Firebird

----- Original Message ----- 
From: "Remco Seesink" <raseesink@...>
To: <pkg-firebird-general@...>;
<firebird-devel@...>
Sent: Sunday, July 11, 2004 11:56 AM
Subject: Re: [Pkg-firebird-general] Re: beginning manuals for gdef and gpre


> On Sat, 10 Jul 2004 20:23:21 -0400
> "Grzegorz B. Prokopski" <gadek@...> wrote:
>
> > W li_cie z czw, 08-07-2004, godz. 16:44, Daniel Urban pisze:
> > > Hi
> > > I've found documentation for gbak, bfix, gsec
> > > http://www.destructor.de/firebird/index.htm
> >
> > I can not find any information under what license these documents are
> > available.  For any document to be included in Debian package it should
> > also be available under terms compatible with DFSG requirements.
>
> I will (try to) contact the website owner to find out where they came
from.
> Would it be smart to place all documentation written by us under the
> Interbase Public License?

In my opinion  IDPL (Initial Developer's Public License) used by firebird
team
for new code (not from interbase) would be better.

> I just noticed qli has some self documentation
> which could be copied into the manual.
Where? in itself code?

>
> I think we should coordinate with upstream for producing manuals. They may
> already have something going on. CC-ing this mail to firebird-devel.
> Please reply to:
> pkg-firebird-general@...
> firebird-devel@...

no to firebird-devel but firebird-doc
firebird-doc@...

Best regards
Daniel Urban

Hi Iman Biglari,

Iman Biglari <ibiglari@...> wrote:
>    I don't know how to proide the order, because in Arabic and Farsi the
> characters change when they are mixed together. Anyway, I'll write the
> persian alphabet in the order they should be, with their UNICODE number.=
 I
> hope you can use this information. Let me know if there's anything else =
I
> must provide:
> 
> U+0622 (0xC2): =D8=A2
> U+0627 (0xC7): =D8=A7
> U+0628 (0xC8): =D8=A8
[...]

Thank you for collecting this list. I assume I can get a first 
approximation working, based on this list, which is better 
than the current situation.

But I suspect, there may be more things which must be observed, when 
aiming for complete consistencies with Arabic dictionaries, library sortin=
g
or just phone books. Perhaps you can locate some article about this.

And a most basic question:
You seem to want a collation for the Windows 1256 code page. Do you think,=
 
this is the most often used character set for Arabic?

Regards,
Peter Jacobi

Hi;
   I don't think there should be any difference between sortings in 
phonebooks, listings, and so on, but I'll keep looking. About the code 
page, you're right, but the order in the code page is not correct. As 
you can see in the list, some characters are not in orther (in fact, the 
four persian characters: P, CH, ZH and G), and at the end the two 
letters V and H are misplaced. But in Windows, 1256 is the only 
character set which is appliable to both Arabic and Farsi.
   Thanks;
--
PS: I can send you a sample database, with the correct order in an 
integer field if your mailbox accepts attachments. Does it help you?

Peter Jacobi wrote:
> Hi Iman Biglari,
> 
> Iman Biglari <ibiglari@...> wrote:
> 
>>   I don't know how to proide the order, because in Arabic and Farsi the
>>characters change when they are mixed together. Anyway, I'll write the
>>persian alphabet in the order they should be, with their UNICODE number. I
>>hope you can use this information. Let me know if there's anything else I
>>must provide:
>>
>>U+0622 (0xC2): آ
>>U+0627 (0xC7): ا
>>U+0628 (0xC8): ب
> 
> [...]
> 
> Thank you for collecting this list. I assume I can get a first 
> approximation working, based on this list, which is better 
> than the current situation.
> 
> But I suspect, there may be more things which must be observed, when 
> aiming for complete consistencies with Arabic dictionaries, library sorting
> or just phone books. Perhaps you can locate some article about this.
> 
> And a most basic question:
> You seem to want a collation for the Windows 1256 code page. Do you think, 
> this is the most often used character set for Arabic?
> 
> Regards,
> Peter Jacobi
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email sponsored by Black Hat Briefings & Training.
> Attend Black Hat Briefings & Training, Las Vegas July 24-29 - 
> digital self defense, top technical experts, no vendor pitches, 
> unmatched networking opportunities. Visit http://www.blackhat.com
> Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
> 

-- 
If you want the RAINBOW, you have to pass through the rain.

