  There are
fixes, there are enhancements....

I did not test with multiple browsers.  It's all jQuery
so it's supposed to work but it wouldn't hurt to
get more testing before committing.  Tested with FF 3.5.


There are 4 patches:

1) Remove jQuery 1.8 .toggle() dependency from js/database.js

This turned out to be really annoying.  See the commit
message.  

I strongly suspect that this fixes a bug
that causes js to constantly run, forever, even
when the refresh seems to be stopped.  Often
this seems to manifest itself in my browser as
an alert in the morning that says a script is not
responding.  On the other hand, maybe some other
random web page is doing that to me.

This also re-orders when display is done and
when the ajax requests are started with an
eye toward making things less racy.  It probably
does not matter, but if you're interested
a 2nd set of eyes on this wouldn't hurt.


2) Add style to the links made active by js.


3) Have refresh happen instantly when "start"
is clicked.  

The window.setInterval() docs
seem to deliberately leave unspecified
whether the function is called instantly
or the first call happens after a delay.
Regardless, this does no harm.  Right?
It makes my browser more snappy.


4) Fix bug where clicking on "stop" would
display an error message.

This is not a particularly elegant solution,
but it works.

Regards,

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

[ppa-dev] Fix to reporting process waiting status

[Karl O. P. <ko...@me...>]

Hi,

I've a new branch on github "procs91":

https://github.com/kpinc/phppgadmin/commits/procs91

Have PG > 8.1 && <= 9.1 reporting waiting processes on process page.

Robert's "waiting" enhancement patch forgot to include the
Postgres91.php file.

Tested with 8.4.

Regards,

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

Re: [ppa-dev] Fix so signals work

[Karl O. P. <ko...@me...>]

On 05/24/2013 03:56:16 PM, Robert Treat wrote:
> applied and backpatched to REL_5-1. As a side note, if you are fixing
> a bug that is in REL_5-1, I think it'd be better to make the fix 
> based
> on the REL_5-1 branch, and  we can pick from there back into master.

Will do.


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

Re: [ppa-dev] [ppa-commit] [phppgadmin/phppgadmin] 974abe: Make process screen report information in a way th...

[Karl O. P. <ko...@me...>]

Interesting.  I was just working toward something similar.

I've not tested with 9.2 but with 9.3 the process list
reports with a state of idle the last process executed.
I figure it's important to frob the output so that
it's clear what's running and what's not but had
not yet come to any conclusions.

On 05/24/2013 05:04:42 PM, GitHub wrote:
>   Branch: refs/heads/master
>   Home:   https://github.com/phppgadmin/phppgadmin
>   Commit: 974abee83d88534339c2aa77137c2a89993dc06e
>       https://github.com/phppgadmin/phppgadmin/
> commit/974abee83d88534339c2aa77137c2a89993dc06e
>   Author: Robert Treat <ro...@xz...>
>   Date:   2013-05-24 (Fri, 24 May 2013)
> 
>   Changed paths:
>     M classes/database/Postgres.php
>     M classes/database/Postgres81.php
>     M database.php
>     M lang/english.php
> 
>   Log Message:
>   -----------
>   Make process screen report information in a way that makes sense in
> postgres 9.2
> Also add in the waiting option, making it work back to at least 8.1


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

Re: [ppa-dev] Add support for user-sent signals

[Robert T. <ro...@xz...>]

applied.

Robert Treat

On Fri, May 24, 2013 at 1:11 AM, Karl O. Pinc <ko...@me...> wrote:
> Hi,
>
> >From 9.2 up PG now supports users sending cancel and
> terminate signals to processes run by their roles.
>
> I've a new branch "sigprocs" which adds support
> for this functionality:
>
> https://github.com/kpinc/phppgadmin/commits/sigprocs
>
> 1 simple patch.
>
> Regards,
>
> Karl <ko...@me...>
> Free Software:  "You don't pay back, you pay forward."
>                  -- Robert A. Heinlein
>
>
> ------------------------------------------------------------------------------
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_may
> _______________________________________________
> phpPgAdmin-devel mailing list
> php...@li...
> https://lists.sourceforge.net/lists/listinfo/phppgadmin-devel

Re: [ppa-dev] Fix so signals work

[Robert T. <ro...@xz...>]

applied and backpatched to REL_5-1. As a side note, if you are fixing
a bug that is in REL_5-1, I think it'd be better to make the fix based
on the REL_5-1 branch, and  we can pick from there back into master.

