From cezanmeyen at gmail.com  Fri Apr  5 09:44:24 2019
From: cezanmeyen at gmail.com (Cezan Meyen)
Date: Fri Apr  5 10:02:45 2019
Subject: [odb-users] Unable to map a member of a member via #pragma
Message-ID: <CACZnhn25BO5qBP+TDvgV8ChChMbsK8XB3bUoNM21Uf-KYn5B-w@mail.gmail.com>

Hello there,

I am currently implementing ODB in a existing program and therefore I am
unable to change the models. And it is working great so far but I am
currently  stuck on a single object member. I need in some way access to
the member or a member via the pragma's. I don?t know if this is possible
at all but I cannot find a way to map this member. I might be doing
something completly wrong. I some posted example code underneath here for
some context.

template<size_t max_length>
struct FixString
{
    public:
        sampleget();
        sampleSet();
    private
        char m_str[max_length + 1];

}
typedef FixString<24> name_string;



struct string_id
{
    public:
    FixString<24> uuid;
}
typedef string_id ID;


struct main
{
    name_string name;
    ID id;
}

//pragmas contained in a different file
#pragma db object(main)

#pragma db value(name_string)
#pragma db member(name_string::m_str) get(sampleget()) set(sampleSet())
// this works okay!

#pragma db value(ID)
#pragma db member(ID::uuid)                           //unable to map C++
type 'FixString<24>'   <--- expected error
#pragma db member(ID::uuid::m_str)                //unable to resolve type
name '::ID::uuid' in db pragma member
I can't seem to access m_str form ID which is necessary to define the get
and set.

Kind regards,
Cezan
From boris at codesynthesis.com  Mon Apr  8 07:19:55 2019
From: boris at codesynthesis.com (Boris Kolpackov)
Date: Mon Apr  8 07:38:08 2019
Subject: [odb-users] Unable to map a member of a member via #pragma
In-Reply-To: <CACZnhn25BO5qBP+TDvgV8ChChMbsK8XB3bUoNM21Uf-KYn5B-w@mail.gmail.com>
References: <CACZnhn25BO5qBP+TDvgV8ChChMbsK8XB3bUoNM21Uf-KYn5B-w@mail.gmail.com>
Message-ID: <boris.20190408131602@codesynthesis.com>

Cezan Meyen <cezanmeyen@gmail.com> writes:

> typedef FixString<24> name_string;
> 
> struct string_id
> {
>     public:
>     FixString<24> uuid;
> }
> typedef string_id ID;
> 
> 
> //pragmas contained in a different file
> #pragma db object(main)
> 
> #pragma db value(name_string)
> #pragma db member(name_string::m_str) get(sampleget()) set(sampleSet())
> // this works okay!
> 
> #pragma db value(ID)
> #pragma db member(ID::uuid)                           //unable to map C++
> type 'FixString<24>'   <--- expected error

You should use name_string (or some other named and mapped value type)
for string_id::uuid. Or, better, yet, map FixString<24> as described
here:

https://www.codesynthesis.com/~boris/blog/2012/10/16/custom-cxx-to-database-type-mapping-in-odb/

Or, if you are using the 2.5.0 pre-release, using '#pragma db map'.

From lfiedler at informatik.uni-leipzig.de  Thu Apr 18 07:29:48 2019
From: lfiedler at informatik.uni-leipzig.de (Lisa Fiedler)
Date: Thu Apr 18 07:48:48 2019
Subject: [odb-users] Use of postgresql array mapping for native views
Message-ID: <adba4c2c-a621-09b4-d8fa-34906015b977@informatik.uni-leipzig.de>

Hello,

I was trying to use a postgresql array mapping to a vector within a 
native view.

