Why is Extract DDL option gone from Database menu?

[Milan Babuskov-4]

Hi,

In FlameRobin 0.8.3 (I cannot dig through painfully slow SVN, but I
found this in changelog) I added an option to Extract DDL for entire
database to Advanced menu. This opens directly in SQL editor instead
of going through wxHTML which is painfully slow on databases with many
objects. I just had to export entire database (about 300 tables, 200
procedures and 50 views) and I waited FlameRobin about 20 minutes (on
a 2.16GHz machine) before killing it and doing the dump with isql. FR
is essentially useless this way. FR is linked against wxWidgets 2.8.12
BTW.

It looks like since Database menu has been restructured, this option is gone?

I'm thinking we shouldn't open the DDL HTML page at all, but always go
to SQL editor. HTML is fine for individual objects, but it is useless
for databases.

P.S. While trying to use some rarely used features and fix some bugs
lately, it seems to me that we have so many half-baked "improvements"
that were unfinished or done in haste that only lowered the quality of
FR. I don't feel like releasing any new version soon.

--
Milan Babuskov
http://www.guacosoft.com

------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn
about Cisco certifications, training, and career opportunities.
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Gordon Niessen-2]

In 0.9.3.2106: Database/Advanced/Extract metadata DDL

On 10/22/2011 5:06 PM, Milan Babuskov wrote:

> Hi,
>
> In FlameRobin 0.8.3 (I cannot dig through painfully slow SVN, but I
> found this in changelog) I added an option to Extract DDL for entire
> database to Advanced menu. This opens directly in SQL editor instead
> of going through wxHTML which is painfully slow on databases with many
> objects. I just had to export entire database (about 300 tables, 200
> procedures and 50 views) and I waited FlameRobin about 20 minutes (on
> a 2.16GHz machine) before killing it and doing the dump with isql. FR
> is essentially useless this way. FR is linked against wxWidgets 2.8.12
> BTW.
>
> It looks like since Database menu has been restructured, this option is gone?
>
> I'm thinking we shouldn't open the DDL HTML page at all, but always go
> to SQL editor. HTML is fine for individual objects, but it is useless
> for databases.
>
> P.S. While trying to use some rarely used features and fix some bugs
> lately, it seems to me that we have so many half-baked "improvements"
> that were unfinished or done in haste that only lowered the quality of
> FR. I don't feel like releasing any new version soon.
>

--
Thanks,

Gordon


------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn
about Cisco certifications, training, and career opportunities.
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Milan Babuskov-4]

On Sun, Oct 23, 2011 at 2:02 PM, Gordon Niessen <[hidden email]> wrote:
> In 0.9.3.2106: Database/Advanced/Extract metadata DDL

In 0.9.3.2190 there is no Database/Advanced anymore.

--
Milan Babuskov
http://www.guacosoft.com

------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn
about Cisco certifications, training, and career opportunities.
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Gordon Niessen-2]

On 10/23/2011 8:21 AM, Milan Babuskov wrote:
> On Sun, Oct 23, 2011 at 2:02 PM, Gordon Niessen<[hidden email]>  wrote:
>> In 0.9.3.2106: Database/Advanced/Extract metadata DDL
> In 0.9.3.2190 there is no Database/Advanced anymore.
>
I am glad I didn't upgrade.

--
Thanks,

Gordon


------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn
about Cisco certifications, training, and career opportunities.
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Ismael L. Donis Garcia]

Is it placed on the page of sourseforge for his unloading?

Best Regards
=========
|| ISMAEL ||
=========
----- Original Message -----
From: "Milan Babuskov" <[hidden email]>
To: "Development list" <[hidden email]>
Sent: Sunday, October 23, 2011 9:21 AM
Subject: Re: [Flamerobin-devel] Why is Extract DDL option gone from
Databasemenu?

> On Sun, Oct 23, 2011 at 2:02 PM, Gordon Niessen <[hidden email]>
> wrote:
>> In 0.9.3.2106: Database/Advanced/Extract metadata DDL
>
> In 0.9.3.2190 there is no Database/Advanced anymore.
>
> --
> Milan Babuskov
> http://www.guacosoft.com
>

------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn
about Cisco certifications, training, and career opportunities.
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Nando Dessena]

Hello Milan,
sorry for being absent, this is a very very busy period for me.

