Hey all,
I'm using Fedora 8 on my laptop. The library known as "libstdc++-libc6.1-1.so.2" isn't there, which I need to run Maple 7 on my computer. This should be a part of the compat-libstdc++-296 RPM, as this library was originally a part of RHL 7.3 and 8.0 in libstdc++, and available in the compat-libstdc++-7.3-2.96.118 RPM for RHL 9. Does anybody know why this file was removed from the compat-libstdc++-296 RPM, and what I can do to get it from the SRPM again? (I'm guessing theoretcially it should still be part of the source code.)
Thanks William
William M. Quarles wrote:
I'm using Fedora 8 on my laptop. The library known as "libstdc++-libc6.1-1.so.2" isn't there, which I need to run Maple 7 on my computer. This should be a part of the compat-libstdc++-296 RPM, as this library was originally a part of RHL 7.3 and 8.0 in libstdc++, and available in the compat-libstdc++-7.3-2.96.118 RPM for RHL 9. Does anybody know why this file was removed from the compat-libstdc++-296 RPM, and what I can do to get it from the SRPM again? (I'm guessing theoretcially it should still be part of the source code.)
Apparently that library was intentionally removed: https://bugzilla.redhat.com/show_bug.cgi?id=225371
...which seems odd. I can't see how the compat-libstdc++-296 package actually provides compatibility with binaries that were built with gcc 2.96. It appears to be useless? Brilliant!
Gordon Messmer wrote:
Apparently that library was intentionally removed: https://bugzilla.redhat.com/show_bug.cgi?id=225371
...which seems odd. I can't see how the compat-libstdc++-296 package actually provides compatibility with binaries that were built with gcc 2.96. It appears to be useless? Brilliant!
Sent too soon: I meant to suggest that you get the compat-libstdc++ from an older release of Fedora. Try FC5.
Gordon Messmer <yinyang <at> eburg.com> writes:
...which seems odd. I can't see how the compat-libstdc++-296 package actually provides compatibility with binaries that were built with gcc 2.96.
It does. libstdc++-libc6.2-2.so.3 is the version which actually shipped with GCC 2.96. (At least the RH one. There was no formal 2.96 release, so other distributions may have been shipping other 2.96 variants with different libraries.)
The missing libstdc++-libc6.1-1.so.2 is actually from GCC 2.95 (or EGCS, which was the project which lead to GCC 2.95 and used the same soname).
Upstreams still building their binaries with GCC 2.95 (or 2.96 for that matter) should really be told to get with the times. GCC is at 4.3 now, 2.95 is just a long gone memory from the distant past.
Kevin Kofler
Kevin Kofler wrote:
Gordon Messmer <yinyang <at> eburg.com> writes:
...which seems odd. I can't see how the compat-libstdc++-296 package actually provides compatibility with binaries that were built with gcc 2.96.
It does. libstdc++-libc6.2-2.so.3 is the version which actually shipped with GCC 2.96. (At least the RH one. There was no formal 2.96 release, so other distributions may have been shipping other 2.96 variants with different libraries.)
The missing libstdc++-libc6.1-1.so.2 is actually from GCC 2.95 (or EGCS, which was the project which lead to GCC 2.95 and used the same soname).
Upstreams still building their binaries with GCC 2.95 (or 2.96 for that matter) should really be told to get with the times. GCC is at 4.3 now, 2.95 is just a long gone memory from the distant past.
Unfortunately people and companies who have critical commercial software for which upgrades are unavailable or unafordable remember those days well. I have FC4 and RH8 running in virtual machines for that very reason.
I agree that no one should be building with those libraries, but in fact people do use binaries already using them.
On Thu, 23 Oct 2008 14:47:57 -0400 Bill Davidsen davidsen@tmr.com wrote:
I agree that no one should be building with those libraries, but in fact people do use binaries already using them.
Yea. For instance, I'm sticking with a very old version of nxclient/server because it works flawlessly for what I want to do and I have no idea if I'd run into problems with an upgrade. (Always a risk when you have clients and servers on different machines with different update schedules and administrators involved).
