Thread: Naming of the PostreSQL ODBC .dll file
I can understand why you changed the name of the DLL to psqlodbc.dll the trouble with that though is it gives no sense of 16 vs 32 bit .dll. The podbc32.dll did such... Can we agree on a *32.dll name? And this weekend when I'm working on everything else I have I'll do a *16.dll Julie -- [ Julia Anne Case ] [ Ships are safe inside the harbor, ] [Programmer at large] [ but is that what ships are really for. ] [ Admining Linux ] [ To thine own self be true. ] [ Windows/WindowsNT ] [ Fair is where you take your cows to be judged. ]
Julia, That was one of the biggest changes with this driver. It no longer supports 16 bit windows. All the messy Dummy Handle Lists that used to be mainted were eliminated and now the handle is the pointer to the structure for all the Statement functions. The old driver used GlobalAlloc and GlobalLock to handle 16 bit because the pointer returned by malloc may be moved and is not sufficient to be used for a handle. In 32 bit windows, you can safely use malloc. Our web site explains this. I sure hope this driver doesn't have to support 16 bit! Byron Julia A.Case wrote: > I can understand why you changed the name of the DLL to > psqlodbc.dll the trouble with that though is it gives no sense of 16 vs 32 > bit .dll. The podbc32.dll did such... Can we agree on a *32.dll name? > And this weekend when I'm working on everything else I have I'll do a > *16.dll > > Julie > > -- > [ Julia Anne Case ] [ Ships are safe inside the harbor, ] > [Programmer at large] [ but is that what ships are really for. ] > [ Admining Linux ] [ To thine own self be true. ] > [ Windows/WindowsNT ] [ Fair is where you take your cows to be judged. ]
Quoting Byron Nikolaidis (byronn@insightdist.com):
> That was one of the biggest changes with this driver. It no longer supports
> 16 bit windows. All the messy Dummy Handle Lists that used to be mainted were
> eliminated and now the handle is the pointer to the structure for all the
> Statement functions. The old driver used GlobalAlloc and GlobalLock to handle
> 16 bit because the pointer returned by malloc may be moved and is not
> sufficient to be used for a handle. In 32 bit windows, you can safely use
> malloc.
>
The last time I made mumblings of droping 16 bit support I got
roasted pretty good. There seem to still be a lot of 16 bit users out
there that would like to use the driver.
Julie
--
[ Julia Anne Case ] [ Ships are safe inside the harbor, ]
[Programmer at large] [ but is that what ships are really for. ]
[ Admining Linux ] [ To thine own self be true. ]
[ Windows/WindowsNT ] [ Fair is where you take your cows to be judged. ]
Julie, Well, the driver is now 32 bit and 16 bit support would be very difficult to add and would degrade the driver's performance and supportability, as I already stated. I think it would be moving backwards for this development effort. We haven't heard any complaints yet about the lack of 16 bit support. Why dont we just keep going in a forward direction and see what happens. It seems to work for Bill Gates. He's no friend of mine, but he sure knows how to get software to market. Byron Julia A.Case wrote: > Quoting Byron Nikolaidis (byronn@insightdist.com): > > That was one of the biggest changes with this driver. It no longer supports > > 16 bit windows. All the messy Dummy Handle Lists that used to be mainted were > > eliminated and now the handle is the pointer to the structure for all the > > Statement functions. The old driver used GlobalAlloc and GlobalLock to handle > > 16 bit because the pointer returned by malloc may be moved and is not > > sufficient to be used for a handle. In 32 bit windows, you can safely use > > malloc. > > > The last time I made mumblings of droping 16 bit support I got > roasted pretty good. There seem to still be a lot of 16 bit users out > there that would like to use the driver. > > Julie
On Fri, 17 Apr 1998, Byron Nikolaidis wrote:
> Julie,
>
> Well, the driver is now 32 bit and 16 bit support would be very difficult to add and
> would degrade the driver's performance and supportability, as I already stated. I
> think it would be moving backwards for this development effort.
>
> We haven't heard any complaints yet about the lack of 16 bit support. Why dont we
> just keep going in a forward direction and see what happens. It seems to work for
> Bill Gates. He's no friend of mine, but he sure knows how to get software to
> market.
Yeah, but look what he turns out.
Vince.
--
==========================================================================
Vince Vielhaber -- KA8CSH email: vev@michvhf.com flame-mail: /dev/null
# include <std/disclaimers.h> TEAM-OS2
Online Searchable Campground Listings http://www.camping-usa.com
"I'm just not a fan of promoting stupidity!
