naviserver-devel — Developer discussion

Re: [naviserver-devel] whole lotta errors

[Zoran V. <zv...@ar...>]

On 13.08.2006, at 17:55, Vlad Seryakov wrote:

> i think i found it, it is the change about switching into chunked mode
> if encoding is used, thus returning the result with length  
> prepended, so
> all test are broken now.
>
> when i disabled it in the sources, some tests are fine now, http.test,
> but still others are broken

As I see from the CVS, there were no changes to the C-code there?
Did you have to change the C-code??
I guess, Bernd will have to explain what he ment to do with that...

Cheers,
Zoran

Re: [naviserver-devel] whole lotta errors

[Vlad S. <vl...@cr...>]

i think i found it, it is the change about switching into chunked mode 
if encoding is used, thus returning the result with length prepended, so 
all test are broken now.

when i disabled it in the sources, some tests are fine now, http.test, 
but still others are broken

Zoran Vasiljevic wrote:
> On 13.08.2006, at 17:28, Vlad Seryakov wrote:
>
>   
>> I remember my last chnage in keepalive but i am sure i ran make  
>> test at
>> that time, it was month ago, after that not sure about other changes
>> really, not many last time
>>     
>
> Your last change was:
>
> 2006-07-21  Vlad Seryakov  <ser...@us...>
>
> 	* tcl/file.tcl: Use wrapped Tcl proc aroun d.tcl file instead of
> 	not exactly as needed global ns_cache.
>
> Since then it was only me fiddling with the ns_log and ns_proxy
> arround. But, I'm sure I did not get thos errors some times
> ago, actually at after 2006-08-04 as this was the last time
> I checked in something, NOT counting the last nsproxylib.c change
> which is totally unrelated to the core.
>
> That means: somebody did change the code (or the testsuite)
> w/o updating the ChangeLog file. I think I will have to go
> to the mail list archive and check the commit maillist as I
> did not get any emails about what was changed since....
>
> I would ask the commiter (we will found who that is soon) to always
> update ChangeLog file so we know what is going on. Otherwise it
> is just a waste of time and nerves...
>
> Cheers
> Zoran
>
>   
>> Zoran Vasiljevic wrote:
>>     
>>> On 13.08.2006, at 17:21, Vlad Seryakov wrote:
>>>
>>>
>>>       
>>>> i am suspecting this happened after encoding patches
>>>>
>>>>         
>>> WHAT encoding patches??? I see nothing in the ChangeLog!
>>> Zoran
>>>
>>> --------------------------------------------------------------------- 
>>> ----
>>> 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
>>> _______________________________________________
>>> naviserver-devel mailing list
>>> nav...@li...
>>> https://lists.sourceforge.net/lists/listinfo/naviserver-devel
>>>
>>>
>>>       
>> ---------------------------------------------------------------------- 
>> ---
>> 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
>> _______________________________________________
>> naviserver-devel mailing list
>> nav...@li...
>> https://lists.sourceforge.net/lists/listinfo/naviserver-devel
>>     
>
>
> -------------------------------------------------------------------------
> 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
> _______________________________________________
> naviserver-devel mailing list
> nav...@li...
> https://lists.sourceforge.net/lists/listinfo/naviserver-devel
>
>   

Re: [naviserver-devel] whole lotta errors

[Zoran V. <zv...@ar...>]

On 13.08.2006, at 18:04, Andrew Piskorski wrote:
>

> Using it shows that between 7/21 and 8/13, there was only one commit
> to Naviserver (ignoring its various modules) by anyone other than
> "vasiljevic", this one:
>
>     2006-08-08 12:07  eide
>
Ah, Bernd!

>     * tests/: encoding.test, tclresp.test, test.nscfg,
>     testserver/modules/testhttp.tcl,
>     testserver/pages/encoding.iso2iso_adp,
>     testserver/pages/encoding.iso2utf_adp,
>     testserver/pages/encoding.iso_adp,
>     testserver/pages/encoding.utf2iso_adp,
>     testserver/pages/encoding.utf2utf_adp,
>     testserver/pages/encoding.utf_adp: *
>     tests/testserver/pages/encoding.*: Various helper files.  *
>     tests/testserver/modules/testhttp.tcl: nstest_http is now encoding
>     aware.  * tests/test.nscfg: Sections for mimetypes and encodings
>     added, along with parameters URLCharset, OutputCharset and
>     HackContentType.  * tests/encoding.test: tests for ns_return and
>     ns_write, changing charsets and verifying URL and ADP encoding.  *
>     tests/tclresp.test: minor modification in tests 3.1 and 4.2 after
>     change in test.nscfg.  Thanks to Michael Lex for writing the  
> tests.