That is I generated a file "traits-pgsql.hxx" with content:

#ifndef TRAITS_PGSQL_HXX #define TRAITS_PGSQL_HXX #include <cstddef> // 
std::size_t #include <cstring> // std::strncmp, std::memset, std::memcpy 
#include <iostream> #include <fstream> #include <odb/pgsql/traits.hxx> 
#include <odb/pgsql/details/endian-traits.hxx> #include 
"../configdb.hpp" #include <ctime>?? // localtime, mktime, time_t, tm 
#include <sstream> namespace odb { ? namespace pgsql ? { ??? template <> 
 ??? class value_traits<std::vector<u_int64_t>, id_string> ??? { ??? 
public: ????? typedef std::vector<u_int64_t> value_type; ????? typedef 
value_type query_type; ????? typedef details::buffer image_type; ????? 
static void ????? set_value (value_type& v, ???????????????? const 
details::buffer& b, ???????????????? std::size_t n, ???????????????? 
bool is_null) ????? { ??????? v.clear (); ??????? if (!is_null) ??????? 
{ ????????? char c; ????????? std::istringstream is (std::string (b.data 
(), n)); ????????? is >> c; // '{' ????????? for (c = static_cast<char> 
(is.peek ()); c != '}'; is >> c) ????????? { ??????????? v.push_back 
(u_int64_t ()); ??????????? is >> v.back (); ????????? } ??????? } ????? 
} ????? static void ????? set_image (details::buffer& b, 
 ???????????????? std::size_t& n, ???????????????? bool& is_null, 
 ???????????????? const value_type& v) ????? { ??????? is_null = false; 
 ??????? std::ostringstream os; ??????? os << '{'; ??????? for 
(value_type::const_iterator i (v.begin ()), e (v.end ()); ???????????? i 
!= e;) ??????? { ????????? os << *i; ????????? if (++i != e) ??????????? 
os << ','; ??????? } ??????? os << '}'; ??????? const std::string& s 
(os.str ()); ??????? n = s.size (); ??????? if (n > b.capacity ()) 
 ????????? b.capacity (n); ??????? std::memcpy (b.data (), s.c_str (), 
n); ????? } ??? }; ? } } #endif // TRAITS_PGSQL_HXX

and in taxonomy.hxx, where I was planning to use this mapping I wrote:

typedef std::vector<u_int64_t> id_vector; #pragma db value(id_vector) 
type("BIGINT[]")

and tried to use this in a native query like this:

#pragma db view struct queryTaxonomyPath_view{ ??? #pragma db 
type("BIGINT[]") ??? id_vector path; };

The error message I get is:

./include/complete/taxonomy.hxx:241:15: error: invalid PostgreSQL type 
declaration: unexpected character '['

Makefile:60: recipe for target 'include/complete/taxonomy-odb.cxx' failed

make: *** [include/complete/taxonomy-odb.cxx] Error 1

make: *** Waiting for unfinished jobs....

./include/complete/taxonomy.hxx:241:15: error: invalid PostgreSQL type 
declaration: unexpected character '['

terminate called after throwing an instance of 'cutl::fs::error'

what(): filesystem error

*** WARNING *** there are active plugins, do not report this as a bug 
unless you can reproduce it without enabling any plugins.

Event | Plugins

PLUGIN_START_UNIT | odb

PLUGIN_PRAGMAS | odb

PLUGIN_OVERRIDE_GATE | odb

In file included from 
/usr/local/odb-2.5.0-b.13/libodb/include/odb/container-traits.hxx:217:0,

from <standard-odb-epilogue>:4:

/usr/local/odb-2.5.0-b.13/libodb/include/odb/std-unordered-set-traits.hxx: 
In function ?void __static_initialization_and_destruction_0(int, int)?:

/usr/local/odb-2.5.0-b.13/libodb/include/odb/std-unordered-set-traits.hxx:122:1: 
internal compiler error: Aborted

}

^

To generate the *odb.ixx and *odb.cxx files I used:

/usr/local/odb-2.5.0-b.13/odb/bin/odb -d pgsql 
-I/usr/local/odb-2.5.0-b.13/libodb/include --std c++11 --generate-query 
--generate-prepared --hxx-prologue "#include \"traits-pgsql.hxx\"" 
./include/complete/properties.hxx ./include/complete/structures.hxx 
./include/complete/taxonomy.hxx ./include/complete/edge.hxx 
./include/complete/record.hxx

If I remove the native view there are no error messages.

Thank you in advance!