I've just started making my own "compat" rpm with libraries from older fedora versions packaged up to be installed on the new fedora to allow my older binaries to run. If it ever reaches the point where that isn't good enough, I guess I'll have to investigate upgrading after all :-).
Tom Horsley <tom.horsley <at> att.net> writes:
Yea. For instance, I'm sticking with a very old version of nxclient/server because it works flawlessly for what I want to do and I have no idea if I'd run into problems with an upgrade.
An old version of software which talks over the network? You DO realize that this is a major security risk? Even more so when you package old versions of random system libraries with no security updates either to get it running.
Kevin Kofler
Tom Horsley wrote:
On Thu, 23 Oct 2008 14:47:57 -0400 Bill Davidsen davidsen@tmr.com wrote:
I agree that no one should be building with those libraries, but in fact people do use binaries already using them.
Yea. For instance, I'm sticking with a very old version of nxclient/server because it works flawlessly for what I want to do and I have no idea if I'd run into problems with an upgrade. (Always a risk when you have clients and servers on different machines with different update schedules and administrators involved).
I've just started making my own "compat" rpm with libraries from older fedora versions packaged up to be installed on the new fedora to allow my older binaries to run. If it ever reaches the point where that isn't good enough, I guess I'll have to investigate upgrading after all :-).
You might think about where to release that for other to find.
Bill Davidsen wrote:
Tom Horsley wrote:
On Thu, 23 Oct 2008 14:47:57 -0400 Bill Davidsen davidsen@tmr.com wrote:
I agree that no one should be building with those libraries, but in fact people do use binaries already using them.
<snip>
I've just started making my own "compat" rpm with libraries from older fedora versions packaged up to be installed on the new fedora to allow my older binaries to run. If it ever reaches the point where that isn't good enough, I guess I'll have to investigate upgrading after all :-).
You might think about where to release that for other to find.
WORD! Ditto on that one, that could be really useful to a number of people. You might want to see who would be willing to post that for you on a site for others to download.
Peace, William
Tom Horsley wrote:
I've just started making my own "compat" rpm with libraries from older fedora versions packaged up to be installed on the new fedora to allow my older binaries to run. If it ever reaches the point where that isn't good enough, I guess I'll have to investigate upgrading after all :-).
Think you could send me some of those? I'm running Fedora 8 right now, but I'm about ready to switch to Fedora 9 thanks to all of the complications of wireless networking that are fixed in Fedora 9. What version of Fedora are you running?
Peace, William
On Sun, 09 Nov 2008 09:35:08 -0500 "William M. Quarles" walrus@bellsouth.net wrote:
Think you could send me some of those?
I doubt it is generally useful, right now it only has /lib/libcrypto.so.0.9.8b in it (which the old version of NX I'm sticking to needs). I guess they switched to something other than openssl in the new release, so the library NX looks for isn't there anymore.
Bill Davidsen <davidsen <at> tmr.com> writes:
Unfortunately people and companies who have critical commercial software for which upgrades are unavailable or unafordable remember those days well.
That's exactly why relying on proprietary software is a bad idea.
Kevin Kofler
Kevin Kofler wrote:
Bill Davidsen <davidsen <at> tmr.com> writes:
Unfortunately people and companies who have critical commercial software for which upgrades are unavailable or unafordable remember those days well.
That's exactly why relying on proprietary software is a bad idea.
That's a really offensive comment. It assumes that business people who use commercial software are just a bunch of clueless assholes. Writing a major applications (a) can cost an order of magnitude more than buying, (b) delays the solution, sometimes for years, and (c) often requires information about hardware which just isn't available. If it weren't for proprietary software many things would never get done at all.
Leave business to business people, who have to do things on time and under budget, and who choose achievable solutions. I don't know what you're good at, but tact and business reality doesn't seem to be in your skill set.
And I'm not much pleased with your counterparts to say that's why relying on open source software is bad. A different set of smug, opinionated blowhards, I heard from them elsewhere on the same topic.