Thanks for getting this one!

So, whatever this means: "nstest_http is now encoding aware"
it either shows that it is (itself) broken or that our core
code is broken :-(
Bernd, which one of those it is?

Cheers
Zoran

Re: [naviserver-devel] whole lotta errors

[Andrew P. <at...@pi...>]

On Sun, Aug 13, 2006 at 05:36:33PM +0200, Zoran Vasiljevic wrote:

> That means: somebody did change the code (or the testsuite)
> w/o updating the ChangeLog file. I think I will have to go
> to the mail list archive and check the commit maillist as I
> did not get any emails about what was changed since....

But Zoran, CVS is the master copy of such information, so ask CVS.
The "cvs2cl.pl" script is useful for getting a by-date changelog style
list of all commits:

  http://www.red-bean.com/cvs2cl/

Using it shows that between 7/21 and 8/13, there was only one commit
to Naviserver (ignoring its various modules) by anyone other than
"vasiljevic", this one:

    2006-08-08 12:07  eide

    * tests/: encoding.test, tclresp.test, test.nscfg,
    testserver/modules/testhttp.tcl,
    testserver/pages/encoding.iso2iso_adp,
    testserver/pages/encoding.iso2utf_adp,
    testserver/pages/encoding.iso_adp,
    testserver/pages/encoding.utf2iso_adp,
    testserver/pages/encoding.utf2utf_adp,
    testserver/pages/encoding.utf_adp: *
    tests/testserver/pages/encoding.*: Various helper files.  *
    tests/testserver/modules/testhttp.tcl: nstest_http is now encoding
    aware.  * tests/test.nscfg: Sections for mimetypes and encodings
    added, along with parameters URLCharset, OutputCharset and
    HackContentType.  * tests/encoding.test: tests for ns_return and
    ns_write, changing charsets and verifying URL and ADP encoding.  *
    tests/tclresp.test: minor modification in tests 3.1 and 4.2 after
    change in test.nscfg.  Thanks to Michael Lex for writing the tests.

(Unfortunately the ancient cvs2cl.pl version 1.2 I'm using tends to
mis-format the the cvs commit messages.)

-- 
Andrew Piskorski <at...@pi...>
http://www.piskorski.com/

Re: [naviserver-devel] whole lotta errors

[Zoran V. <zv...@ar...>]

On 13.08.2006, at 17:28, Vlad Seryakov wrote:

> I remember my last chnage in keepalive but i am sure i ran make  
> test at
> that time, it was month ago, after that not sure about other changes
> really, not many last time

Your last change was:

2006-07-21  Vlad Seryakov  <ser...@us...>

	* tcl/file.tcl: Use wrapped Tcl proc aroun d.tcl file instead of
	not exactly as needed global ns_cache.

Since then it was only me fiddling with the ns_log and ns_proxy
arround. But, I'm sure I did not get thos errors some times
ago, actually at after 2006-08-04 as this was the last time
I checked in something, NOT counting the last nsproxylib.c change
which is totally unrelated to the core.

That means: somebody did change the code (or the testsuite)
w/o updating the ChangeLog file. I think I will have to go
to the mail list archive and check the commit maillist as I
did not get any emails about what was changed since....

I would ask the commiter (we will found who that is soon) to always
update ChangeLog file so we know what is going on. Otherwise it
is just a waste of time and nerves...

Cheers
Zoran

>
> Zoran Vasiljevic wrote:
>> On 13.08.2006, at 17:21, Vlad Seryakov wrote:
>>
>>
>>> i am suspecting this happened after encoding patches
>>>
>>
>> WHAT encoding patches??? I see nothing in the ChangeLog!
>> Zoran
>>
>> --------------------------------------------------------------------- 
>> ----
>> 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
>> _______________________________________________
>> naviserver-devel mailing list
>> nav...@li...
>> https://lists.sourceforge.net/lists/listinfo/naviserver-devel
>>
>>
>
>
> ---------------------------------------------------------------------- 
> ---
> 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
> _______________________________________________
> naviserver-devel mailing list
> nav...@li...
> https://lists.sourceforge.net/lists/listinfo/naviserver-devel

