Thread: Solaris9 Sparc64 edition
Hi Raphaël. I could do this preparation. But, the condition of this is poor. :-( gcc version 3.3.2 atk-1.2.4 gtk+-2.2.4 pango-1.2.5 glib-2.2.3 wxWindows-20040128-CVS See this. http://cre-ent.skcapi.co.jp/~saito/pgadmin3/solaris9_2.jpg Are there any hints? However, There is a business until Monday, and it is delayed... regards, Hiroshi Saito
Hiroshi Saito wrote: >Hi Raphaël. > > > Hi Hiroshi, >I could do this preparation. >But, the condition of this is poor. >:-( > >See this. >http://cre-ent.skcapi.co.jp/~saito/pgadmin3/solaris9_2.jpg >Are there any hints? > > mmh no idea. When I first tried the run of pgAdmin3 under Solaris, I had wxAsserts which I first thought to be the results of my quick wx patching... Nothing I could not "cancel" to go further in the program. Is it possible to get an account on you box or not ? BTW, I'm gonna try the build of pgAdmin3 on Solaris x86, I hope to get quite similar behaviours we have on the sparc release. We stay in touch. Raphaël Private joke: I didn't tell but I loved you smile on our first Hall of fame picture! ;) Who will be the next one.....
Hiroshi Saito wrote: >Hi Raphaël. > > I found necessary liblary from the gnome environment. > > Yes, I think it would be interesting to use libs provided published by sun (www.sun.com/gnome) but we must verify that they didn't build them with their workshop. Otherwise we will for sure enter compatibility problems between gcc generated code and forte developer compiler generated one. >However, >I ignore it, and it seems that I met with the rivalry problem by creating it separately from source. >So, it was decided again that I installed it. >I am very busy, and I don't have time. >Wait for a while now. > > keep cool! I'm sure Dave can wait a day or two ;p Regards, Raphaël
Hi Raphaël. From: "Raphaël Enrici" <blacknoz@club-internet.fr> (snip) > > I found necessary liblary from the gnome environment. > > > > > Yes, I think it would be interesting to use libs provided published by > sun (www.sun.com/gnome) but we must verify that they didn't build them > with their workshop. Otherwise we will for sure enter compatibility > problems between gcc generated code and forte developer compiler > generated one. It was really unpleasant though it was seen by using it.:-( i10n was not a desirable environment. It is as it is, and font-conversion of resource doesn't move.?? So, I tried it with without-unicode. http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_debug4.png liblary of the offer doesn't work so that I may think. Let's break off causal relation. It is chased again because it is very interesting.:-) regards, Hiroshi Saito
They are new conditions. http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_debug5.png http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_debug6.png It seems that even i10n can be used somehow. However, I must study it a little more.:-) regards, Hiroshi Saito > Hi Raphaël. > > From: "Raphaël Enrici" <blacknoz@club-internet.fr> > (snip) > > > I found necessary liblary from the gnome environment. > > > > > > > > Yes, I think it would be interesting to use libs provided published by > > sun (www.sun.com/gnome) but we must verify that they didn't build them > > with their workshop. Otherwise we will for sure enter compatibility > > problems between gcc generated code and forte developer compiler > > generated one. > > It was really unpleasant though it was seen by using it.:-( > i10n was not a desirable environment. > It is as it is, and font-conversion of resource doesn't move.?? > So, I tried it with without-unicode. > http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_debug4.png > > liblary of the offer doesn't work so that I may think. > Let's break off causal relation. > It is chased again because it is very interesting.:-) > > regards, > Hiroshi Saito > > > > > ---------------------------(end of broadcast)--------------------------- > TIP 9: the planner will ignore your desire to choose an index scan if your > joining column's datatypes do not match