> It looks like since Database menu has been restructured, this option
> is gone?

Must have been an oversight on my part, or maybe I thought that doing
the same thing in two places was overkill and removed the wrong one,
apparently. I can't remember anything, just speculating.

> I'm thinking we shouldn't open the DDL HTML page at all, but always
> go to SQL editor. HTML is fine for individual objects, but it is
> useless for databases.

I tend to agree. In order to avoid confusing people used to find the
option there, maybe we can keep the DDL page with just a link in it that
extracts the DDL in a SQL editor, without ever rendering it in the HTML
frame. Would that satisfy everyone, in your opinion?

> P.S. While trying to use some rarely used features and fix some bugs
> lately, it seems to me that we have so many half-baked
> "improvements" that were unfinished or done in haste that only
> lowered the quality of FR. I don't feel like releasing any new
> version soon.

OTOH there hasn't been a release of FR in quite a while. Can you list
the quality issues you are talking about and say what you feel is needed
to fix them?

P.S. I can work on a release but not in the next few weeks.

Ciao
--
Nando Dessena


------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn
about Cisco certifications, training, and career opportunities.
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Gordon Niessen-2]

On 10/24/2011 9:04 AM, Nando Dessena wrote:
> I tend to agree. In order to avoid confusing people used to find the
> option there, maybe we can keep the DDL page with just a link in it that
> extracts the DDL in a SQL editor, without ever rendering it in the HTML
> frame. Would that satisfy everyone, in your opinion?
That works for me.

--
Thanks,

Gordon


------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn
about Cisco certifications, training, and career opportunities.
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Milan Babuskov-4]

On Mon, Oct 24, 2011 at 2:29 PM, Ismael L. Donis Garcia
<[hidden email]> wrote:
> Is it placed on the page of sourseforge for his unloading?

No.

We're just considering making a new official release, and we need to
clear all the regressions at least.

--
Milan Babuskov
http://www.guacosoft.com

------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn
about Cisco certifications, training, and career opportunities.
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Milan Babuskov-4]

On Mon, Oct 24, 2011 at 4:04 PM, Nando Dessena <[hidden email]> wrote:
> I tend to agree. In order to avoid confusing people used to find the
> option there, maybe we can keep the DDL page with just a link in it that
> extracts the DDL in a SQL editor, without ever rendering it in the HTML
> frame. Would that satisfy everyone, in your opinion?

I believe it would. It just isn't consistent with the rest of property
pages, but if that does not bother anyone then I believe it is the
best solution.

> OTOH there hasn't been a release of FR in quite a while. Can you list
> the quality issues you are talking about and say what you feel is needed
> to fix them?

> P.S. I can work on a release but not in the next few weeks.

I'm thinking about making a new release during Firebird Conference,
because there might be a new release of FB at the time. Beside few
issues raise in the past month on this mailing list, I'd like to also
go through Firebird 2.5 changes and add full support for FB 2.5.

--
Milan Babuskov
http://www.guacosoft.com

------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn
about Cisco certifications, training, and career opportunities.
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Milan Babuskov-4]

On Mon, Oct 24, 2011 at 4:04 PM, Nando Dessena <[hidden email]> wrote:
> OTOH there hasn't been a release of FR in quite a while. Can you list
> the quality issues you are talking about and say what you feel is needed
> to fix them?

Here's a list of things we should fix (IMHO) before 0.9.4 release:

- extract DDL for database should go directly to SQL editor
- fix bug from e-mail: SQL generation bug for nullability change
- fix loss of collate for stored procedure parameters (FB2.5)
- fix loss of default values for stored procedure parameters (FB2.5)

There's sf.net bug #3070458 that might relate to this, we need to check.

Everything else is either fixed, or is a feature request, or is a minor bug.

--
Milan Babuskov
http://www.guacosoft.com

------------------------------------------------------------------------------
RSA&reg; Conference 2012
Save &#36;700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Valdir Marcos da Silva]

In SQL Editor, could the final message "Total execution time: 0:00:01
(hh:mm:ss)" show complete timestamp such as hh:mm:ss.nnnn?