From becoggins at hotmail.com  Sat Apr 20 12:20:58 2019
From: becoggins at hotmail.com (Brian Coggins)
Date: Sat Apr 20 12:39:58 2019
Subject: [odb-users] Testing whether lazy_shared_ptr is null
Message-ID: <BN8PR16MB2802110F0EF1A6DF2F69F759B0200@BN8PR16MB2802.namprd16.prod.outlook.com>

Imagine I have two types of object A and B with a unidirectional A to B relationship.  If I load an object A containing a lazy_shared_ptr< B >, what is the best way to determine whether this lazy_shared_ptr< B > represents a relationship to an instance of B (of any kind, transient or persistent, loaded or unloaded), or if it is null?

>From reading the docs, I think I could call lazy_shared_ptr< B >::load() and then test whether the returned shared_ptr< B > is nullptr.  However, if lazy_shared_ptr< B > points to a B, this has the side effect of loading the instance of B, and I don?t actually need the B.  I just need to know whether A has a relationship to B.  It seems like there must be a way to check this without loading the pointed-to object?

Of course, at the database level one could query for whether the field in A representing the relationship is NULL, but I was hoping to find a way to get this info from the C++ object.

Thanks,
Brian




From boris at codesynthesis.com  Mon Apr 22 12:02:43 2019
From: boris at codesynthesis.com (Boris Kolpackov)
Date: Mon Apr 22 12:21:43 2019
Subject: [odb-users] Use of postgresql array mapping for native views
In-Reply-To: <adba4c2c-a621-09b4-d8fa-34906015b977@informatik.uni-leipzig.de>
References: <adba4c2c-a621-09b4-d8fa-34906015b977@informatik.uni-leipzig.de>
Message-ID: <boris.20190422175950@codesynthesis.com>

Lisa Fiedler <lfiedler@informatik.uni-leipzig.de> writes:

> value_traits<std::vector<u_int64_t>, id_string>
>
> [...]
>
> typedef std::vector<u_int64_t> id_vector;
> #pragma db value(id_vector) type("BIGINT[]")

This doesn't make much sense to me: you are providing a mapping
traits for vector<u_int64_t> to string but then trying to use
BIGINT[] as the database type?

I suggest that you read through this article and see if it
clarifies things:

https://www.codesynthesis.com/~boris/blog/2012/07/18/custom-database-to-cxx-type-mapping-in-odb/

From boris at codesynthesis.com  Mon Apr 22 12:17:19 2019
From: boris at codesynthesis.com (Boris Kolpackov)
Date: Mon Apr 22 12:36:16 2019
Subject: [odb-users] Testing whether lazy_shared_ptr is null
In-Reply-To: <BN8PR16MB2802110F0EF1A6DF2F69F759B0200@BN8PR16MB2802.namprd16.prod.outlook.com>
References: <BN8PR16MB2802110F0EF1A6DF2F69F759B0200@BN8PR16MB2802.namprd16.prod.outlook.com>
Message-ID: <boris.20190422181016@codesynthesis.com>

Brian Coggins <becoggins@hotmail.com> writes:

> Imagine I have two types of object A and B with a unidirectional A to B
> relationship.  If I load an object A containing a lazy_shared_ptr< B >,
> what is the best way to determine whether this lazy_shared_ptr< B >
> represents a relationship to an instance of B (of any kind, transient
> or persistent, loaded or unloaded), or if it is null?

Hm, if you want to check whether the pointer is not NULL (i.e., either
contains an object id or points to a transient object), then that would
simply be:

if (p != nullptr)

If you want to know if the above mentioned object id actually references
something real in the database, then, naturally, the only way to answer
that question is to try to load it.

Also, just to add a point about weak pointers, to correctly test whether
it is NULL you need to call it's lock() function, get the strong pointer,
and then proceed as above.