Hi Iman Biglari,

Iman Biglari <ibiglari@...> wrote:
>    I don't think there should be any difference between sortings in 
> phonebooks, listings, and so on, but I'll keep looking. 

I didn't necessarily imply, that they should be different, but wanted to 
say, that it may be more complicated than assigning a single number
to every character.

Most references on Arabic collation that I found by a first
use of Google, all say, that correct Arabic collation is terribly
complicated and involves first extracting the "stem" of the word,
whatever this means in this context.

> About the code page,
> you're right, but the order in the code page is not correct. 

That is completely normal. No code page or ISO character set has in
itself the correct order of characters.

> PS: I can send you a sample database, with the correct order in an 
> integer field if your mailbox accepts attachments. Does it help you?

Your text list is perfectly O.K. It will be stuffed into a Python program 
for generating the necessary tables.

Regards,
Peter Jacobi

Hi;
   can you send me the links you found in Google? My mother language is 
persian and I don't know what "stem" does mean!

Peter Jacobi wrote:
> Hi Iman Biglari,
> 
> Iman Biglari <ibiglari@...> wrote:
> 
>>   I don't think there should be any difference between sortings in 
>>phonebooks, listings, and so on, but I'll keep looking. 
> 
> 
> I didn't necessarily imply, that they should be different, but wanted to 
> say, that it may be more complicated than assigning a single number
> to every character.
> 
> Most references on Arabic collation that I found by a first
> use of Google, all say, that correct Arabic collation is terribly
> complicated and involves first extracting the "stem" of the word,
> whatever this means in this context.
> 
> 
>>About the code page,
>>you're right, but the order in the code page is not correct. 
> 
> 
> That is completely normal. No code page or ISO character set has in
> itself the correct order of characters.
> 
> 
>>PS: I can send you a sample database, with the correct order in an 
>>integer field if your mailbox accepts attachments. Does it help you?
> 
> 
> Your text list is perfectly O.K. It will be stuffed into a Python program 
> for generating the necessary tables.
> 
> Regards,
> Peter Jacobi
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email sponsored by Black Hat Briefings & Training.
> Attend Black Hat Briefings & Training, Las Vegas July 24-29 - 
> digital self defense, top technical experts, no vendor pitches, 
> unmatched networking opportunities. Visit http://www.blackhat.com
> Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
> 

-- 
If you want the RAINBOW, you have to pass through the rain.

