Mailing List Archive: 49091 messages
  • Home
  • Script library
  • AltME Archive
  • Mailing list
  • Articles Index
  • Site search
 

RE : Re: R: fixes - Re: MySQL 1.0.0 beta driver released .... (BUG REPORT)

 [1/11] from: alain:goye:free at: 6-Feb-2006 11:39


Hi Dockimbel, Trying to run mysql-protocol.r I get the "Bad handshake" error everytime, together with strange rubbish printouts, as in the following (running on Win2K): REBOL/View 1.3.2.3.1 5-Dec-2005 Core 2.6.3 Copyright 2000-2005 REBOL Technologies. All rights reserved. REBOL is a trademark of REBOL Technologies. WWW.REBOL.COM *** Visit http://www.REBOL.com for View/Pro features. Type desktop to start the Viewtop.
>> do %mysql-protocol.r
Script: "MySQL Protocol" (25-Jan-2006) MySQL protocol loaded
>> db: open mysql://root:?-localhost/cist
Password: ****** connecting to: localhost {^K^-^U^-^-root^-st=C3=AD=E2=80=A0=C3=85^!^R=C3=BE=C3=BAxc[B=C3=BB=C2=A3=C2=9D4si=C3=9D^-cist^-}
>> insert db "SELECT * FROM staff"
connecting to: localhost {^K^-^U^-^-root^-=CB=9C=C2=B4^Z^A=C3=A4jR=C2=B4=C3=A7 =C2=A5=C3=98=C3=99v=C3=BAk=C2=A7=C3=B1:#^-cist^-} ** User Error: ERROR 1043 : Bad handshake ** Near: insert db "SELECT * FROM staff" Actually I get the same thing whatever (good or bad) password I give in the "open" command, and it never complains when I close the port (though it should not succeed in opening the port if the password is bad?) Do I miss something or is it a bug? Alain G.

 [2/11] from: petr:krenzelok:trz:cz at: 6-Feb-2006 6:10


Hi, are you using the latest version? http://softinnov.org/tmp/mysql-protocol.r Cheers, -pekr- Alain Goy=C3=A9 wrote:

 [3/11] from: alain::goye::free::fr at: 6-Feb-2006 13:55


Yes, I checked that. This is the version with the "random 1043 and 1045 error" bug fix... But my error is not random, it is deterministic! (although the jammed string 1 get when trying to connect seems to be partly random...) Alain. -----Message d'origine----- De : rebol-bounce-rebol.com [mailto:rebol-bounce-rebol.com] De la part de Petr Krenzelok Envoy=E9 : lundi 6 f=E9vrier 2006 12:10 =C0 : rebolist-rebol.com Objet : [REBOL] Re: RE : Re: R: fixes - Re: MySQL 1.0.0 beta driver released .... (BUG REPORT) Hi, are you using the latest version? http://softinnov.org/tmp/mysql-protocol.r Cheers, -pekr- Alain Goy=C3=A9 wrote:

 [4/11] from: anton:wilddsl:au at: 6-Feb-2006 22:49


Hi Alain, what is the version of MySQL ? Anton.

 [5/11] from: alain:goye:free at: 6-Feb-2006 19:25


Hi Anton, This is MySQL 4.1.14-nt ... Alain. -----Message d'origine----- De : rebol-bounce-rebol.com [mailto:rebol-bounce-rebol.com] De la part de Anton Rolls Envoy=E9 : lundi 6 f=E9vrier 2006 18:49 =C0 : rebolist-rebol.com Objet : [REBOL] Re: RE : Re: R: fixes - Re: MySQL 1.0.0 beta driver released .... (BUG REPORT) Hi Alain, what is the version of MySQL ? Anton.

 [6/11] from: petr:krenzelok:trz:cz at: 6-Feb-2006 13:38


Alain Goy=E9 wrote:
> Hi Anton, > > This is MySQL 4.1.14-nt ... >
Hmm, strange a bit, worked for me, although can't remember right now, as I installed 5.x in the meantime. Are you using empty password? Or are you using "old_password" setting in mySQL? -pekr-

 [7/11] from: alain:goye:free at: 6-Feb-2006 20:06


No, I'm not using an empty password, and my "old_passwords" variable is OFF... But here is something new: I installed the same mySql version on another machine (Windows server 2003) and (accessing it over the LAN with the same Rebol) now it works better... Except for the rubbish printouts which are still here:
>> do %mysql-protocol.r
Script: "MySQL Protocol" (25-Jan-2006) MySQL protocol loaded
>> ; good password: >> db: open mysql://root:mySqlAdmin-10.10.10.14/cist
connecting to: 10.10.10.14 {^K^-^Y^-^-root^-=C3=85z0=C2=81^U =C5=A0#=C2=B2=C3=AB^X:=C3=BF=C3=9A=C3=A0^XBYX=C3=83^-cist^-}
>> insert db "SELECT * FROM staff"
== none
>> first db
** Script Error: Invalid argument: 0000-00-00 00:00:00 ** Where: convert-types ** Near: change at row 17 to Alain. -----Message d'origine----- De : rebol-bounce-rebol.com [mailto:rebol-bounce-rebol.com] De la part de Petr Krenzelok Envoy=C3=A9 : lundi 6 f=C3=A9vrier 2006 19:38 =C3=80 : rebolist-rebol.com Objet : [REBOL] Re: RE : Re: RE : Re: R: fixes - Re: MySQL 1.0.0 beta driver released .... (BUG REPORT) Alain Goy=E9 wrote:
> Hi Anton, > > This is MySQL 4.1.14-nt ... >
Hmm, strange a bit, worked for me, although can't remember right now, as=20 I installed 5.x in the meantime. Are you using empty password? Or are=20 you using "old_password" setting in mySQL? -pekr-

