cppcms-users — Disussions of users of CppCMS

Re: [Cppcms-users] CppDB on github

[mawan s. <ma...@gm...>]

Cppcms the best

On Mar 9, 2017 09:16, "Shiv Shankar Dayal" <shi...@gm...>
wrote:

>

> Artyom,
>
> Can you please being CppDB on github as well? Mighty thanks.
>
> I do not have Paypal login. I tried to donate but ... I will make login
and do some donation for your hard work.
>
> Please note that I have never made a penny with CppCMS apps. But this is
just for your motivation and hard work.
>
> CppCMS is finest out there. Make it better.
>
> --
> Respect,
> Shiv Shankar Dayal
>
>
------------------------------------------------------------------------------
> Announcing the Oxford Dictionaries API! The API offers world-renowned
> dictionary content that is easy and intuitive to access. Sign up for an
> account today to start using our lexical data to power your apps and
> projects. Get started today and enter our developer competition.
> http://sdm.link/oxford <http://sdm.link/oxford>
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li... <Cpp...@li...>
> https://lists.sourceforge.net/lists/listinfo/cppcms-users
<https://lists.sourceforge.net/lists/listinfo/cppcms-users>
>

[Cppcms-users] CppDB on github

[Shiv S. D. <shi...@gm...>]

Artyom,

Can you please being CppDB on github as well? Mighty thanks.

I do not have Paypal login. I tried to donate but ... I will make login and
do some donation for your hard work.

Please note that I have never made a penny with CppCMS apps. But this is
just for your motivation and hard work.

CppCMS is finest out there. Make it better.

-- 
Respect,
Shiv Shankar Dayal

Re: [Cppcms-users] CppCMS code migrated to GitHub

[Nazım C. B. <naz...@ne...>]

Artyom,

Could you share your thoughts about the patch? If you think you may 
consider accepting the proposal, I accept to provide patch for other 
backends (if it's required).

Regards,
Nazim Can.

On 04/03/17 21:11, CN wrote:
> On Fri, Mar 3, 2017, at 03:43 PM, Nazım Can Bedir wrote:
>>
>> And about your patch, if you can share your patch in PostgreSQL 
>> backend; there is a chance that I may help you about the other backends.
>>
>
> Hi!
>
> Since it will take a while to see CppDB moved to Github according to 
> Artyom, I can't wait to send the patch to Sourceforge. Hopefully it 
> will be accepted and then someone else will take over the rest of the 
> dirty half done job :-)
>
> Best Regards,
> CN
> -- 
> http://www.fastmail.com - A no graphics, no pop-ups email service
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>
>
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users

Re: [Cppcms-users] CppCMS code migrated to GitHub

[CN <cn...@fa...>]

On Fri, Mar 3, 2017, at 03:43 PM, Nazım Can Bedir wrote:

> And about your patch, if you can share your patch in PostgreSQL
> backend; there is a chance that I may help you about the other
> backends.


Hi!



Since it will take a while to see CppDB moved to Github according to
Artyom, I can't wait to send the patch to Sourceforge. Hopefully it will
be accepted and then someone else will take over the rest of the dirty
half done job :-)


Best Regards,

CN

-- 
http://www.fastmail.com - A no graphics, no pop-ups email service

Re: [Cppcms-users] CppCMS code migrated to GitHub

[Artyom B. <art...@gm...>]

Nothing,

I accepted  patch files without any problems :-)
Just svn become less popular in comparison to GIT and...

I had my own reasons unrelated to community requests or to quality of
svn vs git.

Artyom

On Thu, Mar 2, 2017 at 6:48 PM, Nazım Can Bedir
<naz...@ne...> wrote:
> Not specific to your mail, but I just don't understand what was
> preventing people to send patch from Sourceforge? It was even possible
> to open/comment tickets on Sourceforge, but not possible with GitHub
> (which one I am not going to create an account in any way).
>
>
> On 02/03/17 19:42, CN wrote:
>> On Wed, Mar 1, 2017, at 03:18 PM, Artyom Beilis wrote:
>>> After multiple requests and my final decision the CppCMS web framework
>>> code migrated to GitHub
>>>
>>> https://github.com/artyom-beilis/cppcms
>>>
>> I voted star just now.
>>
>>> 1. Only CppCMS framework migrated, other subprojects like CppDB or
>>> Wikipp are still pending conversion
>> My local version of CppDB enabled with bool data type has been running
>> smoothly for some time without problem.  Once CppDB is also there and
>> after I have learned how to use GitHub, I will try to send a pull
>> request for this patch.
>>
>> However, there is one problem:  Because I do not have experience on any
>> back end except PostgreSQL, I can add this data type only to PostgreSQL.
>>   Such half done work will apparently break the project making it unable
>> to compile for other back ends.
>> Any thoughts are welcome!
>>
>> Best Regards,
>> CN
>>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users

Re: [Cppcms-users] CppCMS code migrated to GitHub

[Artyom B. <art...@gm...>]

Bool is tricky as it isn't native for many DBs.