On Wednesday 07 July 2004 18:06, Treeve Jelbert wrote:
> On Wednesday 07 July 2004 14:29, Treeve Jelbert wrote:
> > I am reporting this here, since I may be the only one currently using
> > gcc-3.4.1 on linux-2.6.7
> >
> > Using the latest CVS-HEAD, I get a number of warnings about invalid code,
> > followed by a compiler crash
> >
> > several
> >
> > ../src/jrd/../jrd/evl_like.h:193: warning: comparison is always true due
> > to limited range of data type
> >
> > and
> >
> > ../src/jrd/pag.cpp:1144: warning: invalid access to non-static data
> > member `Ods::page_inv_page::pip_bits' of NULL object
> > ../src/jrd/pag.cpp:1144: warning: (perhaps the `offsetof' macro was used
> > incorrectly)
> >
> > full error list is attached
> >
> > g++ -O -march=athlon-tbird -mmmx -m3dnow -pipe -O3 -O -march=athlon-tbird
> > -mmmx -m3dnow -pipe -O3 -I../src/include/gen -I../src/include -O
> > -march=athlon-tbird -mmmx -m3dnow -pipe -O3  -fno-omit-frame-pointer
> > -fno-builtin -DNDEBUG -DLINUX -pipe -MMD -fPIC -fmessage-length=0
> > -DPROD_BUILD -DSUPERSERVER -I../src/include/gen -I../src/include -O
> > -march=athlon-tbird -mmmx -m3dnow -pipe -O3  -fno-omit-frame-pointer
> > -fno-builtin -DNDEBUG -DLINUX -pipe -MMD -fPIC -fmessage-length=0
> > -DPROD_BUILD -c ../src/jrd/svc.cpp -o ../temp/superserver/jrd/svc.o
> > ../src/jrd/svc.cpp: In function `void* SVC_read_ib_log(void*)':
> > ../src/jrd/svc.cpp:2106: internal compiler error: in delete_insn, at
> > cfgrtl.c:146
>
> I managed to bypass the compiler crash, by removing the
> -fno-omit-frame-pointer -fno-builtin
>
> I now get the following error
>
> ../src/utilities/ibmgr/srvrmgr.cpp: In function `bool
> print_pool(ibmgr_data_t*)':
> ../src/utilities/ibmgr/srvrmgr.cpp:662: error: invalid initialization of
> non-const reference of type 'UCHAR*&' from a temporary of type 'UCHAR*'
> ../src/include/../common/stuff.h:38: error: in passing argument 1 of `void
> add_word(UCHAR*&, int)'
> make[3]: *** [../temp/std/utilities/ibmgr/srvrmgr.o] Error 1
>

build 7954 gives the following error:

../src/utilities/ibmgr/srvrmgr.cpp: In function `bool print_pool(const 
ibmgr_data_t*)':
../src/utilities/ibmgr/srvrmgr.cpp:669: error: invalid conversion from `void* 
const*' to `void**'
../src/utilities/ibmgr/srvrmgr.cpp:669: error:   initializing argument 2 of 
`ISC_STATUS isc_service_query(ISC_STATUS*, void**, isc_resv_handle*, short 
unsigned int, const ISC_SCHAR*, short unsigned int, const ISC_SCHAR*, short 
unsigned int, ISC_SCHAR*)'
make[3]: *** [../temp/std/utilities/ibmgr/srvrmgr.o] Error 1



> > I can supply a full cmpile log if anyone wants it.
> >
> >
> >
> > Regards,  Treeve
>
> Regards,  Treeve
Regards,  Treeve 

-- 
PGP Key ID: AB929B24
PGP Key Fingerprint:31D9 D22F 42E6 F545 662E  AB6F 9697 34C5 AB92 9B24

On Wednesday 07 July 2004 18:06, Treeve Jelbert wrote:
> On Wednesday 07 July 2004 14:29, Treeve Jelbert wrote:
> > I am reporting this here, since I may be the only one currently using
> > gcc-3.4.1 on linux-2.6.7
> >
> > Using the latest CVS-HEAD, I get a number of warnings about invalid code,
> > followed by a compiler crash
> >
> > several
> >
> > ../src/jrd/../jrd/evl_like.h:193: warning: comparison is always true due
> > to limited range of data type
> >
> > and
> >
> > ../src/jrd/pag.cpp:1144: warning: invalid access to non-static data
> > member `Ods::page_inv_page::pip_bits' of NULL object
> > ../src/jrd/pag.cpp:1144: warning: (perhaps the `offsetof' macro was used
> > incorrectly)
> >
> > full error list is attached
> >
> > g++ -O -march=athlon-tbird -mmmx -m3dnow -pipe -O3 -O -march=athlon-tbird
> > -mmmx -m3dnow -pipe -O3 -I../src/include/gen -I../src/include -O
> > -march=athlon-tbird -mmmx -m3dnow -pipe -O3  -fno-omit-frame-pointer
> > -fno-builtin -DNDEBUG -DLINUX -pipe -MMD -fPIC -fmessage-length=0
> > -DPROD_BUILD -DSUPERSERVER -I../src/include/gen -I../src/include -O
> > -march=athlon-tbird -mmmx -m3dnow -pipe -O3  -fno-omit-frame-pointer
> > -fno-builtin -DNDEBUG -DLINUX -pipe -MMD -fPIC -fmessage-length=0
> > -DPROD_BUILD -c ../src/jrd/svc.cpp -o ../temp/superserver/jrd/svc.o
> > ../src/jrd/svc.cpp: In function `void* SVC_read_ib_log(void*)':
> > ../src/jrd/svc.cpp:2106: internal compiler error: in delete_insn, at
> > cfgrtl.c:146
>
> I managed to bypass the compiler crash, by removing the
> -fno-omit-frame-pointer -fno-builtin
>
> I now get the following error
>
> ../src/utilities/ibmgr/srvrmgr.cpp: In function `bool
> print_pool(ibmgr_data_t*)':
> ../src/utilities/ibmgr/srvrmgr.cpp:662: error: invalid initialization of
> non-const reference of type 'UCHAR*&' from a temporary of type 'UCHAR*'
> ../src/include/../common/stuff.h:38: error: in passing argument 1 of `void
> add_word(UCHAR*&, int)'
> make[3]: *** [../temp/std/utilities/ibmgr/srvrmgr.o] Error 1
>