From lfiedler at informatik.uni-leipzig.de  Tue Apr 23 03:05:40 2019
From: lfiedler at informatik.uni-leipzig.de (Lisa Fiedler)
Date: Tue Apr 23 03:24:54 2019
Subject: [odb-users] Use of postgresql array mapping for native views
In-Reply-To: <boris.20190422175950@codesynthesis.com>
References: <adba4c2c-a621-09b4-d8fa-34906015b977@informatik.uni-leipzig.de>
	<boris.20190422175950@codesynthesis.com>
Message-ID: <39c72fab-64f4-e9af-b252-37514454263c@informatik.uni-leipzig.de>


On 4/22/19 6:02 PM, Boris Kolpackov wrote:
> Lisa Fiedler <lfiedler@informatik.uni-leipzig.de> writes:
>
>> value_traits<std::vector<u_int64_t>, id_string>
>>
>> [...]
>>
>> typedef std::vector<u_int64_t> id_vector;
>> #pragma db value(id_vector) type("BIGINT[]")
> This doesn't make much sense to me: you are providing a mapping
> traits for vector<u_int64_t> to string but then trying to use
> BIGINT[] as the database type?

I agree. I am sorry.

But even after writing the following lines (note I am using vector<int> 
now to reproduce the instructions in the article):

#pragma db map type("BIGINT *\\[(\\d*)\\]") \
 ?????????????? as("TEXT")??????????????????? \
 ?????????????? to("(?)::BIGINT[$1]")??????? \
 ?????????????? from("(?)::TEXT")

typedef std::vector<int> id_vector;
#pragma db value(id_vector) type("BIGINT[]")

and then using precisely the same value_traits as in

https://www.codesynthesis.com/~boris/blog/2012/07/18/custom-database-to-cxx-type-mapping-in-odb/

and then trying to do

#pragma db view
struct queryTaxonomyPath_view{

 ??? #pragma db type("BIGINT[]")
 ??? id_vector path;
};

and

 ??? odb::result<queryTaxonomyPath_view> res
 ??????????? (db_->query<queryTaxonomyPath_view>("SELECT * FROM 
complete.generate_taxonomic_path(" +to_string(taxId) + "::bigint)"));
 ??? cout << "size: " << res.size() << endl;

 ??? for(auto & it : res){
 ??????? cout << it.path.size() << endl;

 ??? }

does compile, however never terminates upon execution. res.size() is 
still printed, but at the for loop the program gets stuck.

generate_taxonomic_path is a postgresql user defined function, which 
returns a table with an attribute of type BIGINT[].

I don't understand why this does not work out ?

>
> I suggest that you read through this article and see if it
> clarifies things:
>
> https://www.codesynthesis.com/~boris/blog/2012/07/18/custom-database-to-cxx-type-mapping-in-odb/

From lfiedler at informatik.uni-leipzig.de  Tue Apr 23 03:43:14 2019
From: lfiedler at informatik.uni-leipzig.de (Lisa Fiedler)
Date: Tue Apr 23 04:02:22 2019
Subject: [odb-users] Use of postgresql array mapping for native views
In-Reply-To: <39c72fab-64f4-e9af-b252-37514454263c@informatik.uni-leipzig.de>
References: <adba4c2c-a621-09b4-d8fa-34906015b977@informatik.uni-leipzig.de>
	<boris.20190422175950@codesynthesis.com>
	<39c72fab-64f4-e9af-b252-37514454263c@informatik.uni-leipzig.de>
Message-ID: <5f3d3e2b-6ba1-6bc6-1385-c7490acf832c@informatik.uni-leipzig.de>

Never mind, it works now. Thank you very much for pointing me to this 
article!!