In any case cppdb is still in svn. So it'll take some time.

You can always prepare a patch and I'll integrate it.

One thing - adding bool would break ABI so I need new to release
a new major version so I may also consider adding some more features currently
not present in CppDB - like C++11 support etc.

Artyom

> My local version of CppDB enabled with bool data type has been running
> smoothly for some time without problem.  Once CppDB is also there and
> after I have learned how to use GitHub, I will try to send a pull
> request for this patch.
>
> However, there is one problem:  Because I do not have experience on any
> back end except PostgreSQL, I can add this data type only to PostgreSQL.
>  Such half done work will apparently break the project making it unable
> to compile for other back ends.
> Any thoughts are welcome!
>
> Best Regards,
> CN
>
> --
> http://www.fastmail.com - Send your email first class
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users

Re: [Cppcms-users] FreeBSD 11+clang fails to build backtrace.cpp

[Nazım C. B. <naz...@ne...>]

Thank you very much.

Regards,
Nazim Can.


On 04/03/17 01:17, Artyom Beilis wrote:
> Fixed in master... now backtrace works in BSD and not just disabled as
> per your patch
>
> Artyom
>
>>> Hi Artyom,
>>>
>>> This looks like related with closed bug "#78 FreeBSD+clang w/ c++11 fails to
>>> build backtrace.cpp...".
>>>
>>> Operating system:
>>>
>>> $ uname -a
>>> FreeBSD web1.bhyve.local 11.0-RELEASE-p1 FreeBSD 11.0-RELEASE-p1 #0 r306420:
>>> Thu Sep 29 01:43:23 UTC 2016
>>> ro...@re...:/usr/obj/usr/src/sys/GENERIC  amd64
>>>
>>> Compiler:
>>>
>>> $ clang --version
>>> FreeBSD clang version 3.8.0 (tags/RELEASE_380/final 262564) (based on LLVM
>>> 3.8.0)
>>> Target: x86_64-unknown-freebsd11.0
>>> Thread model: posix
>>> InstalledDir: /usr/bin
>>>
>>> When trying to compile CppCMS on this system, compiler generates following
>>> errors:
>>>
>>> $ make
>>> [  0%] Building CXX object
>>> booster/CMakeFiles/booster.dir/lib/backtrace/src/backtrace.cpp.o
>>> /home/nazim/cppcms/booster/lib/backtrace/src/backtrace.cpp:51:16: error:
>>> functions that differ only in their return type cannot be overloaded
>>>          extern void * _Unwind_GetIP (void *);
>>>                 ~~~~~~ ^
>>> /usr/include/c++/v1/unwind-itanium.h:92:22: note: previous declaration is
>>> here
>>> extern unsigned long _Unwind_GetIP (struct _Unwind_Context *);
>>>         ~~~~~~~~      ^
>>> /home/nazim/cppcms/booster/lib/backtrace/src/backtrace.cpp:80:48: error: no
>>> matching function for call to '_Unwind_GetIP'
>>>                          d->array[d->total++] = _Unwind_GetIP(context);
>>>                                                 ^~~~~~~~~~~~~
>>> /usr/include/c++/v1/unwind-itanium.h:92:22: note: candidate function not
>>> viable: cannot convert argument of incomplete type 'void *' to
>>>        'struct _Unwind_Context *'
>>> extern unsigned long _Unwind_GetIP (struct _Unwind_Context *);
>>>                       ^
>>> 2 errors generated.
>>> *** Error code 1
>>>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users

Re: [Cppcms-users] FreeBSD 11+clang fails to build backtrace.cpp

[Artyom B. <art...@gm...>]

Fixed in master... now backtrace works in BSD and not just disabled as
per your patch

Artyom

>> Hi Artyom,
>>
>> This looks like related with closed bug "#78 FreeBSD+clang w/ c++11 fails to
>> build backtrace.cpp...".
>>
>> Operating system:
>>
>> $ uname -a
>> FreeBSD web1.bhyve.local 11.0-RELEASE-p1 FreeBSD 11.0-RELEASE-p1 #0 r306420:
>> Thu Sep 29 01:43:23 UTC 2016
>> ro...@re...:/usr/obj/usr/src/sys/GENERIC  amd64
>>
>> Compiler:
>>
>> $ clang --version
>> FreeBSD clang version 3.8.0 (tags/RELEASE_380/final 262564) (based on LLVM
>> 3.8.0)
>> Target: x86_64-unknown-freebsd11.0
>> Thread model: posix
>> InstalledDir: /usr/bin
>>
>> When trying to compile CppCMS on this system, compiler generates following
>> errors:
>>
>> $ make
>> [  0%] Building CXX object
>> booster/CMakeFiles/booster.dir/lib/backtrace/src/backtrace.cpp.o
>> /home/nazim/cppcms/booster/lib/backtrace/src/backtrace.cpp:51:16: error:
>> functions that differ only in their return type cannot be overloaded
>>         extern void * _Unwind_GetIP (void *);
>>                ~~~~~~ ^
>> /usr/include/c++/v1/unwind-itanium.h:92:22: note: previous declaration is
>> here
>> extern unsigned long _Unwind_GetIP (struct _Unwind_Context *);
>>        ~~~~~~~~      ^
>> /home/nazim/cppcms/booster/lib/backtrace/src/backtrace.cpp:80:48: error: no
>> matching function for call to '_Unwind_GetIP'
>>                         d->array[d->total++] = _Unwind_GetIP(context);
>>                                                ^~~~~~~~~~~~~
>> /usr/include/c++/v1/unwind-itanium.h:92:22: note: candidate function not
>> viable: cannot convert argument of incomplete type 'void *' to
>>       'struct _Unwind_Context *'
>> extern unsigned long _Unwind_GetIP (struct _Unwind_Context *);
>>                      ^
>> 2 errors generated.
>> *** Error code 1
>>