Robert Treat

On Fri, May 24, 2013 at 1:18 AM, Karl O. Pinc <ko...@me...> wrote:
> Hi,
>
> This branch "sigfix" fixes the signal system.
> Signals were being sent to the wrong pid.
>
> https://github.com/kpinc/phppgadmin/commits/sigfix
>
> Karl <ko...@me...>
> Free Software:  "You don't pay back, you pay forward."
>                  -- Robert A. Heinlein
>
>
> ------------------------------------------------------------------------------
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_may
> _______________________________________________
> phpPgAdmin-devel mailing list
> php...@li...
> https://lists.sourceforge.net/lists/listinfo/phppgadmin-devel

Re: [ppa-dev] Fix bugs in <= 9.1 processes page

[Robert T. <ro...@xz...>]

Applied and backpatched to REL_5-1.


Robert Treat

On Fri, May 24, 2013 at 12:50 PM, Karl O. Pinc <ko...@me...> wrote:
> Hi,
>
> I've a new branch "sigtidy" that fixes bugs in the
> <= 9.1 process page.  The database string is cleaned
> and processes are restricted to the given db (when
> one is specified in the call.)
>
> https://github.com/kpinc/phppgadmin/commits/sigtidy
>
> I looked back at the history and can't tell why the
> bugs crept in.
>
> I've tested with 8.4 and verified (on IRC with RhodiumToad)
> that using datname for the db name column is correct back to 7.4.
>
> 1 patch:
> ------
> Processes: Fix <= 9.1 to show only db related processes & clean dbname.
>
>     Fix bugs introduced in commit d571ecae7b by ioguix on 2012-09-21.
>
> Karl <ko...@me...>
> Free Software:  "You don't pay back, you pay forward."
>                  -- Robert A. Heinlein
>
>
> ------------------------------------------------------------------------------
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_may
> _______________________________________________
> phpPgAdmin-devel mailing list
> php...@li...
> https://lists.sourceforge.net/lists/listinfo/phppgadmin-devel

[ppa-dev] Fix bugs in <= 9.1 processes page

[Karl O. P. <ko...@me...>]

Hi,

I've a new branch "sigtidy" that fixes bugs in the
<= 9.1 process page.  The database string is cleaned
and processes are restricted to the given db (when
one is specified in the call.)

https://github.com/kpinc/phppgadmin/commits/sigtidy

I looked back at the history and can't tell why the
bugs crept in.

I've tested with 8.4 and verified (on IRC with RhodiumToad)
that using datname for the db name column is correct back to 7.4.

1 patch:
------
Processes: Fix <= 9.1 to show only db related processes & clean dbname.
    
    Fix bugs introduced in commit d571ecae7b by ioguix on 2012-09-21.

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

[ppa-dev] Fix so signals work

[Karl O. P. <ko...@me...>]

Hi,

This branch "sigfix" fixes the signal system.
Signals were being sent to the wrong pid.

https://github.com/kpinc/phppgadmin/commits/sigfix

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

[ppa-dev] Add support for user-sent signals

[Karl O. P. <ko...@me...>]

Hi,

From 9.2 up PG now supports users sending cancel and
terminate signals to processes run by their roles.

I've a new branch "sigprocs" which adds support
for this functionality:

https://github.com/kpinc/phppgadmin/commits/sigprocs

1 simple patch.