On 4/23/19 9:05 AM, Lisa Fiedler wrote:
>
> On 4/22/19 6:02 PM, Boris Kolpackov wrote:
>> Lisa Fiedler <lfiedler@informatik.uni-leipzig.de> writes:
>>
>>> value_traits<std::vector<u_int64_t>, id_string>
>>>
>>> [...]
>>>
>>> typedef std::vector<u_int64_t> id_vector;
>>> #pragma db value(id_vector) type("BIGINT[]")
>> This doesn't make much sense to me: you are providing a mapping
>> traits for vector<u_int64_t> to string but then trying to use
>> BIGINT[] as the database type?
>
> I agree. I am sorry.
>
> But even after writing the following lines (note I am using 
> vector<int> now to reproduce the instructions in the article):
>
> #pragma db map type("BIGINT *\\[(\\d*)\\]") \
> ?????????????? as("TEXT")??????????????????? \
> ?????????????? to("(?)::BIGINT[$1]")??????? \
> ?????????????? from("(?)::TEXT")
>
> typedef std::vector<int> id_vector;
> #pragma db value(id_vector) type("BIGINT[]")
>
> and then using precisely the same value_traits as in
>
> https://www.codesynthesis.com/~boris/blog/2012/07/18/custom-database-to-cxx-type-mapping-in-odb/ 
>
>
> and then trying to do
>
> #pragma db view
> struct queryTaxonomyPath_view{
>
> ??? #pragma db type("BIGINT[]")
> ??? id_vector path;
> };
>
> and
>
> ??? odb::result<queryTaxonomyPath_view> res
> ??????????? (db_->query<queryTaxonomyPath_view>("SELECT * FROM 
> complete.generate_taxonomic_path(" +to_string(taxId) + "::bigint)"));
> ??? cout << "size: " << res.size() << endl;
>
> ??? for(auto & it : res){
> ??????? cout << it.path.size() << endl;
>
> ??? }
>
> does compile, however never terminates upon execution. res.size() is 
> still printed, but at the for loop the program gets stuck.
>
> generate_taxonomic_path is a postgresql user defined function, which 
> returns a table with an attribute of type BIGINT[].
>
> I don't understand why this does not work out ?
>
>>
>> I suggest that you read through this article and see if it
>> clarifies things:
>>
>> https://www.codesynthesis.com/~boris/blog/2012/07/18/custom-database-to-cxx-type-mapping-in-odb/ 
>>
>

From becoggins at hotmail.com  Tue Apr 23 08:58:57 2019
From: becoggins at hotmail.com (Brian Coggins)
Date: Tue Apr 23 09:18:06 2019
Subject: [odb-users] Testing whether lazy_shared_ptr is null
In-Reply-To: <boris.20190422181016@codesynthesis.com>
References: <BN8PR16MB2802110F0EF1A6DF2F69F759B0200@BN8PR16MB2802.namprd16.prod.outlook.com>
	<boris.20190422181016@codesynthesis.com>
Message-ID: <BN8PR16MB2802CA16D97E2311171C9C2EB0230@BN8PR16MB2802.namprd16.prod.outlook.com>

On Apr 22, 2019, at 12:17 PM, Boris Kolpackov <boris@codesynthesis.com> wrote:

> Also, just to add a point about weak pointers, to correctly test whether
> it is NULL you need to call it's lock() function, get the strong pointer,
> and then proceed as above.

Ah, my mistake!  This was the problem.  I have a number of lazy_shared_ptr and lazy_weak_ptr going in various directions in this schema, and the relationship I wanted to test was actually a lazy_weak_ptr.  I couldn?t figure out why it wouldn?t convert to bool, but it makes sense that it has to be locked first.

Thanks,
Brian

From noah.sk at protonmail.ch  Fri Apr 26 04:01:21 2019
From: noah.sk at protonmail.ch (noah.sk)
Date: Fri Apr 26 04:20:40 2019
Subject: [odb-users] CMake/autotools/meson support in release
Message-ID: <Hp7dewk9dUFb5Cqfbc36Ro1e0JT2T5W2O_wo6PkYQyufq6FwA60dVZxp8g5VBw4LXEqyTc7NyDmp4UP0J0-D8J1fOczPJfiSGvaJ68jSE54=@protonmail.ch>

Hi! We look to use ODB as C++ ORM but currently we have discovered lack of support widely used build systems in ODB beta and git versions.

Unfortunately, we are limited to use build tools like CMake, autotools or meson to make own build of ODB.

Do you plan to continue support only build2?

Is there any chance of adding CMake/autotools/maven support to ODB 2.5.0 release?

Thanks. Best regards, Noah.