Re: [Cppcms-users] CppCMS code migrated to GitHub

[Nazım C. B. <naz...@ne...>]

Hi,

Well I am not from that far, mine was 3.5" :)) I used 5.25" ones only in 
high-school computer lab.

If you can use Facebook, Twitter, or any other social networking 
website; then it's not too much different from those. In normal cases, 
90% of coding is done with your own development environment; and the web 
interface is mostly used for social interaction with other developers 
and end-users (like issue management, code review, brainstorming, 
visually navigating Git history, etc.).

Best wishes,
Nazim Can.


On 03/03/17 12:54, CN wrote:
> On Fri, Mar 3, 2017, at 03:43 PM, Nazım Can Bedir wrote:
>> About the GitHub,
>>
>> This is probably related with my age [I am from floppy era, :)] and 
>> my political view (which is very close to BSD philosophy); but I 
>> don't like GitHub because,
>
> Hi! Interesting! What were the dimension of those *diskettes* you used 
> to have? Mine were 5.25" in diameter. I throw them into trash 
> collector about 3 years ago, at last. That action was really a big 
> relief in terms of both living space and mental state.  However, I 
> still keep several 3.5" ones for no obvious reason.
>
>> But I had first-hand experiences with on-premise GitLab installation 
>> (both Community and Enterprise Edition). And if you are using Git 
>> DVCS as a primary development tool, I can really recommend GitLab. At 
>> the worst case, you have an option to reach Dimitri (the original 
>> creator and current CTO of GitLab) via online channels and get a 
>> reply (as we can do same with Artyom).
>>
>
> Thanks for sharing the precious experience!
> Coincidentally I registered an account in Gitlab 2 days ago. It looks 
> to me that I will have to learn both Github and Gitlab. I have a 
> feeling that it is more or less inhuman for an old man to suffer 
> torture like that.
>
> Best Regards,
> CN
> -- 
> http://www.fastmail.com - Send your email first class
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>
>
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users

Re: [Cppcms-users] CppCMS code migrated to GitHub

[CN <cn...@fa...>]

On Fri, Mar 3, 2017, at 03:43 PM, Nazım Can Bedir wrote:

> About the GitHub,

> 

>  This is probably related with my age [I am from floppy era, :)] and
>  my political view (which is very close to BSD philosophy); but I
>  don't like GitHub because,


Hi! Interesting! What were the dimension of those *diskettes* you used
to have? Mine were 5.25" in diameter. I throw them into trash collector
about 3 years ago, at last. That action was really a big relief in terms
of both living space and mental state.  However, I still keep several
3.5" ones for no obvious reason.


> But I had first-hand experiences with on-premise GitLab installation
> (both Community and Enterprise Edition). And if you are using Git DVCS
> as a primary development tool, I can really recommend GitLab. At the
> worst case, you have an option to reach Dimitri (the original creator
> and current CTO of GitLab) via online channels and get a reply (as we
> can do same with Artyom).


Thanks for sharing the precious experience!

Coincidentally I registered an account in Gitlab 2 days ago. It looks
to me that I will have to learn both Github and Gitlab. I have a
feeling that it is more or less inhuman for an old man to suffer
torture like that.


Best Regards,

CN

-- 
http://www.fastmail.com - Send your email first class

Re: [Cppcms-users] CppCMS code migrated to GitHub

[CN <cn...@fa...>]

On Fri, Mar 3, 2017, at 12:48 AM, Nazım Can Bedir wrote:
> Not specific to your mail, but I just don't understand what was 
> preventing people to send patch from Sourceforge? It was even possible 
> to open/comment tickets on Sourceforge, but not possible with GitHub 
> (which one I am not going to create an account in any way).

I speak only for myself.
Sourceforge does not in any way prevent me from sending my patch. I have
not done it only because:
1. My patch is only a minor improvement.
2. As mentioned in my previous message, my half done patch makes me
hesitate to show it under sunlight.

My real intention for sending out previous message has two fold:

1.  to solicit this community to work with me to complete this patch, or
even better:
2.  to encourage someone more competent than myself to complete the task
by himself/herself alone.