Regards,

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

Re: [ppa-dev] Fix of race condition in db browser frame

[Karl O. P. <ko...@me...>]

On 05/20/2013 08:53:59 PM, Karl O. Pinc wrote:
> On 05/20/2013 03:25:13 PM, Jehan-Guillaume (ioguix) de Rorthais 
> wrote:
> > On 15/05/2013 05:57, Karl O. Pinc wrote:
> > > Hi,
> > > 
> > > I've a new branch on github, browserace:
> > > 
> > > https://github.com/kpinc/phppgadmin/commits/browserace
> > > 
> > > Branches off origin/master, as do all my branches unless
> > > I mention otherwise.
> > > 
> > > This fixes what seems to be a race condition.  Browsers
> > > may load multiple html <script> tags asynchronously,
> > > and so may execute them in an indeterminate order.
> > > Given the way the js code is written the
> > > xloadtree/xtree2.js file must execute before the
> > > xloadtree/xloadtree2.js file, and this has
> > > not been guaranteed.
> > > 
> > > jQuery.getScript() serializes js script loading, but
> > > not execution.  So, I beat the problem
> > > over the head with a php based solution.

> > 
> > Most of our jQuery code should be in a .ready() method anyway.

I've been looking at the html 4.1 and html5 specs and
the script execution model.  They say that scripts evaluate
in order.  This must be regardless of parallel asynchronous
retrieval.  At least if the spec is followed.  So ioguix is 
right and this patch is useless.  The code works fine the way it is.

There must be some other cause of my browser tree problems.
(Maybe it's a broken old browser.)

I'll remove the branch.  (Yay!  One down.)

Using a .ready() method is a good idea but is a separate issue.
There won't be much benefit given the current way things
are done with frames and broswer.php but should definitely
be part of the process of getting rid of the frames. 

Regards,

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

[ppa-dev] Patches which supply a useless XMLHttpRequest plugin

[Karl O. P. <ko...@me...>]

Hi,

I've a new git branch "uselessplugin":

https://github.com/kpinc/phppgadmin/commits/uselessplugin

It creates a plugin that makes useless periodic
web requests of the server in manner configurable
per PPA page.

I'm not sure why we'd want this committed except
as an example.  And we already have working examples,
although I suppose not ones that trigger javascript.

Anyhow, since I wrote it and it "works nicely"
I figured I may as well make it available.

Regards,


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

Re: [ppa-dev] Patches for ppa self-tests

[Jehan-Guillaume (i. de R. <io...@fr...>]

On 22/05/2013 05:39, Robert Treat wrote:
> On Mon, May 20, 2013 at 5:53 PM, Karl O. Pinc <ko...@me...> wrote:
>> P.S.  Where is the code that creates the release?
> 
> http://phppgadmin.cvs.sourceforge.net/viewvc/phppgadmin/project/release.sh?revision=1.9&content-type=text%2Fplain

Wrong, at least for me. The "offical" script (at least for me), since we
escape from CVS and put the website on a SCM wiki is here:

  http://phppgadmin.sourceforge.net/doku.php?id=project:release_script

(requires login)

> I actually thought I had fixed some other issue with it during the
> last release, but that should pretty much work, barring OS difference
> like my sed :-\

yeah, I didn't check the packages you created...anyway, we'll pay
attention for the next release.

> Robert Treat
> play: xzilla.net
> work: omniti.com

Re: [ppa-dev] Patches for ppa self-tests

[Karl O. P. <ko...@me...>]

On 05/21/2013 10:39:51 PM, Robert Treat wrote:
> On Mon, May 20, 2013 at 5:53 PM, Karl O. Pinc <ko...@me...> wrote:
> > On 05/20/2013 03:16:13 PM, Jehan-Guillaume (ioguix) de Rorthais
> wrote:
> >> On 30/04/2013 21:57, Karl O. Pinc wrote:
> <snip>
> >> > 3 patches.
> >> > Internationalizes the selenium link on the ppa intro page.
> >> > Makes the link "hot" only when selenium has been configured.
> >>
> >> These two patch seems useless to me: we remove the selenium links
> >> before
> >> each release of PPA. This would require to remove the i18n as 
> well.
> >
> > Yes...  But this last release selenium was not removed.  It seems
> > useful to be able to run the test suite on released code
> > as an end-user.
> >
> 
> why?

Dunno.  To see if it works given what's installed on the box
and how it's configured?

> I asked some end users, none of them could see any value in it. 
> I
> thought about the software I run, I can't think of any that ships 
> with
> a test suite, nor any i would run. I think this would be different if
> the test were testing for system dependencies or something, but that
> isn't what they do.
> 
> So, after thinking about this, I like the idea of the disable link
> based on config, but I'd still want to remove it all in final 
> release.

