Thread: Vote: Adding flex/bison derived files in WIN32_DEV
Because MinGW/Msys doesn't come with flex/bison by default, I have added those derived files to the WIN32_DEV branch in CVS. It makes it easier for people to install _just_ MinGW and compile PostgreSQL on Win32. The branch will live for only 1-2 months until we start 7.5 development. Those files will not be moved into the main branch. Should those files be in WIN32_DEV CVS, or should they be removed, and require people to install bison/flex on MinGW. Even though they will be updated infrequently, there is concern about CVS file bloat. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001+ If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania19073
On Wed, 10 Sep 2003, Bruce Momjian wrote: > Because MinGW/Msys doesn't come with flex/bison by default, I have added > those derived files to the WIN32_DEV branch in CVS. It makes it easier > for people to install _just_ MinGW and compile PostgreSQL on Win32. The > branch will live for only 1-2 months until we start 7.5 development. > Those files will not be moved into the main branch. > > Should those files be in WIN32_DEV CVS, or should they be removed, and > require people to install bison/flex on MinGW. Even though they will > be updated infrequently, there is concern about CVS file bloat. I think a more appropriate question is how many ppl are working on WIN32_DEV *from* CVS that don't have flex/bison available ... if nobody, having those files in CVS is *really* useless and the whole argument is moot ...
Marc G. Fournier wrote: > > On Wed, 10 Sep 2003, Bruce Momjian wrote: > > > Because MinGW/Msys doesn't come with flex/bison by default, I have added > > those derived files to the WIN32_DEV branch in CVS. It makes it easier > > for people to install _just_ MinGW and compile PostgreSQL on Win32. The > > branch will live for only 1-2 months until we start 7.5 development. > > Those files will not be moved into the main branch. > > > > Should those files be in WIN32_DEV CVS, or should they be removed, and > > require people to install bison/flex on MinGW. Even though they will > > be updated infrequently, there is concern about CVS file bloat. > > I think a more appropriate question is how many ppl are working on > WIN32_DEV *from* CVS that don't have flex/bison available ... if nobody, > having those files in CVS is *really* useless and the whole argument is > moot ... I don't know. We would have to ask on the Win32 list, but the files were added specifically because several people asked about those missing files, and didn't/couldn't get bison/flex. Now that we have snapshots, I don't know how many have switched to those. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001+ If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania19073
Bruce Momjian wrote: >Because MinGW/Msys doesn't come with flex/bison by default, I have added >those derived files to the WIN32_DEV branch in CVS. It makes it easier >for people to install _just_ MinGW and compile PostgreSQL on Win32. The >branch will live for only 1-2 months until we start 7.5 development. >Those files will not be moved into the main branch. > >Should those files be in WIN32_DEV CVS, or should they be removed, and >require people to install bison/flex on MinGW. Even though they will >be updated infrequently, there is concern about CVS file bloat. > > > I vote for requiring MinGW+flex+bison. On *nix you need a similar combination, so why should MinGW be different? If installing flex and bison on MinGW is too hard you probably shouldn't be building out of CVS, ISTM. cheers andrew
On Wed, 2003-09-10 at 12:03, Bruce Momjian wrote: > Because MinGW/Msys doesn't come with flex/bison by default, I have added > those derived files to the WIN32_DEV branch in CVS. I'm confused. Right on the MinGW download page is a link for bison-1.875.
On Thursday 11 September 2003 17:30, Matthew T. O'Connor wrote: > On Wed, 2003-09-10 at 12:03, Bruce Momjian wrote: > > Because MinGW/Msys doesn't come with flex/bison by default, I have added > > those derived files to the WIN32_DEV branch in CVS. > > I'm confused. Right on the MinGW download page is a link for > bison-1.875. > > Yep, but I had problems with it. Author confirmed that there could be some problems creating processes (for example M4). However if You make it work, I'll be interested to know how. Check the MinGW mailing list for last few weeks, and see what I'm talking about. By the way the last tip of the author : "reconfigure & make", was not working for me :-( Regards !
> > I'm confused. Right on the MinGW download page is a link for bison-1.875. > > > > > Yep, but I had problems with it. Author confirmed that there could be some > problems creating processes (for example M4). However if You make it work, > I'll be interested to know how. Check the MinGW mailing list for last few > weeks, and see what I'm talking about. By the way the last tip of the > author : "reconfigure & make", was not working for me :-( Download the bin package at http://prdownloads.sf.net/mingw/bison-1.875.0-2003.02.10-1.exe?download and use it. You do not need to compile bison yourself. Andreas
On Thursday 11 September 2003 18:00, Zeugswetter Andreas SB SD wrote: > > > I'm confused. Right on the MinGW download page is a link for > > > bison-1.875. > > > > Yep, but I had problems with it. Author confirmed that there could be > > some problems creating processes (for example M4). However if You make it > > work, I'll be interested to know how. Check the MinGW mailing list for > > last few weeks, and see what I'm talking about. By the way the last tip > > of the author : "reconfigure & make", was not working for me :-( > > Download the bin package at > http://prdownloads.sf.net/mingw/bison-1.875.0-2003.02.10-1.exe?download > and use it. You do not need to compile bison yourself. > > Andreas Did You actually succeed to make output from postgresql yacc definition files with that package ? Can You explain how? Regards !