By the way, please pardon me for derail the topic!
Nowadays perhaps dozens of "cutting edge" new services and technologies
emerge almost every day.  As a result, considering my age, I am often
reluctant to learn so many new stuffs.  However, I am just curious what
makes you not to create an account in GitHub? Have you seen some
weakness in its services? In fact I am studying the possibilities of
leveraging Gitlab's services for myself.

Best Regards,
CN

-- 
http://www.fastmail.com - The professional email service

Re: [Cppcms-users] CppCMS code migrated to GitHub

[Nazım C. B. <naz...@ne...>]

Not specific to your mail, but I just don't understand what was 
preventing people to send patch from Sourceforge? It was even possible 
to open/comment tickets on Sourceforge, but not possible with GitHub 
(which one I am not going to create an account in any way).


On 02/03/17 19:42, CN wrote:
> On Wed, Mar 1, 2017, at 03:18 PM, Artyom Beilis wrote:
>> After multiple requests and my final decision the CppCMS web framework
>> code migrated to GitHub
>>
>> https://github.com/artyom-beilis/cppcms
>>
> I voted star just now.
>
>> 1. Only CppCMS framework migrated, other subprojects like CppDB or
>> Wikipp are still pending conversion
> My local version of CppDB enabled with bool data type has been running
> smoothly for some time without problem.  Once CppDB is also there and
> after I have learned how to use GitHub, I will try to send a pull
> request for this patch.
>
> However, there is one problem:  Because I do not have experience on any
> back end except PostgreSQL, I can add this data type only to PostgreSQL.
>   Such half done work will apparently break the project making it unable
> to compile for other back ends.
> Any thoughts are welcome!
>
> Best Regards,
> CN
>

Re: [Cppcms-users] CppCMS code migrated to GitHub

[CN <cn...@fa...>]

On Wed, Mar 1, 2017, at 03:18 PM, Artyom Beilis wrote:
> After multiple requests and my final decision the CppCMS web framework
> code migrated to GitHub
> 
> https://github.com/artyom-beilis/cppcms
> 

I voted star just now.

> 1. Only CppCMS framework migrated, other subprojects like CppDB or
> Wikipp are still pending conversion

My local version of CppDB enabled with bool data type has been running
smoothly for some time without problem.  Once CppDB is also there and
after I have learned how to use GitHub, I will try to send a pull
request for this patch.

However, there is one problem:  Because I do not have experience on any
back end except PostgreSQL, I can add this data type only to PostgreSQL.
 Such half done work will apparently break the project making it unable
to compile for other back ends.
Any thoughts are welcome!

Best Regards,
CN

-- 
http://www.fastmail.com - Send your email first class

Re: [Cppcms-users] CppCMS code migrated to GitHub

[Hery M. <hma...@gm...>]

Yes and thanks a lot.

2017-03-01 10:18 GMT+03:00 Artyom Beilis <art...@gm...>:

> After multiple requests and my final decision the CppCMS web framework
> code migrated to GitHub
>
> https://github.com/artyom-beilis/cppcms
>
> Please note:
>
> 1. Only CppCMS framework migrated, other subprojects like CppDB or
> Wikipp are still pending conversion
> 2. The main bug tracker is still on source-forge - however I'll relate
> to issues opened on GitHub
>
> Regards,
>   Artyom Beilis
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>

Re: [Cppcms-users] CppCMS code migrated to GitHub

[Shiv S. D. <shi...@gm...>]

Yes!!!

On Wed, Mar 1, 2017 at 12:48 PM, Artyom Beilis <art...@gm...>
wrote:

> After multiple requests and my final decision the CppCMS web framework
> code migrated to GitHub
>
> https://github.com/artyom-beilis/cppcms
>
> Please note:
>
> 1. Only CppCMS framework migrated, other subprojects like CppDB or
> Wikipp are still pending conversion
> 2. The main bug tracker is still on source-forge - however I'll relate
> to issues opened on GitHub
>
> Regards,
>   Artyom Beilis
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>



-- 
Respect,
Shiv Shankar Dayal

[Cppcms-users] CppCMS code migrated to GitHub

[Artyom B. <art...@gm...>]

After multiple requests and my final decision the CppCMS web framework
code migrated to GitHub

https://github.com/artyom-beilis/cppcms

Please note:

1. Only CppCMS framework migrated, other subprojects like CppDB or
Wikipp are still pending conversion
2. The main bug tracker is still on source-forge - however I'll relate
to issues opened on GitHub

Regards,
  Artyom Beilis

Re: [Cppcms-users] [BUG] Custom 'syslog.id' and 'openlog'

[Nazım C. B. <naz...@ne...>]

Hi,

For the bug report, please see #148 Custom 'syslog.id' and openlog 
(https://sourceforge.net/p/cppcms/bugs/148).

And while you are looking at this, could you please look at this small 
feature request:
#37 Custom syslog Facility Support 
(https://sourceforge.net/p/cppcms/feature-requests/37)

Thanks in advance.

Regards,
Nazim Can.