Re: [naviserver-devel] whole lotta errors

[Vlad S. <vl...@cr...>]

I remember my last chnage in keepalive but i am sure i ran make test at 
that time, it was month ago, after that not sure about other changes 
really, not many last time

Zoran Vasiljevic wrote:
> On 13.08.2006, at 17:21, Vlad Seryakov wrote:
>
>   
>> i am suspecting this happened after encoding patches
>>     
>
> WHAT encoding patches??? I see nothing in the ChangeLog!
> Zoran
>
> -------------------------------------------------------------------------
> 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
> _______________________________________________
> naviserver-devel mailing list
> nav...@li...
> https://lists.sourceforge.net/lists/listinfo/naviserver-devel
>
>   

Re: [naviserver-devel] whole lotta errors

[Zoran V. <zv...@ar...>]

On 13.08.2006, at 17:21, Vlad Seryakov wrote:

> i am suspecting this happened after encoding patches

WHAT encoding patches??? I see nothing in the ChangeLog!
Zoran

Re: [naviserver-devel] whole lotta errors

[Vlad S. <vl...@cr...>]

i am suspecting this happened after encoding patches

Zoran Vasiljevic wrote:
> Hi!
>
> Please(!) update ChangeLog when commiting new changes.
>
> There is a whole lotta new errors in the test suite
> in the form of:
>
> ---- Result was:
> 200 {} {1
> x
> 0
>
> }
> ---- Result should have been (exact matching):
> 200 1 x
> ==== tclresp-4.2 FAILED
>
> Now, what IS this? I cannot remember seeing those
> errors couple of days ago when I last commited.
> Since then I saw no entries in the ChangeLog so
> I wonder where is this comming from???
>
> Cheers
> Zoran
>
> -------------------------------------------------------------------------
> 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
> _______________________________________________
> naviserver-devel mailing list
> nav...@li...
> https://lists.sourceforge.net/lists/listinfo/naviserver-devel
>
>   

[naviserver-devel] whole lotta errors

[Zoran V. <zv...@ar...>]

Hi!

Please(!) update ChangeLog when commiting new changes.

There is a whole lotta new errors in the test suite
in the form of:

---- Result was:
200 {} {1
x
0

}
---- Result should have been (exact matching):
200 1 x
==== tclresp-4.2 FAILED

Now, what IS this? I cannot remember seeing those
errors couple of days ago when I last commited.
Since then I saw no entries in the ChangeLog so
I wonder where is this comming from???

Cheers
Zoran

[naviserver-devel] Encoding: default charset

[Michael L. <mic...@gm...>]

When no OutpuCharset is defined, naviserver will send the content
without transformation, that means utf-8. But according to the RFC
2616 (HTTP 1.1) all content without an explicit "charset: ..." in the
Content-Type-Header should be treated as iso-8859-1 by the clients.
This causes problems when you have an incomplete configuration (w/o
OutputCharset).

Michael

[naviserver-devel] Encoding

[Michael L. <mic...@gm...>]

I'm new to Naviserver and I'm not quite sure about the intended use of ns_conn
encoding.

There are three situation where it is possible to change the encoding by hand:

1. When using ns_write, you can set the encoding with ns_startcontent
-type. You could also use ns_conn encoding, but ns_startcontent was
designed only for this purpose, so i think it would be better to use
this function.

2. Inside adp-files one can use ns_conn encoding. But it is necessary
to change the Content-Type-Header explicitly. The function
ns_adp_mimetype will set the Content-Header and change the encoding,
so the use of this function is easier and less dangerous.

3. ns_return expects a type-argument which will be used to create the
Content-Type-Header. If you specify a charset, naviserver will
automatically load the corresponding tcl-encoding. So here you have
also no need to change the encoding with ns_conn encoding (btw if you
do so, the result is very confusing).

For me it looks like the use cases for ns_conn encoding are rare, e.g.
if you use it in a context where the client knows exactly what encoding to
expect. The documentation should mention the above alternatives and warn not
to use ns_conn encoding if you "don't know all side effects."

Michael

Re: [naviserver-devel] Encoding

[Bernd E. <eid...@we...>]

Hi Stephen,

I asked my colleague Michael to apply your patch and start writing tests for 
the encoding. He wrote some and it looks like there should some other things 
be considered/fixed first to make them meaningful.

After that we can discuss a solution for the Content-Length/Encoding problem 
which was introduced by the scatter/gather algorithms.