Note: I rarely offend people by accident, if you don't like the tone of my reply, be assured that was my intention, I really didn't like the implications of yours.
Bill Davidsen <davidsen <at> tmr.com> writes:
That's a really offensive comment. It assumes that business people who use commercial software are just a bunch of clueless assholes.
No, it just says that they didn't consider the consequences of using proprietary software when they made their decision and they got what they deserved.
It doesn't mean they are "assholes" or even necessarily "clueless", just that they have no right to complain because that situation is of their own chosing.
Writing a major applications (a) can cost an order of magnitude more than buying, (b) delays the solution, sometimes for years, and (c) often requires information about hardware which just isn't available. If it weren't for proprietary software many things would never get done at all.
Using Free Software does not necessarily mean writing it from scratch, often there's an existing Free Software project one can use and/or contribute to.
I can understand using a proprietary tool if there's really nothing available in Free Software, but I doubt this is the majority of the cases.
Leave business to business people, who have to do things on time and under budget, and who choose achievable solutions. I don't know what you're good at, but tact and business reality doesn't seem to be in your skill set.
And this is a completely unwarranted personal attack. Do you really have to resort to ad hominem attacks? It shows that you're running out of arguments.
All I'm saying is that the ability to rebuild for newer versions of system libraries is a very practical argument for choosing Free Software. Face it, GNU/Linux is not going to be binary-compatible forever, not even proprietary OSes do that (just look at how Apple dropped Classic and is now phasing out Carbon, Mac OS is not even source compatible with older versions, let alone binary!), maybe with the exception of one (and I guess I don't have to tell you which one... its the one whose legacy cruft impedes progress significantly, for example it's the only major modern non-embedded OS still not using UTF-8 as the default 8-bit charset). Being able to rebuild your software against newer libraries is essential for continued use of it.
Kevin Kofler
On Fri, Oct 24, 2008 at 8:41 PM, Kevin Kofler kevin.kofler@chello.at wrote:
Bill Davidsen <davidsen <at> tmr.com> writes:
That's a really offensive comment. It assumes that business people who use commercial software are just a bunch of clueless assholes.
No, it just says that they didn't consider the consequences of using proprietary software when they made their decision and they got what they deserved.
It doesn't mean they are "assholes" or even necessarily "clueless", just that they have no right to complain because that situation is of their own chosing.
Writing a major applications (a) can cost an order of magnitude more than buying, (b) delays the solution, sometimes for years, and (c) often requires information about hardware which just isn't available. If it weren't for proprietary software many things would never get done at all.
Using Free Software does not necessarily mean writing it from scratch, often there's an existing Free Software project one can use and/or contribute to.
I can understand using a proprietary tool if there's really nothing available in Free Software, but I doubt this is the majority of the cases.
Leave business to business people, who have to do things on time and under budget, and who choose achievable solutions. I don't know what you're good at, but tact and business reality doesn't seem to be in your skill set.
And this is a completely unwarranted personal attack. Do you really have to resort to ad hominem attacks? It shows that you're running out of arguments.
All I'm saying is that the ability to rebuild for newer versions of system libraries is a very practical argument for choosing Free Software. Face it, GNU/Linux is not going to be binary-compatible forever, not even proprietary OSes do that (just look at how Apple dropped Classic and is now phasing out Carbon, Mac OS is not even source compatible with older versions, let alone binary!), maybe with the exception of one (and I guess I don't have to tell you which one... its the one whose legacy cruft impedes progress significantly, for example it's the only major modern non-embedded OS still not using UTF-8 as the default 8-bit charset). Being able to rebuild your software against newer libraries is essential for continued use of it.
Kevin Kofler
First, do not confuse "Free" with "Open". They are not the same.
Let's not stretch advocacy to ridiculous levels. Open source has limits. It's nice to have the source code. However, the source code does no good if you have the neither the resources and/or the skills to do something with it.