On 22/02/17 09:42, Artyom Beilis wrote:
> Ok, Thanks for the report!!!
>
> Can you please open the ticket?
>
> I'll fix it in trunk,
>
> Artyom
>
> On Tue, Feb 21, 2017 at 1:49 AM, Nazım Can Bedir
> <naz...@ne...> wrote:
>> Hi Artyom,
>>
>> For any reason, if user needs to specify custom syslog.id value; CppCMS is
>> calling 'openlog' function with id variable stored on the stack.
>>
>> According to
>>
>> man page at https://linux.die.net/man/3/openlog, and
>> this Debian bug report at
>> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=102350
>>
>> 'openlog' is free to use given 'ident' pointer as-is (without making a copy
>> of it). Even on Ubuntu 16.04 system, it's not copying given ident.
>>
>> In the following code excerpt, the pointer given at the line colored with
>> red will point to invalid memory location at the moment code reach to end of
>> active if block.
>>
>> service.cpp
>>
>>      if(settings.get("logging.syslog.enable",false)==true) {
>>          #ifndef CPPCMS_POSIX
>>              throw cppcms_error("Syslog is not availible on Windows");
>>          #else
>>          std::string id = settings.get("logging.syslog.id","");
>>          std::vector<std::string> vops =
>> settings.get("logging.syslog.options",std::vector<std::string>());
>>          std::string sfacility = settings.get("logging.syslog.options","");
>>          int ops = 0;
>>          for(unsigned i=0;i<vops.size();i++) {
>>              std::string const &op=vops[i];
>>              if(op=="LOG_CONS") ops|=LOG_CONS;
>>              else if(op=="LOG_NDELAY") ops|=LOG_NDELAY;
>>              else if(op=="LOG_NOWAIT") ops|=LOG_NOWAIT;
>>              else if(op=="LOG_ODELAY") ops|=LOG_ODELAY;
>>              #ifdef LOG_PERROR
>>              else if(op=="LOG_PERROR") ops|=LOG_PERROR;
>>              #endif
>>              else if(op=="LOG_PID") ops|=LOG_PID;
>>          }
>>          if(id.empty())
>>              ::openlog(0,ops,0);
>>          else
>>              ::openlog(id.c_str(),ops,0);
>>          logger::instance().add_sink(booster::shared_ptr<sink>(new
>> sinks::syslog()));
>>          #endif
>>      }
>>
>>      std::string log_file;
>>
>> Regards,
>> Nazim Can.
>>
>> ------------------------------------------------------------------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> _______________________________________________
>> Cppcms-users mailing list
>> Cpp...@li...
>> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users

Re: [Cppcms-users] FreeBSD 11+clang fails to build backtrace.cpp

[Artyom B. <art...@gm...>]

Ok thanks...

I'll apply the patch

Artyom

On Sat, Feb 11, 2017 at 9:02 PM, Nazım Can Bedir
<naz...@ne...> wrote:
> Hi Artyom,
>
> This looks like related with closed bug "#78 FreeBSD+clang w/ c++11 fails to
> build backtrace.cpp...".
>
> Operating system:
>
> $ uname -a
> FreeBSD web1.bhyve.local 11.0-RELEASE-p1 FreeBSD 11.0-RELEASE-p1 #0 r306420:
> Thu Sep 29 01:43:23 UTC 2016
> ro...@re...:/usr/obj/usr/src/sys/GENERIC  amd64
>
> Compiler:
>
> $ clang --version
> FreeBSD clang version 3.8.0 (tags/RELEASE_380/final 262564) (based on LLVM
> 3.8.0)
> Target: x86_64-unknown-freebsd11.0
> Thread model: posix
> InstalledDir: /usr/bin
>
> When trying to compile CppCMS on this system, compiler generates following
> errors:
>
> $ make
> [  0%] Building CXX object
> booster/CMakeFiles/booster.dir/lib/backtrace/src/backtrace.cpp.o
> /home/nazim/cppcms/booster/lib/backtrace/src/backtrace.cpp:51:16: error:
> functions that differ only in their return type cannot be overloaded
>         extern void * _Unwind_GetIP (void *);
>                ~~~~~~ ^
> /usr/include/c++/v1/unwind-itanium.h:92:22: note: previous declaration is
> here
> extern unsigned long _Unwind_GetIP (struct _Unwind_Context *);
>        ~~~~~~~~      ^
> /home/nazim/cppcms/booster/lib/backtrace/src/backtrace.cpp:80:48: error: no
> matching function for call to '_Unwind_GetIP'
>                         d->array[d->total++] = _Unwind_GetIP(context);
>                                                ^~~~~~~~~~~~~
> /usr/include/c++/v1/unwind-itanium.h:92:22: note: candidate function not
> viable: cannot convert argument of incomplete type 'void *' to
>       'struct _Unwind_Context *'
> extern unsigned long _Unwind_GetIP (struct _Unwind_Context *);
>                      ^
> 2 errors generated.
> *** Error code 1
>
> Stop.
> make[2]: stopped in /usr/home/nazim/cppcms
> *** Error code 1
>
> Stop.
> make[1]: stopped in /usr/home/nazim/cppcms
> *** Error code 1
>
> Stop.
> make: stopped in /usr/home/nazim/cppcms/booster
>
> I already read the conversion on the bug #78.
> /usr/include/c++/v1/unwind-itanium.h is (indirectly) included by
> /usr/include/c++/v1/cxxabi.h. backtrace.cpp file includes cxxabi.h at the
> line #26, but feature testing code in CMakeLists.txt doesn't have such
> #include directive (which makes CMake to succeed the feature test, while the
> actual compilation fails).
>
> But, if we add following code to the booster's CMakeLists.txt, then the
> feature test fails and the compilation completes without error.
>
>   check_cxx_source_compiles(
>           "#include <execinfo.h>
>           int main() { backtrace(0,0); }"
>           BOOSTER_HAVE_EXECINFO)
>
>   check_cxx_source_compiles(
>           "#include <stdexcept>
>            #include <iostream>
>            #include <stdlib.h>
>            #include <string.h>
>            #include <sstream>
>            #include <typeinfo>
>            #include <new>
> +
> +          #if defined(__GNUC__)
> +          #include <cxxabi.h>
> +          #endif
> +
>             extern \"C\" {
>              extern void* _Unwind_GetIP (void *);
>              extern int _Unwind_Backtrace(int (*)(void *,void *),void *);
>            }
>
>           int main() { _Unwind_Backtrace(0,0); _Unwind_GetIP(0); }"
>           BOOSTER_HAVE_UNWIND_BACKTRACE)
>
>
>   Check_cxx_source_compiles(
>           "#include <atomic.h>
>
>
> I am also sending output of 'svn diff' as attachment, for anyone who may
> experience same issue on the future.
>
> Regards,
> Nazim Can.
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>