From ps.georgiou at gmail.com  Fri Apr 26 04:36:27 2019
From: ps.georgiou at gmail.com (Panayiotis Georgiou)
Date: Fri Apr 26 04:55:56 2019
Subject: [odb-users] CMake/autotools/meson support in release
In-Reply-To: <Hp7dewk9dUFb5Cqfbc36Ro1e0JT2T5W2O_wo6PkYQyufq6FwA60dVZxp8g5VBw4LXEqyTc7NyDmp4UP0J0-D8J1fOczPJfiSGvaJ68jSE54=@protonmail.ch>
References: <Hp7dewk9dUFb5Cqfbc36Ro1e0JT2T5W2O_wo6PkYQyufq6FwA60dVZxp8g5VBw4LXEqyTc7NyDmp4UP0J0-D8J1fOczPJfiSGvaJ68jSE54=@protonmail.ch>
Message-ID: <CAJQ2NJQ5HUsBUDwQ3jidmZa4dbyVv-5SZrnj+MfC4ugKZVRXTQ@mail.gmail.com>

Hi,

Some time ago, I found the following repository:
https://github.com/BtbN/OdbCmake

It includes an example of CMakeLists.txt, and the useful modules
FindODB.cmake and UseODB.cmake.

I've been using it in my project with ODB 2.4.0 and Cmake with good
success. Not all the command line parameters of the ODB compiler are
supported, but if you have a look in UseODB.cmake, it's relatively easy to
add new ones.

P.

On Fri, Apr 26, 2019 at 9:01 AM noah.sk <noah.sk@protonmail.ch> wrote:

> Hi! We look to use ODB as C++ ORM but currently we have discovered lack of
> support widely used build systems in ODB beta and git versions.
>
> Unfortunately, we are limited to use build tools like CMake, autotools or
> meson to make own build of ODB.
>
> Do you plan to continue support only build2?
>
> Is there any chance of adding CMake/autotools/maven support to ODB 2.5.0
> release?
>
> Thanks. Best regards, Noah.
>
>
From noah.sk at protonmail.ch  Fri Apr 26 06:50:32 2019
From: noah.sk at protonmail.ch (noah.sk)
Date: Fri Apr 26 07:09:56 2019
Subject: [odb-users] CMake/autotools/meson support in release
In-Reply-To: <CAJQ2NJQ5HUsBUDwQ3jidmZa4dbyVv-5SZrnj+MfC4ugKZVRXTQ@mail.gmail.com>
References: <Hp7dewk9dUFb5Cqfbc36Ro1e0JT2T5W2O_wo6PkYQyufq6FwA60dVZxp8g5VBw4LXEqyTc7NyDmp4UP0J0-D8J1fOczPJfiSGvaJ68jSE54=@protonmail.ch>
	<CAJQ2NJQ5HUsBUDwQ3jidmZa4dbyVv-5SZrnj+MfC4ugKZVRXTQ@mail.gmail.com>
Message-ID: <PK5APWFCV3f3ww7txb4dT9dW9EuejsMf0GHwmVTPbSJnkJLtvlN7pK8XXO9PotjrO9lPpo_N0-DImjcdCYfTqZfTPuaWu4dQ6CvQrqFIq6w=@protonmail.ch>

It is for building apps with ODB, not for building ODB itself.

??????? Original Message ???????
On Friday, April 26, 2019 11:36 AM, Panayiotis Georgiou <ps.georgiou@gmail.com> wrote:

> Hi,
>
> Some time ago, I found the following repository:
> https://github.com/BtbN/OdbCmake
>
> It includes an example of CMakeLists.txt, and the useful modules FindODB.cmake and UseODB.cmake.
>
> I've been using it in my project with ODB 2.4.0 and Cmake with good success. Not all the command line parameters of the ODB compiler are supported, but if you have a look in UseODB.cmake, it's relatively easy to add new ones.
>
> P.
>
> On Fri, Apr 26, 2019 at 9:01 AM noah.sk <noah.sk@protonmail.ch> wrote:
>
>> Hi! We look to use ODB as C++ ORM but currently we have discovered lack of support widely used build systems in ODB beta and git versions.
>>
>> Unfortunately, we are limited to use build tools like CMake, autotools or meson to make own build of ODB.
>>
>> Do you plan to continue support only build2?
>>
>> Is there any chance of adding CMake/autotools/maven support to ODB 2.5.0 release?
>>
>> Thanks. Best regards, Noah.
From boris at codesynthesis.com  Fri Apr 26 11:08:47 2019
From: boris at codesynthesis.com (Boris Kolpackov)
Date: Fri Apr 26 11:27:59 2019
Subject: [odb-users] CMake/autotools/meson support in release
In-Reply-To: <Hp7dewk9dUFb5Cqfbc36Ro1e0JT2T5W2O_wo6PkYQyufq6FwA60dVZxp8g5VBw4LXEqyTc7NyDmp4UP0J0-D8J1fOczPJfiSGvaJ68jSE54=@protonmail.ch>
References: <Hp7dewk9dUFb5Cqfbc36Ro1e0JT2T5W2O_wo6PkYQyufq6FwA60dVZxp8g5VBw4LXEqyTc7NyDmp4UP0J0-D8J1fOczPJfiSGvaJ68jSE54=@protonmail.ch>
Message-ID: <boris.20190426165941@codesynthesis.com>