Good by me.  Especially if nobody wants it.

What about the il8n-ness?

Do you want a patch to the release code that removes the link?

> 
> > There could be issues, in which case yes these patches are useless.
> > See my response to Robert regards this.
> >
> > It now occurs to me that people run HEAD a lot.  If it's dangerous
> > to have the self-tests on a production system we should probably
> > address this.
> >
> 
> I have sometimes recommended people try head if they are trying to
> verify a problem or some such, but I think the expectation is if
> you're pulling from git, there's no guarantees from us that anything
> will work or be safe (and yes, I personally run head all the time, 
> but
> for client stuff, I tell people to install releases)

I've seen people told to run HEAD on irc.  ;-)

> > P.S.  Where is the code that creates the release?
> 
> http://phppgadmin.cvs.sourceforge.net/viewvc/phppgadmin/project/
> release.sh?revision=1.9&content-type=text%2Fplain

Thanks.


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

Re: [ppa-dev] Patches for ppa self-tests

[Robert T. <ro...@xz...>]

On Mon, May 20, 2013 at 5:53 PM, Karl O. Pinc <ko...@me...> wrote:
> On 05/20/2013 03:16:13 PM, Jehan-Guillaume (ioguix) de Rorthais wrote:
>> On 30/04/2013 21:57, Karl O. Pinc wrote:
<snip>
>> > 3 patches.
>> > Internationalizes the selenium link on the ppa intro page.
>> > Makes the link "hot" only when selenium has been configured.
>>
>> These two patch seems useless to me: we remove the selenium links
>> before
>> each release of PPA. This would require to remove the i18n as well.
>
> Yes...  But this last release selenium was not removed.  It seems
> useful to be able to run the test suite on released code
> as an end-user.
>

why? I asked some end users, none of them could see any value in it. I
thought about the software I run, I can't think of any that ships with
a test suite, nor any i would run. I think this would be different if
the test were testing for system dependencies or something, but that
isn't what they do.

So, after thinking about this, I like the idea of the disable link
based on config, but I'd still want to remove it all in final release.

> There could be issues, in which case yes these patches are useless.
> See my response to Robert regards this.
>
> It now occurs to me that people run HEAD a lot.  If it's dangerous
> to have the self-tests on a production system we should probably
> address this.
>

I have sometimes recommended people try head if they are trying to
verify a problem or some such, but I think the expectation is if
you're pulling from git, there's no guarantees from us that anything
will work or be safe (and yes, I personally run head all the time, but
for client stuff, I tell people to install releases)

> P.S.  Where is the code that creates the release?

http://phppgadmin.cvs.sourceforge.net/viewvc/phppgadmin/project/release.sh?revision=1.9&content-type=text%2Fplain

I actually thought I had fixed some other issue with it during the
last release, but that should pretty much work, barring OS difference
like my sed :-\

Robert Treat
play: xzilla.net
work: omniti.com

Re: [ppa-dev] Patches for ppa self-tests

[Karl O. P. <ko...@me...>]

On 05/20/2013 04:53:31 PM, Karl O. Pinc wrote:
> On 05/20/2013 03:16:13 PM, Jehan-Guillaume (ioguix) de Rorthais 
> wrote:
> > On 30/04/2013 21:57, Karl O. Pinc wrote:
> 
> > > https://github.com/kpinc/phppgadmin/commits/selenium-login
> > > 
> > > 1 patch.
> > > Deletes cookies on selenium login so that previous
> > > failed or paused tests do not affect the execution of
> > > subsequent manually executed tests.  This allows
> > > the user more control over manual test execution.
> > 
> > I don't like this.
> > 
> > I prefer this syntax:
> > 
> >   $this->deleteAllVisibleCookies();
> 
> I'm not sure I understand.  Your suggestion would
> delete all cookies when the Selenium test suite
> is entered, right?
> 
> The point of the patch is
> to be able to make each test independent of the state
> at which the last test leaves the PPA system.  This allows
> tests to fail, be run out of order, and so forth.
> To do this you need to clear the cookies each
> time Selenium starts a test.