Re: [Cppcms-users] [BUG] Custom 'syslog.id' and 'openlog'

[Artyom B. <art...@gm...>]

Ok, Thanks for the report!!!

Can you please open the ticket?

I'll fix it in trunk,

Artyom

On Tue, Feb 21, 2017 at 1:49 AM, Nazım Can Bedir
<naz...@ne...> wrote:
> Hi Artyom,
>
> For any reason, if user needs to specify custom syslog.id value; CppCMS is
> calling 'openlog' function with id variable stored on the stack.
>
> According to
>
> man page at https://linux.die.net/man/3/openlog, and
> this Debian bug report at
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=102350
>
> 'openlog' is free to use given 'ident' pointer as-is (without making a copy
> of it). Even on Ubuntu 16.04 system, it's not copying given ident.
>
> In the following code excerpt, the pointer given at the line colored with
> red will point to invalid memory location at the moment code reach to end of
> active if block.
>
> service.cpp
>
>     if(settings.get("logging.syslog.enable",false)==true) {
>         #ifndef CPPCMS_POSIX
>             throw cppcms_error("Syslog is not availible on Windows");
>         #else
>         std::string id = settings.get("logging.syslog.id","");
>         std::vector<std::string> vops =
> settings.get("logging.syslog.options",std::vector<std::string>());
>         std::string sfacility = settings.get("logging.syslog.options","");
>         int ops = 0;
>         for(unsigned i=0;i<vops.size();i++) {
>             std::string const &op=vops[i];
>             if(op=="LOG_CONS") ops|=LOG_CONS;
>             else if(op=="LOG_NDELAY") ops|=LOG_NDELAY;
>             else if(op=="LOG_NOWAIT") ops|=LOG_NOWAIT;
>             else if(op=="LOG_ODELAY") ops|=LOG_ODELAY;
>             #ifdef LOG_PERROR
>             else if(op=="LOG_PERROR") ops|=LOG_PERROR;
>             #endif
>             else if(op=="LOG_PID") ops|=LOG_PID;
>         }
>         if(id.empty())
>             ::openlog(0,ops,0);
>         else
>             ::openlog(id.c_str(),ops,0);
>         logger::instance().add_sink(booster::shared_ptr<sink>(new
> sinks::syslog()));
>         #endif
>     }
>
>     std::string log_file;
>
> Regards,
> Nazim Can.
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>

[Cppcms-users] [BUG] Custom 'syslog.id' and 'openlog'

[Nazım C. B. <naz...@ne...>]

Hi Artyom,

For any reason, if user needs to specify custom syslog.id value; CppCMS 
is calling 'openlog' function with id variable stored on the stack.

According to

  * man page at https://linux.die.net/man/3/openlog, and
  * this Debian bug report at
    https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=102350

'openlog' is free to use given 'ident' pointer as-is (without making a 
copy of it). Even on Ubuntu 16.04 system, it's not copying given ident.

In the following code excerpt, the pointer given at the line colored 
with red will point to invalid memory location at the moment code reach 
to end of active if block.