noah.sk <noah.sk@protonmail.ch> writes:

> Unfortunately, we are limited to use build tools like CMake, autotools
> or meson to make own build of ODB.
> 
> Do you plan to continue support only build2?

Yes, currently we only plan to support build2 as a way to build ODB
(which is not to say that you cannot use other build systems for your
own project that use ODB).

Note also that quite a few (also initially skeptical) users have
successfully integrated a build2-based ODB build into their setup.
The fact that build2 is dependency-free (all you need is a C++
compiler) and can easily be invoked from other build systems makes
this fairly straightforward. So maybe give it a try?

https://codesynthesis.com/products/odb/doc/install-build2.xhtml

From per.edin at sequence-point.se  Sun Apr 28 05:12:08 2019
From: per.edin at sequence-point.se (Per Edin)
Date: Mon Apr 29 07:49:20 2019
Subject: [odb-users] CMake/autotools/meson support in release
In-Reply-To: <boris.20190426165941@codesynthesis.com>
References: <Hp7dewk9dUFb5Cqfbc36Ro1e0JT2T5W2O_wo6PkYQyufq6FwA60dVZxp8g5VBw4LXEqyTc7NyDmp4UP0J0-D8J1fOczPJfiSGvaJ68jSE54=@protonmail.ch>
	<boris.20190426165941@codesynthesis.com>
Message-ID: <CAJA0JPBzcvrpWodxk-wWHZGhkRDYQySuxy0dO47=X=ZB8h1VFw@mail.gmail.com>

On Fri, Apr 26, 2019 at 5:09 PM Boris Kolpackov <boris@codesynthesis.com> wrote:
>
> noah.sk <noah.sk@protonmail.ch> writes:
>
> > Do you plan to continue support only build2?
>
> Yes, currently we only plan to support build2 as a way to build ODB
> (which is not to say that you cannot use other build systems for your
> own project that use ODB).
>

Noah, it's definitely worth the potential "trouble" of switching to
build2. If CMake is the build-system equivalent of what SVN is to CVS,
then build2 is the build-system equivalent of what Git is to SVN.

Advanced, flexible, has a ton of new concepts and documentation, and a
bit of a threshold to learn initilly. But I'll never go back to
anything else for my own projects.

Add the quality of ODB into the mix and it's a no-brainer.

// Per

From aheadsoftwaredeveloper at gmail.com  Tue Apr 30 23:07:47 2019
From: aheadsoftwaredeveloper at gmail.com (Allan Head)
Date: Wed May  1 09:29:03 2019
Subject: [odb-users] Visual Studio 2017 - LNK1104 cannot open file
 'odb-mysql-d.lib' for Hello World
Message-ID: <CAKm0JUCL6oLJbuoJkQgYTWxGrMtQtcmg-RvsFVU=GwvFFqcYiw@mail.gmail.com>

OS:  Win 10 Pro Build 17763 ver 1809
ODB Version:  2.5.0-b.13
Examples File:  odb-examples-2.4.0.zip

After following the Build2 installation instructions, I cannot seem to get
the Hello World example to work.  I opened the hello-mysql-vc12 project and
upgraded it to vc141.  The program compiles, but during the build it
generates the error LINK:  fatal error LNK1104: cannot open file
'odb-mysql-d.lib'.  I do not have an 'odb-mysql-d.lib' file, but I do have
an odb-mysql.lib file and an odb-mysql.dll.lib file.

Should the odb-example-2.4.0 work under ODB 2.5.0-b.13 or am I trying to do
the impossible?

Allan