build 7954 gives the following error


../src/utilities/ibmgr/srvrmgr.cpp: In function `bool print_pool(const 
ibmgr_data_t*)':
../src/utilities/ibmgr/srvrmgr.cpp:669: error: invalid conversion from `void* 
const*' to `void**'
../src/utilities/ibmgr/srvrmgr.cpp:669: error:   initializing argument 2 of 
`ISC_STATUS isc_service_query(ISC_STATUS*, void**, isc_resv_handle*, short 
unsigned int, const ISC_SCHAR*, short unsigned int, const ISC_SCHAR*, short 
unsigned int, ISC_SCHAR*)'
make[3]: *** [../temp/std/utilities/ibmgr/srvrmgr.o] Error 1



> > I can supply a full cmpile log if anyone wants it.
> >
> >
> >
> > Regards,  Treeve
>
> Regards,  Treeve
Regards,  Treeve 

-- 
PGP Key ID: AB929B24
PGP Key Fingerprint:31D9 D22F 42E6 F545 662E  AB6F 9697 34C5 AB92 9B24

On Wednesday 07 July 2004 18:06, Treeve Jelbert wrote:
> On Wednesday 07 July 2004 14:29, Treeve Jelbert wrote:
> > I am reporting this here, since I may be the only one currently using
> > gcc-3.4.1 on linux-2.6.7
> >
> > Using the latest CVS-HEAD, I get a number of warnings about invalid code,
> > followed by a compiler crash
> >
> > several
> >
> > ../src/jrd/../jrd/evl_like.h:193: warning: comparison is always true due
> > to limited range of data type
> >
> > and
> >
> > ../src/jrd/pag.cpp:1144: warning: invalid access to non-static data
> > member `Ods::page_inv_page::pip_bits' of NULL object
> > ../src/jrd/pag.cpp:1144: warning: (perhaps the `offsetof' macro was used
> > incorrectly)
> >
> > full error list is attached
> >
> > g++ -O -march=athlon-tbird -mmmx -m3dnow -pipe -O3 -O -march=athlon-tbird
> > -mmmx -m3dnow -pipe -O3 -I../src/include/gen -I../src/include -O
> > -march=athlon-tbird -mmmx -m3dnow -pipe -O3  -fno-omit-frame-pointer
> > -fno-builtin -DNDEBUG -DLINUX -pipe -MMD -fPIC -fmessage-length=0
> > -DPROD_BUILD -DSUPERSERVER -I../src/include/gen -I../src/include -O
> > -march=athlon-tbird -mmmx -m3dnow -pipe -O3  -fno-omit-frame-pointer
> > -fno-builtin -DNDEBUG -DLINUX -pipe -MMD -fPIC -fmessage-length=0
> > -DPROD_BUILD -c ../src/jrd/svc.cpp -o ../temp/superserver/jrd/svc.o
> > ../src/jrd/svc.cpp: In function `void* SVC_read_ib_log(void*)':
> > ../src/jrd/svc.cpp:2106: internal compiler error: in delete_insn, at
> > cfgrtl.c:146
>
> I managed to bypass the compiler crash, by removing the
> -fno-omit-frame-pointer -fno-builtin
>
> I now get the following error
>
> ../src/utilities/ibmgr/srvrmgr.cpp: In function `bool
> print_pool(ibmgr_data_t*)':
> ../src/utilities/ibmgr/srvrmgr.cpp:662: error: invalid initialization of
> non-const reference of type 'UCHAR*&' from a temporary of type 'UCHAR*'
> ../src/include/../common/stuff.h:38: error: in passing argument 1 of `void
> add_word(UCHAR*&, int)'
> make[3]: *** [../temp/std/utilities/ibmgr/srvrmgr.o] Error 1
>