I may not recall correctly, but I believe I had
a php bug that resulted in no output.  This borked
the selenium tests since there was nothing to click
on to begin the login process, even though I'd
fixed the php.  Clearing the cookies I could re-run
just the test I was interested in.

Due to my slow X rendering it takes about 25
minutes to run through the selenium tests
if I leave ppa's browser tab in the foreground.
It makes a whole lot of sense to be able
to re-run a single test just by clicking on
it in the selenium control panel.  I'm sure
there's a work-around via re-loading the
entire selenium panel and re-generating
the tests and so forth, but why do I want
to have to think about that when I'm
fixing php code?


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

Re: [ppa-dev] Fix of race condition in db browser frame

[Karl O. P. <ko...@me...>]

On 05/20/2013 03:25:13 PM, Jehan-Guillaume (ioguix) de Rorthais wrote:
> On 15/05/2013 05:57, Karl O. Pinc wrote:
> > Hi,
> > 
> > I've a new branch on github, browserace:
> > 
> > https://github.com/kpinc/phppgadmin/commits/browserace
> > 
> > Branches off origin/master, as do all my branches unless
> > I mention otherwise.
> > 
> > This fixes what seems to be a race condition.  Browsers
> > may load multiple html <script> tags asynchronously,
> > and so may execute them in an indeterminate order.
> > Given the way the js code is written the
> > xloadtree/xtree2.js file must execute before the
> > xloadtree/xloadtree2.js file, and this has
> > not been guaranteed.
> > 
> > jQuery.getScript() serializes js script loading, but
> > not execution.  So, I beat the problem
> > over the head with a php based solution.
> 
> I never have been bitten by this

Every now and then my browse tree has been
completely broken, repeatedly.  It all began ages
ago after the first major xorg release of X,
when the video driver for the card in my ancient xserver
got replace with something non-optimized
for the card.  Since the X has rendered on
the P-III cpu and redisplay of the browser
is slow.  Especially when there's lots of
tabs or windows.

I can't say what breaks it, but this sounds
likely.

> and I think the xtree should be drop
> altogether with the frames anyway. 

I'm all for getting rid of the frames.
But I'm not sure what this has to do with
xtree.  If you want a browse tree you need
some js code to do it.

> As the tree is in Full js anyway,

What is "Full"?

> we
> should be able to make it hide-able with js instead of using ugly,
> heavy, useless, unconfortable and troublesome frames.

The xtree stuff does use js to hide parts of the tree.
I'm not clear on what you're talking about hiding here.

> 
> I would prefer to see someone working on this instead of fixing again
> (potential) bugs with xtree. Let's save some time.

I'm interested if you would explain a bit more.

I want to be able to display all relations tables,
views, and materialize views in a single list and
separately.  This means in the browse tree too.
I already have a design and an implementation
plan.  It means decorating the browse tree
with an additional icon to control what's mixed
together and what's separated.  The xtree2.js
and xloadtree.js code is nice and modular and
straightforward to extend in this fashion.

My plan was to throw some code together so
y'all could see how useful it is to be
able to see all the relations at once
when you want to see things that way.
If you're dead set against this I'd like
to hear of it.

If you're talking about throwing xtree away I need
to know what would replace it.

> 
> Most of our jQuery code should be in a .ready() method anyway.

We could simply put both xloadtree/* files in a single file,
or do something else.  Using php seemed a simple solution.
There's surely a way to use jQuery to solve this problem
but it's not immediately obvious how to do it.
Putting all the xloadtree/* code into a method raises
lexical scoping issues.  Although I suppose that if
there's a single top-level object that's invoked
from the html <script> entity this wouldn't be too
tricky.  Right now browser.php works with several
top-level js objects from the xloadtree/* files.

In any case as I move forward with tables/views/materalized
view display I'm going to be adding js.  I need to
know where to put it, what js code should be extended
to manage the tree, and how the various js files will
be put together.  I will happily make a jQuery.ready()
method out of the xloadtree/* stuff if that's what
you guys want.  But it's going to make browser.php
look more complicated.

It still all seems entirely orthogonal to the frame
issue.  I have enough to do with getting PPA
to support various 9.* view related enhancements that
I don't want to mess with getting rid of frames.

Regards,

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