Thanks,

Valdir


2011/11/1 Milan Babuskov <[hidden email]>:

> On Mon, Oct 24, 2011 at 4:04 PM, Nando Dessena <[hidden email]> wrote:
>> OTOH there hasn't been a release of FR in quite a while. Can you list
>> the quality issues you are talking about and say what you feel is needed
>> to fix them?
>
> Here's a list of things we should fix (IMHO) before 0.9.4 release:
>
> - extract DDL for database should go directly to SQL editor
> - fix bug from e-mail: SQL generation bug for nullability change
> - fix loss of collate for stored procedure parameters (FB2.5)
> - fix loss of default values for stored procedure parameters (FB2.5)
>
> There's sf.net bug #3070458 that might relate to this, we need to check.
>
> Everything else is either fixed, or is a feature request, or is a minor bug.
>
> --
> Milan Babuskov
> http://www.guacosoft.com
>
> ------------------------------------------------------------------------------
> RSA&reg; Conference 2012
> Save &#36;700 by Nov 18
> Register now
> http://p.sf.net/sfu/rsa-sfdev2dev1
> _______________________________________________
> Flamerobin-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/flamerobin-devel
>

------------------------------------------------------------------------------
RSA&#174; Conference 2012
Save $700 by Nov 18
Register now&#33;
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Milan Babuskov-4]

On Tue, Nov 1, 2011 at 7:21 PM, Valdir Marcos <[hidden email]> wrote:
> In SQL Editor, could the final message "Total execution time: 0:00:01
> (hh:mm:ss)" show complete timestamp such as hh:mm:ss.nnnn?

It could, but what's the point?

Query performance can be affected by so many things that sub-second
variations are completely irrelevant to anything.

--
Milan Babuskov
http://www.guacosoft.com

------------------------------------------------------------------------------
RSA&#174; Conference 2012
Save $700 by Nov 18
Register now&#33;
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Valdir Marcos da Silva]

I see your view-point as correct when we think about an isolated
command, but in my scenario, it would help in getting the best result
in single commands that will be used in long stored procedures that
will be called thousands times per minute in a 24x7 system.

Even though performance may vary on many things, I use all available
resources to get my job best done.

Thanks,

Valdir



2011/11/1 Milan Babuskov <[hidden email]>:

> On Tue, Nov 1, 2011 at 7:21 PM, Valdir Marcos <[hidden email]> wrote:
>> In SQL Editor, could the final message "Total execution time: 0:00:01
>> (hh:mm:ss)" show complete timestamp such as hh:mm:ss.nnnn?
>
> It could, but what's the point?
>
> Query performance can be affected by so many things that sub-second
> variations are completely irrelevant to anything.
>
> --
> Milan Babuskov
> http://www.guacosoft.com
>
> ------------------------------------------------------------------------------
> RSA&#174; Conference 2012
> Save $700 by Nov 18
> Register now&#33;
> http://p.sf.net/sfu/rsa-sfdev2dev1
> _______________________________________________
> Flamerobin-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/flamerobin-devel
>

------------------------------------------------------------------------------
RSA&#174; Conference 2012
Save $700 by Nov 18
Register now&#33;
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Milan Babuskov-4]

On Wed, Nov 2, 2011 at 6:02 AM, Valdir Marcos <[hidden email]> wrote:
> I see your view-point as correct when we think about an isolated
> command, but in my scenario, it would help in getting the best result
> in single commands that will be used in long stored procedures that
> will be called thousands times per minute in a 24x7 system.

I would never benchmark a single call of such procedure, because that
is simple not precise and would probably differ in production. If that
is really the case, I would write another procedure or execute block
and run it 100 or 1000 times and then see how much time it took.

Besides, milisecond precision depends on the operating system clock
precision. AFAIK, that's one second on Windows XP and all previous
Windows versions (not sure about Vista and 7) and Linux introduced
microsecond in kernel 2.6 IIRC. I don't think we should include a
feature that is not consistent and reliable with results you get.