We have elected officials for that job!" -- Rock
==========================================================================
On Fri, 17 Apr 1998 julie@hub.org wrote: > The last time I made mumblings of droping 16 bit support I got > roasted pretty good. There seem to still be a lot of 16 bit users out > there that would like to use the driver. > > Julie Could someone forward this to the old list; I'm sorry that I do not remember the address. It is just that I know not all members of the old PostODBC list have migrated to this one, and there may be many of those that have viewpoints on this issue. I, myself, would not have a problem with dropping 16bit; however, I realize not all have windoze users have buckled under the enourmous pressure from MS to convert to 95. Scott
On Fri, 17 Apr 1998, Scott Tomer x6214 Austin wrote:
> On Fri, 17 Apr 1998 julie@hub.org wrote:
>
> > The last time I made mumblings of droping 16 bit support I got
> > roasted pretty good. There seem to still be a lot of 16 bit users out
> > there that would like to use the driver.
> >
> > Julie
>
> Could someone forward this to the old list; I'm sorry that I do not remember
> the address. It is just that I know not all members of the old PostODBC
> list have migrated to this one, and there may be many of those that have
> viewpoints on this issue.
>
> I, myself, would not have a problem with dropping 16bit; however, I realize
> not all have windoze users have buckled under the enourmous pressure from
> MS to convert to 95.
Besides a cocky remark every now and then (don't even need ale for that :)
I've been avoiding most of this discussion. Y'know M$ came up with w95 a
few years ago and you'd think that by now everyone migrated. It's simply
not true. Alot of businesses have decided that 95 isn't for them and that
3.1 is stable enough for their operation and that's that.
Then there's another thing that many people neglect to notice. There's a
TON of people out there that are running OS/2 which does 3.1 only. I did
the port (not that there was alot of work to do that) of psql and libpq to
OS/2. You'd be surprised how many people are still running that operating
system that M$ has repeatedly declared dead. But the fact remains that it
runs 16 bit windows only unless someone does the port (using Open32). If
you want a diverse crowd using it - a crowd that may very well find enough
of an interest in it to help (which seems to be a shortage these days) in
the development - then you sure don't wanna start by isolating perspective
users, do ya?
Possibility for the FAQ. Make the "official site" www.postgresql.org and
use wgets in a cron job to keep it up. That way the most current copy is
always on the main site. If for some reason the keeper can't keep up at
least it's got a current rendition at the main site for someone else to
pick up and run with. Also that way Julie can keep it up at Magenet and
have it local for easy updates. (note: assumptions taken in that last
statement)
Vince.
--
==========================================================================
Vince Vielhaber -- KA8CSH email: vev@michvhf.com flame-mail: /dev/null
# include <std/disclaimers.h> TEAM-OS2
Online Searchable Campground Listings http://www.camping-usa.com
"I'm just not a fan of promoting stupidity!
We have elected officials for that job!" -- Rock
==========================================================================
On Fri, 17 Apr 1998, Byron Nikolaidis wrote:
> Julia,
>
> That was one of the biggest changes with this driver. It no longer supports
> 16 bit windows. All the messy Dummy Handle Lists that used to be mainted were
> eliminated and now the handle is the pointer to the structure for all the
> Statement functions. The old driver used GlobalAlloc and GlobalLock to handle
> 16 bit because the pointer returned by malloc may be moved and is not
> sufficient to be used for a handle. In 32 bit windows, you can safely use
> malloc.
>
> Our web site explains this.
>
> I sure hope this driver doesn't have to support 16 bit!
Hrmmm...just curious, but how many users were lost because of
*this* decision?? Doesn't this sort of assume that *everyone* is running
Win95 or greater?
That's sort of like saying we don't support AIX 3.5 cause AIX 4.1
is out and does things a bit cleaner/differently :(
Marc G. Fournier
Systems Administrator @ hub.org
primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org
On Fri, 17 Apr 1998, Byron Nikolaidis wrote:
> We haven't heard any complaints yet about the lack of 16 bit support.
Actually, you have...Julie just voiced one...:(
> Why dont we just keep going in a forward direction and see what happens.
> It seems to work for Bill Gates. He's no friend of mine, but he sure
> knows how to get software to market.
Ya, he shoves the software down ppls throat even if it isn't very
good...IMHO, a sad day for freeware when we follow Bill's example :(
Marc G. Fournier
Systems Administrator @ hub.org
primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org
On Fri, 17 Apr 1998, Scott Tomer x6214 Austin wrote:
> I, myself, would not have a problem with dropping 16bit; however, I realize
> not all have windoze users have buckled under the enourmous pressure from
> MS to convert to 95.