I'll ask Michael to join the list and summarize what he found in the code.

Bernd.

> >From your explanations of encodings, it sounds like you've got a real
>
> handle on this!  A lot of nice debugging and testing there too.  If
> you turn that into some real tests and put them in tests/ecoding.test,
> I'll give you a big kiss!
>
> Re the open bug: I'm pretty sure removing the shortcut for utf-8 will
> fix it, but I'm having trouble writing tests to prove it!
>
> Attached are some tests I was working on (ouch!) 4 months ago...  Big
> kisses all round for anyone who can figure this out.

Re: [naviserver-devel] Hang in nsproxy tests

[Zoran V. <zv...@ar...>]

On 28.07.2006, at 23:52, Zoran Vasiljevic wrote:

> OK. This should be easy. It does not appear on Darwin or Solaris.

WRONG! This happens on all machines! I guess I have added that
one in the very last (innocent) step to remove round-robin usage
of proxies and hence keep the number of nsproxy processes at miniumum.
This was just a small change and I did not run the test-suite for that.
Good lesson to learn: run the test suite always, regardless how small
the change is.

Anyways, the fix is now checked in.

Cheers,
Zoran

Re: [naviserver-devel] Hang in nsproxy tests

[Zoran V. <zv...@ar...>]

On 29.07.2006, at 00:00, Stephen Deasey wrote:

>
> If i leave it long enough, it starts to log ADP page cache  
> statistics...!?!?!
>
> Laptop's getting toasty..!

This is some stupid memory thing...

Re: [naviserver-devel] Hang in nsproxy tests

[Stephen D. <sd...@gm...>]

On 7/28/06, Zoran Vasiljevic <zv...@ar...> wrote:
>
> On 28.07.2006, at 23:39, Stephen Deasey wrote:
>
> >
> > It buzzes the CPU.
>
>
> I see. It skyrockets to 100%.
> OK. This should be easy. It does not appear on Darwin or Solaris.
> I will fix this tomorrow...
>

If i leave it long enough, it starts to log ADP page cache statistics...!?!?!

Laptop's getting toasty..!

Re: [naviserver-devel] Hang in nsproxy tests

[Zoran V. <zv...@ar...>]

On 28.07.2006, at 23:39, Stephen Deasey wrote:

>
> It buzzes the CPU.


I see. It skyrockets to 100%.
OK. This should be easy. It does not appear on Darwin or Solaris.
I will fix this tomorrow...

Cheers
Zoran

Re: [naviserver-devel] Hang in nsproxy tests

[Stephen D. <sd...@gm...>]

On 7/28/06, Zoran Vasiljevic <zv...@ar...> wrote:
>
> On 28.07.2006, at 23:39, Stephen Deasey wrote:
>
> > The nsproxy tests are hanging for me soon after this message:
> >
> > ns_proxy.test
> > [28/Jul/2006:22:36:25][3065.2895170464][-nsproxy:reap-] Notice:
> > starting
> >
> > It buzzes the CPU.
> >
> > Looks like memory corruption.
>
> I will look at this tomorrow. BTW what does it mean "buzzes the CPU" ?


100% usage.

Re: [naviserver-devel] Hang in nsproxy tests

[Zoran V. <zv...@ar...>]

On 28.07.2006, at 23:39, Stephen Deasey wrote:

> The nsproxy tests are hanging for me soon after this message:
>
> ns_proxy.test
> [28/Jul/2006:22:36:25][3065.2895170464][-nsproxy:reap-] Notice:  
> starting
>
> It buzzes the CPU.
>
> Looks like memory corruption.

I will look at this tomorrow. BTW what does it mean "buzzes the CPU" ?
Cheers
zoran

[naviserver-devel] Hang in nsproxy tests

[Stephen D. <sd...@gm...>]

The nsproxy tests are hanging for me soon after this message:

ns_proxy.test
[28/Jul/2006:22:36:25][3065.2895170464][-nsproxy:reap-] Notice: starting

It buzzes the CPU.

Looks like memory corruption.

Linux 2.6.