On Thursday 11 September 2003 18:08, Darko Prenosil wrote: > On Thursday 11 September 2003 18:00, Zeugswetter Andreas SB SD wrote: > > > > I'm confused. Right on the MinGW download page is a link for > > > > bison-1.875. > > > > > > Yep, but I had problems with it. Author confirmed that there could be > > > some problems creating processes (for example M4). However if You make > > > it work, I'll be interested to know how. Check the MinGW mailing list > > > for last few weeks, and see what I'm talking about. By the way the last > > > tip of the author : "reconfigure & make", was not working for me :-( > > > > Download the bin package at > > http://prdownloads.sf.net/mingw/bison-1.875.0-2003.02.10-1.exe?download > > and use it. You do not need to compile bison yourself. > > > > Andreas > > Did You actually succeed to make output from postgresql yacc definition > files with that package ? Can You explain how? >I forgot to mention that I succeed to make that output, but only from shell, not from makefile. Maybe there is something wrong with my MinGW setup ? However I do not like Idea that people might give up MinGW bison only because I misinformed the group. Regards !
Andrew Dunstan wrote: > Bruce Momjian wrote: > > >Because MinGW/Msys doesn't come with flex/bison by default, I have added > >those derived files to the WIN32_DEV branch in CVS. It makes it easier > >for people to install _just_ MinGW and compile PostgreSQL on Win32. The > >branch will live for only 1-2 months until we start 7.5 development. > >Those files will not be moved into the main branch. > > > >Should those files be in WIN32_DEV CVS, or should they be removed, and > >require people to install bison/flex on MinGW. Even though they will > >be updated infrequently, there is concern about CVS file bloat. > > > > > > > I vote for requiring MinGW+flex+bison. On *nix you need a similar > combination, so why should MinGW be different? If installing flex and > bison on MinGW is too hard you probably shouldn't be building out of > CVS, ISTM. OK, removed. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001+ If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania19073
What is flex/bison ? ""Marc G. Fournier"" <scrappy@postgresql.org> wrote in message news:20030910133350.A57860@ganymede.hub.org... > > On Wed, 10 Sep 2003, Bruce Momjian wrote: > > > Because MinGW/Msys doesn't come with flex/bison by default, I have added > > those derived files to the WIN32_DEV branch in CVS. It makes it easier > > for people to install _just_ MinGW and compile PostgreSQL on Win32. The > > branch will live for only 1-2 months until we start 7.5 development. > > Those files will not be moved into the main branch. > > > > Should those files be in WIN32_DEV CVS, or should they be removed, and > > require people to install bison/flex on MinGW. Even though they will > > be updated infrequently, there is concern about CVS file bloat. > > I think a more appropriate question is how many ppl are working on > WIN32_DEV *from* CVS that don't have flex/bison available ... if nobody, > having those files in CVS is *really* useless and the whole argument is > moot ... > > > ---------------------------(end of broadcast)--------------------------- > TIP 2: you can get off all lists at once with the unregister command > (send "unregister YourEmailAddressHere" to majordomo@postgresql.org) >
After takin a swig o' Arrakan spice grog, "Relaxin" <noname@spam.com> belched out...: > What is flex/bison ? Bison is the GNU Project parser generator, in the style of yacc (Yet Another Compiler Compiler). Flex is the GNU Project "fast lexical analyzer generator," in the style of lex. These are used to generate parser code that knows how to recognize (in this case) the SQL language. Obviously, an SQL database must include some way of reading through queries and parsing out whether they are valid, and if they are, what the database is supposed to do with them. Flex and Bison do that part of the work. -- "cbbrowne","@","acm.org" http://cbbrowne.com/info/x.html Rules of the Evil Overlord #212. "I will not send out battalions composed wholly of robots or skeletons against heroes who have qualms about killing living beings. <http://www.eviloverlord.com/>
After installing the following into my msys/mingw set: http://prdownloads.sourceforge.net/mingw/bison-1.875.0-2003.02.10-1.exe?download and http://www.cs.colorado.edu/~main/mingw32/flex-2_5_4a.zip (Not sure if this was needed, but I did it anyway) and then running "configure --without-zlib --without-readline", I was able to "make" the branch right through to the stage where it fails to link because of missing fork/signmask/etc. Bison ran just fine, apparently. If you do this make sure you install everything in default locations - the bison binary has a hardcoded location for m4 (yuck). cheers andrew ----- Original Message ----- From: "Darko Prenosil" <darko.prenosil@finteh.hr> > On Thursday 11 September 2003 18:08, Darko Prenosil wrote: > > On Thursday 11 September 2003 18:00, Zeugswetter Andreas SB SD wrote: > > > > > I'm confused. Right on the MinGW download page is a link for > > > > > bison-1.875. > > > > > > > > Yep, but I had problems with it. Author confirmed that there could be > > > > some problems creating processes (for example M4). However if You make > > > > it work, I'll be interested to know how. Check the MinGW mailing list > > > > for last few weeks, and see what I'm talking about. By the way the last > > > > tip of the author : "reconfigure & make", was not working for me :-( > > > > > > Download the bin package at > > > http://prdownloads.sf.net/mingw/bison-1.875.0-2003.02.10-1.exe?download > > > and use it. You do not need to compile bison yourself. > > > > > > Andreas > > > > Did You actually succeed to make output from postgresql yacc definition > > files with that package ? Can You explain how? > > > I forgot to mention that I succeed to make that output, but only from shell, > not from makefile. Maybe there is something wrong with my MinGW setup ? > However I do not like Idea that people might give up MinGW bison only because > I misinformed the group. >