Hell, I know ppl that haven't bucked under enornous pressure from
MS to cnvert to Windows, period :(
Marc G. Fournier
Systems Administrator @ hub.org
primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org
> > It no longer supports 16 bit windows. > > I sure hope this driver doesn't have to support 16 bit! > Hrmmm...just curious, but how many users were lost because of > *this* decision?? Doesn't this sort of assume that *everyone* is > running Win95 or greater? > That's sort of like saying we don't support AIX 3.5 cause AIX 4.1 > is out and does things a bit cleaner/differently :( Actually that sounds like a really good idea; we should have thought of that before. AIX-3.5 is a real pain ;-)
On Sat, 18 Apr 1998, Thomas G. Lockhart wrote:
> > > It no longer supports 16 bit windows.
> > > I sure hope this driver doesn't have to support 16 bit!
> > Hrmmm...just curious, but how many users were lost because of
> > *this* decision?? Doesn't this sort of assume that *everyone* is
> > running Win95 or greater?
> > That's sort of like saying we don't support AIX 3.5 cause AIX 4.1
> > is out and does things a bit cleaner/differently :(
>
> Actually that sounds like a really good idea; we should have thought of
> that before. AIX-3.5 is a real pain ;-)
I hope that was a *really* sarcastic grin? :) Could you
imagine...let's just drop support for every old version of an operating
system. That one would go over well...
We only support:
FreeBSD 3.0
Solaris 2.6 (sparc/x86)
AIX 4.1
Linux <insert this weeks version>
That one would go over *really* well, eh?
Marc G. Fournier
Systems Administrator @ hub.org
primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org
> That one would go over well... > > We only support: > > FreeBSD 3.0 > Solaris 2.6 (sparc/x86) > AIX 4.1 > Linux <insert this weeks version> Yeah, and I'm not too sure about FreeBSD, Solaris, and AIX either :))
On Fri, 17 Apr 1998, Vince Vielhaber wrote:
> On Fri, 17 Apr 1998, Scott Tomer x6214 Austin wrote:
>
> > On Fri, 17 Apr 1998 julie@hub.org wrote:
> >
> > > The last time I made mumblings of droping 16 bit support I got
> > > roasted pretty good. There seem to still be a lot of 16 bit users out
> > > there that would like to use the driver.
> > >
> > > Julie
> >
> > Could someone forward this to the old list; I'm sorry that I do not remember
> > the address. It is just that I know not all members of the old PostODBC
> > list have migrated to this one, and there may be many of those that have
> > viewpoints on this issue.
> >
> > I, myself, would not have a problem with dropping 16bit; however, I realize
> > not all have windoze users have buckled under the enourmous pressure from
> > MS to convert to 95.
>
> Besides a cocky remark every now and then (don't even need ale for that :)
> I've been avoiding most of this discussion. Y'know M$ came up with w95 a
> few years ago and you'd think that by now everyone migrated. It's simply
> not true. Alot of businesses have decided that 95 isn't for them and that
> 3.1 is stable enough for their operation and that's that.
... CUT ...
Agree!
I'm one (of the last?) who uses 16-bit too.
I know several companies (esp. with many PC's) which stay still in the
16-bit world.
If it is *really* a problem to support 16-bit AND 32-bit in one
source tree (I didn't see the great problems), a practically way could be
to split it into a 16-bit and 32-bit tree. Anyone who needs a 16-bit
version could (should) support the developement.
IMHO one the *greatest* advantages of PostgreSQL is the extremely wide
range of supported platforms and interfaces!
Hoping OBC16 does NOT die,
Gerhard
+-----------------+ +--- gerhardr@tech-edv.co.at ---+
| Technische EDV \ Reithofer / Technical Sofware Developement |
| A-2136 Laa/Thaya \ Gerhard / Tel +43-2522/8726 +-------------+
| Staatsbahnstr. 100 +-------+ Fax +43-2522/87268 |
+----- http://members.aon.at/tech-edv/Info -------+
The Hermit Hacker wrote: > > On Fri, 17 Apr 1998, Byron Nikolaidis wrote: > > > Julia, > > > > That was one of the biggest changes with this driver. It no longer supports > > 16 bit windows. All the messy Dummy Handle Lists that used to be mainted were > > eliminated and now the handle is the pointer to the structure for all the > > Statement functions. The old driver used GlobalAlloc and GlobalLock to handle > > 16 bit because the pointer returned by malloc may be moved and is not > > sufficient to be used for a handle. In 32 bit windows, you can safely use > > malloc. > > > > Our web site explains this. > > > > I sure hope this driver doesn't have to support 16 bit! > > Hrmmm...just curious, but how many users were lost because of > *this* decision?? Doesn't this sort of assume that *everyone* is running > Win95 or greater? None, if *this* decision has anything to do with the better quality of the current driver. I guess that one working 32 bit driver gets us more users than largely broken ones for both 32 and 16. It may be a good idea to keep the old 16 bit driver around, but I can't see anyone motivated enough to actually do it ;( Hannu