_service.cpp_

     if(settings.get("logging.syslog.enable",false)==true) {
         #ifndef CPPCMS_POSIX
             throw cppcms_error("Syslog is not availible on Windows");
         #else
         std::string id = settings.get("logging.syslog.id","");
         std::vector<std::string> vops = 
settings.get("logging.syslog.options",std::vector<std::string>());
         std::string sfacility = settings.get("logging.syslog.options","");
         int ops = 0;
         for(unsigned i=0;i<vops.size();i++) {
             std::string const &op=vops[i];
             if(op=="LOG_CONS") ops|=LOG_CONS;
             else if(op=="LOG_NDELAY") ops|=LOG_NDELAY;
             else if(op=="LOG_NOWAIT") ops|=LOG_NOWAIT;
             else if(op=="LOG_ODELAY") ops|=LOG_ODELAY;
             #ifdef LOG_PERROR
             else if(op=="LOG_PERROR") ops|=LOG_PERROR;
             #endif
             else if(op=="LOG_PID") ops|=LOG_PID;
         }
         if(id.empty())
             ::openlog(0,ops,0);
         else
             ::openlog(id.c_str(),ops,0);
logger::instance().add_sink(booster::shared_ptr<sink>(new sinks::syslog()));
         #endif
     }

     std::string log_file;

Regards,
Nazim Can.

Re: [Cppcms-users] 404 Not Found even with valid URLs

[Nazım C. B. <naz...@ne...>]

Hello,

There is inconsistency in your dispatcher and mapper code. As far as I 
understand, you would like to use your URLs like "/students/list", 
"/students/add", etc. For this to work, you have two different approaches;

  * You need to mount your application with "/students" prefix. If you
    choose this way, then you don't need to set mapper root, and can
    dispatch and map as "/add", "/list", "/show", etc.
  * You need to maintain prefix manually, and 1) set mapper root to that
    prefix, 2) prefix all of your dispatch rules appropriately.

In your code, you are mounting your application to "/", and setting 
mapper root to "/students" but your dispatch codes doesn't have 
"/students" prefix.

This is your code:

           }

         dispatcher().assign("/add", &students::add, this);
         mapper().assign("add", "/add");

         std::cout << "Added add script" << std::endl;

         dispatcher().assign("", &students::list, this);
         mapper().assign("");

         dispatcher().assign("/list", &students::list, this);
         mapper().assign("list", "/list");

         dispatcher().assign("/show/(\\d)+", &students::show, this, 1);
         mapper().assign("show", "/show/{1}");

         dispatcher().assign("/update/(\\d)+", &students::update, this, 1);
         mapper().assign("update", "/update/{1}");

         mapper().root("/students");
       }

This is how it should look:

           }

         dispatcher().assign("/students/add", &students::add, this);
         mapper().assign("add", "/add");

         std::cout << "Added add script" << std::endl;

         dispatcher().assign("/students", &students::list, this);
         mapper().assign("/students");

         dispatcher().assign("/students/list", &students::list, this);
         mapper().assign("list", "/list");

         dispatcher().assign("/students/show/(\\d)+", &students::show,
    this, 1);
         mapper().assign("show", "/show/{1}");

         dispatcher().assign("/students/update/(\\d)+",
    &students::update, this, 1);
         mapper().assign("update", "/update/{1}");

         mapper().root("/students");
       }

After making changes, I am able to visit addresses and get output from 
the console.

Regards,
Nazim Can.

On 18/02/17 14:52, abi...@ic... wrote:
> Here’s the project. The dispatcher and mapper code is in the root 
> directory in the students.cpp file. 
> http://github.com/abbygriffiths/cppcms-student_record.git
>
> /Abirbhav Goswami/
>
> On Feb 18, 2017, 5:03 PM +0530, Nazım Can Bedir 
> <naz...@ne...>, wrote:
>>
>> Can you share the code which one is believed to be correct?
>>
>>
>> On 18/02/17 14:07, abi...@ic... wrote:
>>> I made a small website and it’s giving me a 404 – Not Found error, 
>>> but I’m confident that the code isn’t wrong. Can someone point me in 
>>> the right direction?
>>>
>>> /Abirbhav Goswami/
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Check out the vibrant tech community on one of the world's most
>>> engaging tech sites, SlashDot.org!http://sdm.link/slashdot
>>>
>>>
>>> _______________________________________________
>>> Cppcms-users mailing list
>>> Cpp...@li...
>>> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>>
>> ------------------------------------------------------------------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> _______________________________________________
>> Cppcms-users mailing list
>> Cpp...@li...
>> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>
>
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users

Re: [Cppcms-users] 404 Not Found even with valid URLs

[<abi...@ic...>]

Here’s the project. The dispatcher and mapper code is in the root directory in the students.cpp file. http://github.com/abbygriffiths/cppcms-student_record.git

Abirbhav Goswami

On Feb 18, 2017, 5:03 PM +0530, Nazım Can Bedir <naz...@ne...>, wrote:
> Can you share the code which one is believed to be correct?
>
> On 18/02/17 14:07, abi...@ic... wrote:
> > I made a small website and it’s giving me a 404 – Not Found error, but I’m confident that the code isn’t wrong. Can someone point me in the right direction?
> >
> > Abirbhav Goswami
> >
> >
> >
> > ------------------------------------------------------------------------------
> > Check out the vibrant tech community on one of the world's most
> > engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> >
> >
> >
> > _______________________________________________
> > Cppcms-users mailing list
> > Cpp...@li...
> > https://lists.sourceforge.net/lists/listinfo/cppcms-users
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users

Re: [Cppcms-users] 404 Not Found even with valid URLs

[Nazım C. B. <naz...@ne...>]

Can you share the code which one is believed to be correct?