Even when the maximum returned column length is set to > 511 characters, the true maximum character length for exported columns is 511 Characters ... I tested this with a value of 65535 characters for the Max Column Length and a 1000 charactger text field. Full Articles stored in Text Fields are very frequently over 511 characters in length. Are there any plans to fix this? CG __________________________________ Do you Yahoo!? Yahoo! Mail SpamGuard - Read only the mail you want. http://antispam.yahoo.com/tools
Chris Gamache wrote: >Even when the maximum returned column length is set to > 511 characters, the >true maximum character length for exported columns is 511 Characters ... I >tested this with a value of 65535 characters for the Max Column Length and a >1000 charactger text field. Full Articles stored in Text Fields are very >frequently over 511 characters in length. > >Are there any plans to fix this? > Use "execute to file" for this. Regards, Andreas
Hm... I don't see that option in my Query Tool. I'm using Version 1.0.2 (Nov 28 2003) is this a new feature? Where do you find it? I remember that there was an "Execute To:" feature in PgAdmin II ... CG --- Andreas Pflug <pgadmin@pse-consulting.de> wrote: > Chris Gamache wrote: > > >Even when the maximum returned column length is set to > 511 characters, the > >true maximum character length for exported columns is 511 Characters ... I > >tested this with a value of 65535 characters for the Max Column Length and a > >1000 charactger text field. Full Articles stored in Text Fields are very > >frequently over 511 characters in length. > > > >Are there any plans to fix this? > > > > Use "execute to file" for this. > > Regards, > Andreas > > __________________________________ Do you Yahoo!? Yahoo! Mail SpamGuard - Read only the mail you want. http://antispam.yahoo.com/tools
Chris Gamache wrote: >Hm... I don't see that option in my Query Tool. I'm using Version 1.0.2 (Nov 28 >2003) is this a new feature? Where do you find it? > > You're right, it's not in 1.0; this is added for 1.1. I made a comment on this in the FAQs. Regards, Andreas
Hi Raphaël. It seemed that I couldn't provide it with this environment.... libc of Solaris9 seems to be special. However, This is a thing which can be spent with Single-byte. http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_Sparc64/ I think that wait for a while.*sigh* Any Comments? regards, Hiroshi Saito
Hi Raphaël. Thank you for taking the time to help me in this. From: "Raphaël Enrici" <blacknoz@club-internet.fr> > >Hi Raphaël. > > > >It seemed that I couldn't provide it with this environment.... > >libc of Solaris9 seems to be special. > > > > > I'm not sure I understand... What do you want to say ? Oh,Sorry I was no explanation. It is this. 1) LANG=ja_JP.UTF-8 pgadmin3 running. http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_Sparc64/Solaris12.png http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_Sparc64/Solaris13.png It doesn't become a character.:-( 2) LANG=ja_JP.eucJP pgadmin3 running. In this case, it seems to go well except for gettext, scintilla. http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_Sparc64/Solaris14.png It seems that gtk2 of Solaris9 can't handle UNICODE well. configure --enable-monolithic --enable-unicode --disable-shared --enable-static --with-gtk --enable-gtk2 --prefix=/usr/local/wxw indows ... _gtk_rc_context_get_default_font_name /usr/local/wxwindows/lib/libwx_gtk2u-2.5.a(monolib_settings.o) _gtk_accel_group_detach /usr/local/wxwindows/lib/libwx_gtk2u-2.5.a(monolib_menu.o) _gtk_accel_group_attach /usr/local/wxwindows/lib/libwx_gtk2u-2.5.a(monolib_menu.o) ld: fatal: Symbol referencing errors. No output written to pgadmin3 collect2: ld returned 1 exit status So, this is being used. configure --enable-monolithic --enable-unicode --disable-shared --enable-static --with-gtk --prefix=/usr/local/wxwindows > > >However, > >This is a thing which can be spent with Single-byte. > >http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_Sparc64/ > > > > > It seems you got it running isn't it ? So, where is the problem ? > You couldn't built it in 64bits that's it ? > > Sorry if my english is not up to date Oh it is me.. regards, Hiroshi Saito