build 7954 gives the following error


../src/utilities/ibmgr/srvrmgr.cpp: In function `bool print_pool(const 
ibmgr_data_t*)':
../src/utilities/ibmgr/srvrmgr.cpp:669: error: invalid conversion from `void* 
const*' to `void**'
../src/utilities/ibmgr/srvrmgr.cpp:669: error:   initializing argument 2 of 
`ISC_STATUS isc_service_query(ISC_STATUS*, void**, isc_resv_handle*, short 
unsigned int, const ISC_SCHAR*, short unsigned int, const ISC_SCHAR*, short 
unsigned int, ISC_SCHAR*)'
make[3]: *** [../temp/std/utilities/ibmgr/srvrmgr.o] Error 1




> > I can supply a full cmpile log if anyone wants it.
> >
> >
> >
> > Regards,  Treeve
>
> Regards,  Treeve
Regards,  Treeve 

-- 
PGP Key ID: AB929B24
PGP Key Fingerprint:31D9 D22F 42E6 F545 662E  AB6F 9697 34C5 AB92 9B24

Treeve Jelbert wrote:
>> 
> build 7954 gives the following error
> 
> g++ -O -march=athlon-tbird -mmmx -m3dnow -pipe  -O
> -march=athlon-tbird -mmmx
> -m3dnow -pipe  -I../src/include/gen -I../src/include -DNDEBUG
> -DLINUX -pipe
> -MMD -fPIC -fmessage-length=0  -DPROD_BUILD -O -march=athlon-tbird
> -mmmx 
> -m3dnow -pipe  -I../src/include/gen -I../src/include -DNDEBUG
> -DLINUX -pipe
> -MMD -fPIC -fmessage-length=0  -DPROD_BUILD -I../src/include/gen
> -I../src/include -DNDEBUG -DLINUX -pipe -MMD -fPIC -fmessage-length=0
> -DPROD_BUILD -c ../src/utilities/ibmgr/srvrmgr.cpp
> -o ../temp/std/utilities/ibmgr/srvrmgr.o
> ../src/utilities/ibmgr/srvrmgr.cpp: In function `bool
> print_pool(const ibmgr_data_t*)':
> ../src/utilities/ibmgr/srvrmgr.cpp:669: error: invalid conversion 
> from `void*
> const*' to `void**'
> ../src/utilities/ibmgr/srvrmgr.cpp:669: error:   initializing
> argument 2 of
> `ISC_STATUS isc_service_query(ISC_STATUS*, void**,
> isc_resv_handle*, short
> unsigned int, const ISC_SCHAR*, short unsigned int, const
> ISC_SCHAR*, short
> unsigned int, ISC_SCHAR*)'
> make[3]: *** [../temp/std/utilities/ibmgr/srvrmgr.o] Error 1

Try again when you can retrieve the latest version.

BTW, what's happening with your mail reader? It generates replies with
Re:<space>subject -> correct
Re:<space><space>subject -> wrong
Re::<space>subject -> wrong

This makes article/email threading a pain.

C.