On 18/02/17 14:07, abi...@ic... wrote:
> I made a small website and it’s giving me a 404 – Not Found error, but 
> I’m confident that the code isn’t wrong. Can someone point me in the 
> right direction?
>
> /Abirbhav Goswami/
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>
>
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users

[Cppcms-users] 404 Not Found even with valid URLs

[<abi...@ic...>]

I made a small website and it’s giving me a 404 – Not Found error, but I’m confident that the code isn’t wrong. Can someone point me in the right direction?

Abirbhav Goswami

[Cppcms-users] FreeBSD 11+clang fails to build backtrace.cpp

[Nazım C. B. <naz...@ne...>]

Hi Artyom,

This looks like related with closed bug "#78 FreeBSD+clang w/ c++11 
fails to build backtrace.cpp...".

Operating system:

    $ uname -a
    FreeBSD web1.bhyve.local 11.0-RELEASE-p1 FreeBSD 11.0-RELEASE-p1 #0 r306420: Thu Sep 29 01:43:23 UTC 2016     ro...@re...:/usr/obj/usr/src/sys/GENERIC  amd64

Compiler:

    $ clang --version
    FreeBSD clang version 3.8.0 (tags/RELEASE_380/final 262564) (based on LLVM 3.8.0)
    Target: x86_64-unknown-freebsd11.0
    Thread model: posix
    InstalledDir: /usr/bin

When trying to compile CppCMS on this system, compiler generates 
following errors:

    $ make
    [  0%] Building CXX object booster/CMakeFiles/booster.dir/lib/backtrace/src/backtrace.cpp.o
    /home/nazim/cppcms/booster/lib/backtrace/src/backtrace.cpp:51:16: error: functions that differ only in their return type cannot be overloaded
             extern void * _Unwind_GetIP (void *);
                    ~~~~~~ ^
    /usr/include/c++/v1/unwind-itanium.h:92:22: note: previous declaration is here
    extern unsigned long _Unwind_GetIP (struct _Unwind_Context *);
            ~~~~~~~~      ^
    /home/nazim/cppcms/booster/lib/backtrace/src/backtrace.cpp:80:48: error: no matching function for call to '_Unwind_GetIP'
                             d->array[d->total++] = _Unwind_GetIP(context);
                                                    ^~~~~~~~~~~~~
    /usr/include/c++/v1/unwind-itanium.h:92:22: note: candidate function not viable: cannot convert argument of incomplete type 'void *' to
           'struct _Unwind_Context *'
    extern unsigned long _Unwind_GetIP (struct _Unwind_Context *);
                          ^
    2 errors generated.
    *** Error code 1

    Stop.
    make[2]: stopped in /usr/home/nazim/cppcms
    *** Error code 1

    Stop.
    make[1]: stopped in /usr/home/nazim/cppcms
    *** Error code 1

    Stop.
    make: stopped in /usr/home/nazim/cppcms/booster

I already read the conversion on the bug #78. 
/usr/include/c++/v1/unwind-itanium.h is (indirectly) included by 
/usr/include/c++/v1/cxxabi.h. backtrace.cpp file includes cxxabi.h at 
the line #26, but feature testing code in CMakeLists.txt doesn't have 
such #include directive (which makes CMake to succeed the feature test, 
while the actual compilation fails).

But, if we add following code to the booster's CMakeLists.txt, then the 
feature test fails and the compilation completes without error.

       check_cxx_source_compiles(
               "#include <execinfo.h>
               int main() { backtrace(0,0); }"
               BOOSTER_HAVE_EXECINFO)

       check_cxx_source_compiles(
               "#include <stdexcept>
                #include <iostream>
                #include <stdlib.h>
                #include <string.h>
                #include <sstream>
                #include <typeinfo>
                #include <new>
    + + #if defined(__GNUC__) + #include <cxxabi.h> + #endif +
                 extern \"C\" {
                  extern void* _Unwind_GetIP (void *);
                  extern int _Unwind_Backtrace(int (*)(void *,void *),void *);
                }

               int main() { _Unwind_Backtrace(0,0); _Unwind_GetIP(0); }"
               BOOSTER_HAVE_UNWIND_BACKTRACE)


       Check_cxx_source_compiles(
               "#include <atomic.h>


I am also sending output of 'svn diff' as attachment, for anyone who may 
experience same issue on the future.

Regards,
Nazim Can.