Re: [naviserver-devel] [naviserver-commits] naviserver/tcl file.tcl, 1.7, 1.8

[Vlad S. <vl...@cr...>]

No problem, i just wanted to remove non-functional global cache

Stephen Deasey wrote:
> Some people 'cache' content to disk in the form of Tcl pages. There's
> the logic in there that needs to run at page delivery time, plus some
> data from a db, for example.  New Tcl pages are created as new content
> arrives, deleted as necessary.
> 
> The ns:tclcache_* procs are only ever created.  This is a cache of
> infinite size.
> 
> It may be easier to just bite the bullet and create an ADP wrapper
> that simulates a single-chunk ADP page...
> 
> 
> 
> On 7/22/06, Vlad Seryakov <ser...@us...> wrote:
>> Update of /cvsroot/naviserver/naviserver/tcl
>> In directory sc8-pr-cvs7.sourceforge.net:/tmp/cvs-serv23802/tcl
>>
>> Modified Files:
>>         file.tcl
>> Log Message:
>> Use wrapped Tcl proc aroun d.tcl file instead of not exactly as needed working global ns_cache for cacheing Tcl bytecode
>>
>>
>> Index: file.tcl
>> ===================================================================
>> RCS file: /cvsroot/naviserver/naviserver/tcl/file.tcl,v
>> retrieving revision 1.7
>> retrieving revision 1.8
>> diff -C2 -d -r1.7 -r1.8
>> *** file.tcl    9 Mar 2006 11:25:55 -0000       1.7
>> --- file.tcl    22 Jul 2006 02:24:09 -0000      1.8
>> ***************
>> *** 53,57 ****
>>       ns_register_proc POST /*.tcl ns_sourceproc
>>       ns_register_proc HEAD /*.tcl ns_sourceproc
>> -     ns_cache_create ns:filecache [ns_config $path filecachesize 5000000]
>>   }
>>
>> --- 53,56 ----
>> ***************
>> *** 86,126 ****
>>   #
>>   # Side effects:
>> ! #   Caches the file content and Tcl_Obj's rep of the
>> ! #   sourced script in per-thread cache (is this true?)
>>   #
>>
>> ! proc ns_sourcefile {filename} {
>> !
>> !     file stat $filename stat
>> !     set current_cookie $stat(mtime):$stat(ctime):$stat(ino):$stat(dev)
>> !
>> !     #
>> !     # Read current cached file
>> !     #
>> !
>> !     set pair [ns_cache_eval ns:filecache $filename {
>> !         list $current_cookie [ns_fileread $filename]
>> !     }]
>>
>> !     #
>> !     # If changed, re-cache it
>> !     #
>>
>> !     if {[lindex $pair 0] ne $current_cookie} {
>> !         ns_cache_flush ns:filecache $filename
>> !         set pair [ns_cache_eval ns:filecache $filename {
>> !             list $current_cookie [ns_fileread $filename]
>> !         }]
>>       }
>> !
>> !     #
>> !     # And here's the magic part. We're using "for" here to translate the
>> !     # text source file into bytecode, which will be associated with the
>> !     # Tcl_Obj we just cached (as its internal representation).  "eval"
>> !     # doesn't do this as the eval provided in Tcl uses the TCL_EVAL_DIRECT
>> !     # flag, and hence interprets the text directly.
>> !     #
>> !
>> !     uplevel [for [lindex $pair 1] {0} {} {}]
>>   }
>>
>> --- 85,107 ----
>>   #
>>   # Side effects:
>> ! #   Each .tcl file will be wrapped into Tcl proc in every thread
>>   #
>>
>> ! proc ns_sourcefile { path } {
>>
>> !     set proc0 [info procs ns:tclcache_$path]
>> !     file stat $path stat
>> !     set cookie0 $stat(mtime):$stat(ctime):$stat(ino):$stat(dev)
>>
>> !     # Verify file modification time
>> !     if { $proc0 == "" || [$proc0 1] != $cookie0 } {
>> !       set code [ns_fileread $path]
>> !       proc ns:tclcache_$path { {getcookie 0} } "
>> !          if { \$getcookie } { return $cookie0 }
>> !          $code
>> !       "
>>       }
>> !     # Run the proc
>> !     ns:tclcache_$path
>>   }
>>
>>
>>
>> -------------------------------------------------------------------------
>> 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
>> _______________________________________________
>> naviserver-commits mailing list
>> nav...@li...
>> https://lists.sourceforge.net/lists/listinfo/naviserver-commits
>>
> 
> -------------------------------------------------------------------------
> 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
> _______________________________________________
> naviserver-devel mailing list
> nav...@li...
> https://lists.sourceforge.net/lists/listinfo/naviserver-devel
> 

-- 
Vlad Seryakov
571 262-8608 office
vl...@cr...
http://www.crystalballinc.com/vlad/

Re: [naviserver-devel] [naviserver-commits] naviserver/tcl file.tcl, 1.7, 1.8

[Stephen D. <sd...@gm...>]

Some people 'cache' content to disk in the form of Tcl pages. There's
the logic in there that needs to run at page delivery time, plus some
data from a db, for example.  New Tcl pages are created as new content
arrives, deleted as necessary.