Kam Leo <kam.leo <at> gmail.com> writes:
First, do not confuse "Free" with "Open". They are not the same.
Indeed, they're not. "Free Software" is the correct term: http://www.gnu.org/philosophy/open-source-misses-the-point.html And Free Software it's about freedom, not price.
Let's not stretch advocacy to ridiculous levels. Open source has limits. It's nice to have the source code. However, the source code does no good if you have the neither the resources and/or the skills to do something with it.
How hard is it to run "make"? Often just recompiling is enough to make the software work on a current distribution. And if it does not build, fixing it is often not rocket science either.
Kevin Kofler
On Sat, Oct 25, 2008 at 3:43 PM, Kevin Kofler kevin.kofler@chello.at wrote:
Kam Leo <kam.leo <at> gmail.com> writes:
First, do not confuse "Free" with "Open". They are not the same.
Indeed, they're not. "Free Software" is the correct term: http://www.gnu.org/philosophy/open-source-misses-the-point.html And Free Software it's about freedom, not price.
There are "BSD", "GNU", "beer", "public domain", etc. styles of freedom. When used loosely "Free Software" can also refer to the "beer" variety. The "beer" variety usually does not provide source code.
Let's not stretch advocacy to ridiculous levels. Open source has limits. It's nice to have the source code. However, the source code does no good if you have the neither the resources and/or the skills to do something with it.
How hard is it to run "make"? Often just recompiling is enough to make the software work on a current distribution. And if it does not build, fixing it is often not rocket science either.
Kevin Kofler
Running "make" is the easy part. Fixing becomes difficult when the kernel, compiler or library goes through a significant change. In that situation you have to know what changed in order to deal with it.
Kevin Kofler wrote:
Upstreams still building their binaries with GCC 2.95 (or 2.96 for that matter) should really be told to get with the times. GCC is at 4.3 now, 2.95 is just a long gone memory from the distant past.
Sorry for taking so long to reply, for some reason my Gmane.org feed wasn't showing the latest replies on this thread.
OK, Windows XP still runs nearly all programs from previous versions Windows, plus it has DOS emulation so that it can run many, but not all DOS programs. What is SO WRONG with some element of reverse compatibility? I know that some of you may feel that there is a hinder to progress there, but there has to be some kind of balance between bleeding-edge and interoperability with other software.
I'm not talking about a need to build new binaries, I'm just talking about getting older software to run on a newer OS. This doesn't just include commercial software such as Maple, but also older open-source projects that haven't been updated in a while, but could still hypothetically work if the proper libraries were provided.
William
OK, so I said the phrase "Windows XP" in a message, and suddenly NO ONE will reply? Seriously, I'll say it again, Microsoft and Sony (e.g. most of the PlayStation series) can do reverse compatibility fairly well, and Fedora is almost totally lacking in that arena. "Get with the times" doesn't always make sense. And if anybody says, "Well, if you like reverse compatibility so much, why don't you shell out for Microsoft software like 90% of the rest of the herd." Uh, no. There are both commercial and "free" software products that need that reverse compatibility.
Example: is there anybody out there doing natural science or engineering on Linux machines right now who is NOT using any commercial software whatsoever?
And just because a "free" software product hasn't been updated in a while does not mean that the software is useless.
I can't quadruple boot my machine just to run all of the software I use. It should not be necessary, especially when most of this software is designed for Red Hat/Fedora distributions, or can be installed and run using WINE.
Any thoughts?
William
William M. Quarles wrote:
Kevin Kofler wrote:
Upstreams still building their binaries with GCC 2.95 (or 2.96 for that matter) should really be told to get with the times. GCC is at 4.3 now, 2.95 is just a long gone memory from the distant past.
Sorry for taking so long to reply, for some reason my Gmane.org feed wasn't showing the latest replies on this thread.
OK, Windows XP still runs nearly all programs from previous versions Windows, plus it has DOS emulation so that it can run many, but not all DOS programs. What is SO WRONG with some element of reverse compatibility? I know that some of you may feel that there is a hinder to progress there, but there has to be some kind of balance between bleeding-edge and interoperability with other software.