--
Milan Babuskov
http://www.guacosoft.com

------------------------------------------------------------------------------
RSA&#174; Conference 2012
Save $700 by Nov 18
Register now&#33;
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Milan Babuskov-4]

On Wed, Nov 2, 2011 at 6:58 AM, Milan Babuskov <[hidden email]> wrote:
> I would never benchmark a single call of such procedure, because that
> is simple not precise and would probably differ in production.

A single call has overhead of transaction start and statement prepare
as well. And Firebird caching can also kick in and give different
results. If we would really include this feature, I'm sure you would
get inconsistent results between any two statement runs in FlameRobin
itself.  And if it involves network access as well, it's completely
unusable.

You might want to benchmark many calls of the same procedure at once
or place the time measurement code inside the procedure (using 'NOW'
vs timestamp) to get better, more precise results.

--
Milan Babuskov
http://www.guacosoft.com

------------------------------------------------------------------------------
RSA&#174; Conference 2012
Save $700 by Nov 18
Register now&#33;
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Michael Hieke]

On 01.11.2011 19:21, Valdir Marcos wrote:

> In SQL Editor, could the final message "Total execution time:
> 0:00:01 (hh:mm:ss)" show complete timestamp such as hh:mm:ss.nnnn?

It does that already.  It uses the following function to format the
elapsed time:

   wxString millisToTimeString(long millis)
   {
     if (millis >= 60 * 1000)
     {
       int hh = millis / (60 * 60 * 1000);
       millis -= 60 * 60 * 1000 * hh;
       int mm = millis / (60 * 1000);
       millis -= 60 * 1000 * mm;
       int ss = (millis + 500) / 1000;
       return wxString::Format(wxT("%d:%.2d:%.2d (hh:mm:ss)"),
         hh, mm, ss);
     }
     else
       return wxString::Format(wxT("%.3fs"), 0.001 * millis);
   }

As you can see there are two code paths, one to show seconds only with
fractional part, the other to show hours, minutes and seconds.  The
cut-off is at one minute.

You will agree that for times from one minute up there is little sense
in showing fractional seconds.  One could even argue that there
shouldn't be milliseconds precision for durations from 10 seconds and
up, a single digit after the decimal separator would probably be good
enough.

Thanks

--
Michael Hieke

------------------------------------------------------------------------------
RSA&#174; Conference 2012
Save $700 by Nov 18
Register now&#33;
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel

[Valdir Marcos da Silva]

You are right.

Thanks,

Valdir


2011/11/2 Michael Hieke <[hidden email]>:

> On 01.11.2011 19:21, Valdir Marcos wrote:
>
>> In SQL Editor, could the final message "Total execution time:
>> 0:00:01 (hh:mm:ss)" show complete timestamp such as hh:mm:ss.nnnn?
>
> It does that already.  It uses the following function to format the
> elapsed time:
>
>   wxString millisToTimeString(long millis)
>   {
>     if (millis >= 60 * 1000)
>     {
>       int hh = millis / (60 * 60 * 1000);
>       millis -= 60 * 60 * 1000 * hh;
>       int mm = millis / (60 * 1000);
>       millis -= 60 * 1000 * mm;
>       int ss = (millis + 500) / 1000;
>       return wxString::Format(wxT("%d:%.2d:%.2d (hh:mm:ss)"),
>         hh, mm, ss);
>     }
>     else
>       return wxString::Format(wxT("%.3fs"), 0.001 * millis);
>   }
>
> As you can see there are two code paths, one to show seconds only with
> fractional part, the other to show hours, minutes and seconds.  The
> cut-off is at one minute.
>
> You will agree that for times from one minute up there is little sense
> in showing fractional seconds.  One could even argue that there
> shouldn't be milliseconds precision for durations from 10 seconds and
> up, a single digit after the decimal separator would probably be good
> enough.
>
> Thanks
>
> --
> Michael Hieke
>
> ------------------------------------------------------------------------------
> RSA&#174; Conference 2012
> Save $700 by Nov 18
> Register now&#33;
> http://p.sf.net/sfu/rsa-sfdev2dev1
> _______________________________________________
> Flamerobin-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/flamerobin-devel
>

------------------------------------------------------------------------------
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Flamerobin-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/flamerobin-devel