The ns:tclcache_* procs are only ever created.  This is a cache of
infinite size.

It may be easier to just bite the bullet and create an ADP wrapper
that simulates a single-chunk ADP page...



On 7/22/06, Vlad Seryakov <ser...@us...> wrote:
> Update of /cvsroot/naviserver/naviserver/tcl
> In directory sc8-pr-cvs7.sourceforge.net:/tmp/cvs-serv23802/tcl
>
> Modified Files:
>         file.tcl
> Log Message:
> Use wrapped Tcl proc aroun d.tcl file instead of not exactly as needed working global ns_cache for cacheing Tcl bytecode
>
>
> Index: file.tcl
> ===================================================================
> RCS file: /cvsroot/naviserver/naviserver/tcl/file.tcl,v
> retrieving revision 1.7
> retrieving revision 1.8
> diff -C2 -d -r1.7 -r1.8
> *** file.tcl    9 Mar 2006 11:25:55 -0000       1.7
> --- file.tcl    22 Jul 2006 02:24:09 -0000      1.8
> ***************
> *** 53,57 ****
>       ns_register_proc POST /*.tcl ns_sourceproc
>       ns_register_proc HEAD /*.tcl ns_sourceproc
> -     ns_cache_create ns:filecache [ns_config $path filecachesize 5000000]
>   }
>
> --- 53,56 ----
> ***************
> *** 86,126 ****
>   #
>   # Side effects:
> ! #   Caches the file content and Tcl_Obj's rep of the
> ! #   sourced script in per-thread cache (is this true?)
>   #
>
> ! proc ns_sourcefile {filename} {
> !
> !     file stat $filename stat
> !     set current_cookie $stat(mtime):$stat(ctime):$stat(ino):$stat(dev)
> !
> !     #
> !     # Read current cached file
> !     #
> !
> !     set pair [ns_cache_eval ns:filecache $filename {
> !         list $current_cookie [ns_fileread $filename]
> !     }]
>
> !     #
> !     # If changed, re-cache it
> !     #
>
> !     if {[lindex $pair 0] ne $current_cookie} {
> !         ns_cache_flush ns:filecache $filename
> !         set pair [ns_cache_eval ns:filecache $filename {
> !             list $current_cookie [ns_fileread $filename]
> !         }]
>       }
> !
> !     #
> !     # And here's the magic part. We're using "for" here to translate the
> !     # text source file into bytecode, which will be associated with the
> !     # Tcl_Obj we just cached (as its internal representation).  "eval"
> !     # doesn't do this as the eval provided in Tcl uses the TCL_EVAL_DIRECT
> !     # flag, and hence interprets the text directly.
> !     #
> !
> !     uplevel [for [lindex $pair 1] {0} {} {}]
>   }
>
> --- 85,107 ----
>   #
>   # Side effects:
> ! #   Each .tcl file will be wrapped into Tcl proc in every thread
>   #
>
> ! proc ns_sourcefile { path } {
>
> !     set proc0 [info procs ns:tclcache_$path]
> !     file stat $path stat
> !     set cookie0 $stat(mtime):$stat(ctime):$stat(ino):$stat(dev)
>
> !     # Verify file modification time
> !     if { $proc0 == "" || [$proc0 1] != $cookie0 } {
> !       set code [ns_fileread $path]
> !       proc ns:tclcache_$path { {getcookie 0} } "
> !          if { \$getcookie } { return $cookie0 }
> !          $code
> !       "
>       }
> !     # Run the proc
> !     ns:tclcache_$path
>   }
>
>
>
> -------------------------------------------------------------------------
> 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
> _______________________________________________
> naviserver-commits mailing list
> nav...@li...
> https://lists.sourceforge.net/lists/listinfo/naviserver-commits
>

Re: [naviserver-devel] another sample config issue and some database related questions...