I'm not talking about a need to build new binaries, I'm just talking about getting older software to run on a newer OS. This doesn't just include commercial software such as Maple, but also older open-source projects that haven't been updated in a while, but could still hypothetically work if the proper libraries were provided.
William
On Fri, 14 Nov 2008 20:19:18 -0500 "William M. Quarles" walrus@bellsouth.net wrote:
And just because a "free" software product hasn't been updated in a while does not mean that the software is useless.
The kernel is actually very backward compatible friendly. Very seldom has the kernel itself busted anything (removing support for old style ptys is one instance I can think of).
The thing that almost always causes problems is shared libraries being updated to a state of incompatibility, then no "compat" packages being maintained to ship the old libraries.
I've often thought it would be better if all software was always shipped with all the shared libs it needs installed in a unique per-program directory, and a system daemon run after each install to determine which libs were actually identical and hard-link them together :-).
If virtualization technology keeps on being pushed, perhaps we'll someday reach the point where every program runs inside its very own virtual machine with the exact environment required by that program always maintained in the perfect state for that one program's requirements.
At one time it was possible to avoid all these problems by static linking your program, but with every library now being written to dlopen some ridiculous plugin, attempts to static link things usually result in nonsense like dlopen dragging in libc.so to a program which was already linked with libc.a, and pretty soon you have two conflicting copies of malloc stomping on each other (I didn't just make this up - I've watched it happen :-).
Example: is there anybody out there doing natural science or engineering on Linux machines right now who is NOT using any commercial software whatsoever?
There are yes but that's really irrelevant to the discussion anyway
And just because a "free" software product hasn't been updated in a while does not mean that the software is useless.
If you want to run prehistoric application software you'll need to install the right ancient supporting libraries. Just like in the windows world, except the windows folk generally seem to have bundled ancient DLLs with the applications while Linux proprietary app providers have assumed the installing user can go find them.
Any thoughts?
I can still happily run 1993 applications on my current system with the right libraries added.
There is very little that doesn't work and that tend to divide into two categories
1. System configuration tools for very old interfaces 2. Programs with fundamental bugs that happen to show up nowdays due to thinks like faster processor speeds, incorrect assumptions in the code etc.
Alan
William M. Quarles wrote:
OK, so I said the phrase "Windows XP" in a message, and suddenly NO ONE will reply? Seriously, I'll say it again, Microsoft and Sony (e.g. most of the PlayStation series) can do reverse compatibility fairly well, and Fedora is almost totally lacking in that arena. "Get with the times" doesn't always make sense. And if anybody says, "Well, if you like reverse compatibility so much, why don't you shell out for Microsoft software like 90% of the rest of the herd." Uh, no. There are both commercial and "free" software products that need that reverse compatibility.
Example: is there anybody out there doing natural science or engineering on Linux machines right now who is NOT using any commercial software whatsoever?
And just because a "free" software product hasn't been updated in a while does not mean that the software is useless.
I can't quadruple boot my machine just to run all of the software I use. It should not be necessary, especially when most of this software is designed for Red Hat/Fedora distributions, or can be installed and run using WINE.
Any thoughts?
William
Vista won't run many old programs. Many old Windows programs have problems with newer ones. We have scientific equipment that requires Windows 98 to run as the software won't run on XP.
If you are using older software, then expect to have issues. This is how Microsoft gets you to spend more money on their products.
I have not purchased any commercial software on my Linux box. The only one that I would like would be Autocad but for 95% of my drafting needs, Qcad works.
As for older OpenSource products, sometimes the only thing needed to do is to update the "make" file.
As others say, run the software in a VM if it won't run in Wine.
When it comes to software, it comes up to the supplier of the software.
-
Alan Cox -
Bill Davidsen -
Gordon Messmer -
Kam Leo -
Kevin Kofler -
Robin Laing -
Tom Horsley -
William M. Quarles