Hi Hiroshi, Andreas & buddies, I recently discovered an interesting project with cool & reactive people involved (http://www.blastwave.org) and got an account on some solaris boxes. I'd like to come back on Solaris port and need to sync what was done since my previous build and preferably Hiroshi's one as it seems he went further. My main goal would be to produce a package in SYSV Solaris format and try to solve the unicode problem if my skill permits it. The main problem I encounter today is (just as before) linked to the wxWidgets build. Actual wxWidgets snapshots don't build straight and our patched one does not neither (that's not new). Hiroshi, which wxWid CVS did you use finally ? The one I tried when I got it running or another one ? Do you still have the patched source somewhere ? Andreas, I contacted Ron recently and he told me there was some (little) hope to get something up and running with official wxWid... Something new with this since your last mail concluding to the fact it is unusable ? (that's of interest for Debian port & Solaris port). Regards, Raphaël Hiroshi Saito wrote: >From: "Raphaël Enrici" <blacknoz@club-internet.fr> > > >>>Hi Raphaël. >>>It seemed that I couldn't provide it with this environment.... >>>libc of Solaris9 seems to be special. >>> >>> >>I'm not sure I understand... What do you want to say ? >> >> > >Oh,Sorry I was no explanation. > >It is this. >1) >LANG=ja_JP.UTF-8 pgadmin3 running. >http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_Sparc64/Solaris12.png >http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_Sparc64/Solaris13.png >It doesn't become a character.:-( >2) >LANG=ja_JP.eucJP pgadmin3 running. >In this case, it seems to go well except for gettext, scintilla. >http://cre-ent.skcapi.co.jp/~saito/pgadmin3/Solaris9_Sparc64/Solaris14.png > >It seems that gtk2 of Solaris9 can't handle UNICODE well. >configure --enable-monolithic --enable-unicode --disable-shared --enable-static --with-gtk --enable-gtk2 --prefix=/usr/local/wxw >indows >.... >_gtk_rc_context_get_default_font_name /usr/local/wxwindows/lib/libwx_gtk2u-2.5.a(monolib_settings.o) >_gtk_accel_group_detach /usr/local/wxwindows/lib/libwx_gtk2u-2.5.a(monolib_menu.o) >_gtk_accel_group_attach /usr/local/wxwindows/lib/libwx_gtk2u-2.5.a(monolib_menu.o) >ld: fatal: Symbol referencing errors. No output written to pgadmin3 >collect2: ld returned 1 exit status > > >So, this is being used. >configure --enable-monolithic --enable-unicode --disable-shared --enable-static --with-gtk --prefix=/usr/local/wxwindows > >
Raphaël Enrici wrote: > Andreas, I contacted Ron recently and he told me there was some > (little) hope to get something up and running with official wxWid... > Something new with this since your last mail concluding to the fact it > is unusable ? (that's of interest for Debian port & Solaris port). > Hi Raphaël, unfortely no real news. The font handling in wx has been changed fundamentally, and it doesn't work at all for us now. The last posting of Vadim seemed to show some understanding of what is going wrong, kind-of accepting that the "native look and feel" attitude has been overstressed leading to unfortunate behaviour, so I reposted a quite old proposal to fix this. I offered testing that immediately when basically implemented despite my lack of time, to get it into 2.5.2. There was no reaction on that so far. Regards, Andreas
Hi Raphaël.
From: "Raphaël Enrici" <blacknoz@club-internet.fr>
> Hi Hiroshi, Andreas & buddies,
>
(snip)
>
> The main problem I encounter today is (just as before) linked to the
> wxWidgets build.
> Actual wxWidgets snapshots don't build straight and our patched one does
> not neither (that's not new). Hiroshi, which wxWid CVS did you use
> finally ? The one I tried when I got it running or another one ? Do you
> still have the patched source somewhere ?
I used this
wxWindows-20040301-CVS
--- src/unix/dialup.cpp.orig Mon May 10 09:07:44 2004
+++ src/unix/dialup.cpp Mon May 10 09:08:47 2004
@@ -716,8 +716,8 @@
#if defined(__SOLARIS__) || defined (__SUNOS__)
// dialup device under SunOS/Solaris
- hasModem = strstr(output,"ipdptp") != (char *)NULL;
- hasLAN = strstr(output, "hme") != (char *)NULL;
+ hasModem = strstr(output.fn_str(),"ipdptp") != (char *)NULL;
+ hasLAN = strstr(output.fn_str(), "hme") != (char *)NULL;
#elif defined(__LINUX__) || defined (__FREEBSD__)
hasModem = strstr(output.fn_str(),"ppp") // ppp
|| strstr(output.fn_str(),"sl") // slip
However,i10n is condition bad.:-(
This environment isn't satisfied even if patch of Andreas is applied.
Can i18n cope with solaris9?
regards,
Hiroshi Saito