[Mike <nee...@gm...>]

On 7/22/06, Vlad Seryakov <vl...@cr...> wrote:
> I agree, actually i have foreach command in OSSWEB and i do not know why
> it is not included in to nstk, let me add it. Also, tell me if what kind
> of functionality you would like to see in nstk without make it bloated
> or big. I know it is very limited, that was the goal.
>
> I am familiar with OpenACS and i created nstk for only reason not to add
> any additional baggage so it can be used standalone with no
> dependencies. Of course it is mutual to having full-blown system like
> OpenACS.

It is not clear to me that it makes sense to enhance nstk if Stephen
is working on something similar.  I think a consensus on the "one true
direction" should be made and efforts placed into it.  Do you agree?

Re: [naviserver-devel] another sample config issue and some database related questions...

[Vlad S. <vl...@cr...>]

I agree, actually i have foreach command in OSSWEB and i do not know why 
it is not included in to nstk, let me add it. Also, tell me if what kind 
of functionality you would like to see in nstk without make it bloated 
or big. I know it is very limited, that was the goal.

I am familiar with OpenACS and i created nstk for only reason not to add 
any additional baggage so it can be used standalone with no 
dependencies. Of course it is mutual to having full-blown system like 
OpenACS.

> 
> Vlad,
>   I am familiar with OpenACS database API.  I am used to it.  nstk is
> very limited as a database API - it is intended to output to a
> template, so, for example, it's missing a db_foreach sort of
> functionality.  I didn't say it is terrible - I said it hasn't been
> terrible working with it - but not nearly as pleasant as the ns_db API
> I am used to.  I also looked at OSSWEB, but it brings with itself a
> whole bunch of baggage that I'm not ready for, e.g.
> ossweb::util::parse_args ...
> 
> -------------------------------------------------------------------------
> 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
> _______________________________________________
> naviserver-devel mailing list
> nav...@li...
> https://lists.sourceforge.net/lists/listinfo/naviserver-devel
> 


-- 
Vlad Seryakov
571 262-8608 office
vl...@cr...
http://www.crystalballinc.com/vlad/

Re: [naviserver-devel] another sample config issue and some database related questions...

[Mike <nee...@gm...>]

On 7/22/06, Vlad Seryakov <vl...@cr...> wrote:
> nstk is quite old and i have not updated in for a long time.
>
> Can you describe what was not pleasant and what was terrible, i may work
> on that to improve it.

Vlad,
  I am familiar with OpenACS database API.  I am used to it.  nstk is
very limited as a database API - it is intended to output to a
template, so, for example, it's missing a db_foreach sort of
functionality.  I didn't say it is terrible - I said it hasn't been
terrible working with it - but not nearly as pleasant as the ns_db API
I am used to.  I also looked at OSSWEB, but it brings with itself a
whole bunch of baggage that I'm not ready for, e.g.
ossweb::util::parse_args ...

Re: [naviserver-devel] another sample config issue and some database related questions...

[Vlad S. <vl...@cr...>]

nstk is quite old and i have not updated in for a long time.

Can you describe what was not pleasant and what was terrible, i may work 
on that to improve it.

Mike wrote:
>>> Right now I am
>>> looking at 4 choices:
>>>
>>> 1) drop nsdb from naviserver and use nstcl after cleaning up any
>>> multi-thread issues
>>> 2) import ad_proc and 00-database from OACS and hack them up until
>>> they work.  (unfortunately the combination of the two is not possible
>>> - nstcl ad_proc does not support positional parameters without default
>>> values after named parameters ith default values)
>>> 3) dig into OSSWEB and import its notion of ad_proc and db_*
>>> 4) nsdbi and friends
>>>
>>> Numbers 3 and 4 sound most appealing to me at this moment because
>>> these are "most supported" by the naviserver "community" - getting
>>> help with 00-database is unlikely to happen for something unrelated to
>>> OACS and nstcl hasn't been touched in 2+ years.  The downside to
>>> Number 3 is that i need to import a bunch of things with OSSWEB that
>>> are not related to the database, and I am hesitant about doing so.
>>> Your thoughts, comments, and opinions are most welcome.
>>>
>> You're using Postgres, right?
> 
> Correct.  Thus far I've been trying to use nstk.  It hasn't been
> terrible, but it hasn't been pleasant either...
> 
> -------------------------------------------------------------------------
> 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
> _______________________________________________
> naviserver-devel mailing list
> nav...@li...
> https://lists.sourceforge.net/lists/listinfo/naviserver-devel
> 


-- 
Vlad Seryakov
571 262-8608 office
vl...@cr...
http://www.crystalballinc.com/vlad/