 [8/11] from: petr:krenzelok:trz:cz at: 6-Feb-2006 15:07


Alain Goy=C3=A9 wrote:
> No, I'm not using an empty password, and my "old_passwords" variable is > OFF...
<<quoted lines omitted: 12>>
> {^K^-^Y^-^-root^-=C3=85z0=C2=81^U > =C5=A0#=C2=B2=C3=AB^X:=C3=BF=C3=9A=C3=A0^XBYX==
C3=83^-cist^-}
oops, seems like some forgotten debug output to me? I get it here too ....
>>> insert db "SELECT * FROM staff" >>>
<<quoted lines omitted: 4>>
> ** Where: convert-types > ** Near: change at row 17 to
try copy db ..... in opposite to /Command, Doc's scheme does not use the principle of: ble: open mysql://user:pass-domain/db conn: first ble copy conn simply try "print copy db" to get the result -pekr-

 [9/11] from: petr::krenzelok::trz::cz at: 6-Feb-2006 15:14


Petr Krenzelok wrote:
> Alain Goy=C3=A9 wrote: >> No, I'm not using an empty password, and my "old_passwords" variable is
<<quoted lines omitted: 28>>
>> > oops, seems like some forgotten debug output to me? I get it here too ...
yes, I seem to get it fixed. Just one forgotten probe in do-handshake function - look at the code below: send-packet port probe rejoin [ write-int client-param write-int24 (length? port/user) + (length? port/pass) + 7 + std-header-length write-string port/user write-string key: scramble port/pass port write-string port/target ] remove the probe .... Now at least the rubbish output should be gone ... Cheers, -pekr-

 [10/11] from: dockimbel:free at: 6-Feb-2006 20:46


Hi Alain, I'm guilty for the "probe" left in the last beta uploaded on our website. I've fixed that. About your handshake issue, I've tried to reproduce that with MySQLd v4.1-16-nt (couldn't find a 1-14 version). I saw no error during the handshake process (with or without password). Yet, I've used tables from my 3.x install with v4+ servers, so it may impact the relevance of my validating tests... Conversion error: the error you get when reading the first record from the result of your query is caused by REBOL not allowing null dates ("0000-00-00 00:00:00"). I'm open to suggestions on how should the driver act upon such not LOAD-able date, besides catching the error. Thanks for your report, -- DocKimbel. Selon Alain Goy=C3=A9 <alain.goye-free.fr>:

 [11/11] from: alain::goye::free::fr at: 7-Feb-2006 8:40


Hi Dockimbel, Petr, Anton, Thanks so much to all for your help with this issue. One mystery remains (why I still have the "Bad handshake" on the local-machine mySql) but at least I could have a version running. As for the "null dates" error, as I already had it with the former version of mysql-protocol.r, I "patched" it this way. Does it make sense that a null date be converted to a none value? conv-model: [ ... date [attempt [to date!]] time [to time!] datetime [attempt [to date!]] ... ] Alain. -----Message d'origine----- De : rebol-bounce-rebol.com [mailto:rebol-bounce-rebol.com] De la part de dockimbel-free.fr Envoy=E9 : mardi 7 f=E9vrier 2006 02:47 =C0 : rebolist-rebol.com Objet : [REBOL] Re: RE : Re: RE : Re: RE : Re: R: fixes - Re: MySQL 1.0.0 beta driver released .... (BUG REPORT) Hi Alain, I'm guilty for the "probe" left in the last beta uploaded on our website.= I've fixed that. About your handshake issue, I've tried to reproduce that with MySQLd v4.1= -16-nt (couldn't find a 1-14 version). I saw no error during the handshake proce= ss (with or without password). Yet, I've used tables from my 3.x install wit= h v4+ servers, so it may impact the relevance of my validating tests... Conversion error: the error you get when reading the first record from th= e result of your query is caused by REBOL not allowing null dates ("0000-00= -00 00:00:00"). I'm open to suggestions on how should the driver act upon suc= h not LOAD-able date, besides catching the error. Thanks for your report, -- DocKimbel. Selon Alain Goy=C3=A9 <alain.goye-free.fr>:
> No, I'm not using an empty password, and my "old_passwords" variable > is
<<quoted lines omitted: 11>>
> connecting to: 10.10.10.14 {^K^-^Y^-^-root^-=C3=85z0=C2=81^U > =C5=A0#=C2=B2=C3=AB^X:=C3=BF=C3=9A=C3=A0^XBYX=

Notes
  • Quoted lines have been omitted from some messages.
    View the message alone to see the lines that have been omitted