June 2012 - petsc-dev - mailman3.cels.anl.gov

Re: [petsc-dev] Patches like this make me want to cry
by Sean Farley 28 Jun '12

28 Jun '12

On Thu, Jun 28, 2012 at 5:41 PM, Paul Mullowney <paulm(a)txcorp.com> wrote: > I moved all CUSPARSE functions into seqcusparse/aijcusparse.cu (and a > similar mpi version). I was asked to do this some months ago. I did my best. > I think it is significantly improved from before. > > I was also asked by multiple people to attempt to get Complex GPU > capabilities working. I succeeded by making changes to petscmath.h. As Jed > pointed out, some of the changes may have not been C99 compliant. > > There is one commented out function in aijcusparse.cu. That can easily be > fixed. > > Many of the other changes (especially to .cu) files were made to protect > against builds breaking because many of the CUSP preconditioners do not > support complex arithmetic. You seem to be addressing the specifics of the code; Matt and Jed are talking about the process: patch reviews. > I have asked for input from multiple people on several occasions and not > gotten responses. > > What am I to do if I ask for input/feedback and don't get it????? I saw no request for reviewing this patch series. How did you ask people to review? I recommend (for everyone here) smaller commits. Commit often. There is a great extension for editing changesets (i.e. folding multiple changesets together, splitting changesets apart, etc.) called histedit: http://bitbucket.org/durin42/histedit In fact, just today it was included into mercurial core for the next version. I also recommend the 'record' extension (bundled with mercurial). The next version of mercurial (2.3) promises better bookmark functionality (for feature branches), built-in history editing, and possibly evolution (for shared, mutable changesets). If you have a bitbucket account (free), I would recommend working on your own fork there so that others could review and you could mutate the history based on said reviews.

1 0

Re: [petsc-dev] Patches like this make me want to cry
by Jed Brown 28 Jun '12

28 Jun '12

On Thu, Jun 28, 2012 at 2:41 PM, Paul Mullowney <paulm(a)txcorp.com> wrote: > I moved all CUSPARSE functions into seqcusparse/aijcusparse.cu (and a > similar mpi version). I was asked to do this some months ago. I did my > best. I think it is significantly improved from before. > > I was also asked by multiple people to attempt to get Complex GPU > capabilities working. I succeeded by making changes to petscmath.h. As Jed > pointed out, some of the changes may have not been C99 compliant. > It needs to work with C89, not just C99. Global changes to handle complex should also be done in separate patches. I hate the idea that CUSP leaks out into user code so I would recommend casting internally rather than changing it globally. That should also enable use of complex without requiring the user to adopt a C++ compiler for their own code (C99 _Complex is binary-compatible with std::complex and cusp::complex). > > There is one commented out function in aijcusparse.cu. That can easily be > fixed. > > Many of the other changes (especially to .cu) files were made to protect > against builds breaking because many of the CUSP preconditioners do not > support complex arithmetic. > Those should absolutely be in separate patches. Lumping it all into one makes it impossible to review or test. > > I have asked for input from multiple people on several occasions and not > gotten responses. > > What am I to do if I ask for input/feedback and don't get it????? > Next time, please provide the patch series. You could use "hg email" or post your branch to bitbucket or elsewhere so that we can comment. Also, please follow the coding guidelines. There is a developer's manual and frequent discussion on petsc-dev about conventions. (If you notice something that is adopted, but not in the developer's manual, please point it out on the mailing list or make a patch.) > > > -Paul > > > > > There are essentially no comments on this massive push. I was just > reading it, getting mad. Jed mailed first. > Pushes need enough comments that I can see exactly what functionality is > being added and how it works. > You will not be the maintainer of this functionality, we will. If we can't > do it, it will be thrown out, and all your > time will be wasted. > > Matt > > >> >>> This is non-portable: >>>> >>>> http://petsc.cs.iit.edu/petsc/petsc-dev/rev/66ca8db0d5f8 >>>> >>>> >>>> Can we please institute some sort of policy on patch >>>> quality/reviewability? This one patch is going to take a significant amount >>>> of fix-up (not made easier by the several merges since) and/or generate >>>> several build failures and user inconvenience (petsc-maints). No doubt the >>>> functionality is important, but we just don't have time to fix these things >>>> line-by-line after they are pushed. >>>> >>> >>> >> > > > -- > What most experimenters take for granted before they begin their > experiments is infinitely more interesting than any results to which their > experiments lead. > -- Norbert Wiener > > >

1 0

Re: [petsc-dev] slow "ar Scq" on cygwin
by Chetan Jhurani 28 Jun '12

28 Jun '12

> > $ PETSC_DIR=/cygdrive/d/petsc-dev-2012-06-26 PETSC_ARCH=test_8 ./config/builder.py > > (The same happens (existing .o are deleted) if ./config/builder2.py build is used) > This is not what I wrote. You used builder.py, which is python 2.4 compatible, but has to do everything without options, > so yes it deletes all the objects and rebuilds everything. That is why I did not suggest you use it. > I said to use builder2.py, which does proper dependency tracking. That is what I used and I mentioned it above. I've tried both separately. Both give the same result that libpetsc.lib is not created. And whether I use builder.py or builder2.py, the existing objects files are deleted in the second build. I did not call for a cleanup. To be sure, I tried the whole thing in a new ARCH with a fresh configure step, and this time builder.py was not used at all (by me) but it led to the same issues with builder2.py. I understand that this is not the most used platform, so let me know if I can be of help. And I forgot to mention that the python build was 4-5 times faster than "ar" based build and that's great... Chetan From: petsc-dev-bounces(a)mcs.anl.gov [mailto:[email protected]] On Behalf Of Matthew Knepley Sent: Thursday, June 28, 2012 3:21 PM To: For users of the development version of PETSc Subject: Re: [petsc-dev] slow "ar Scq" on cygwin On Thu, Jun 28, 2012 at 4:10 PM, Chetan Jhurani <chetan.jhurani(a)gmail.com> wrote: Matt, I was surprised too on both counts. Couldn't find any libpetsc.lib after the python build and I did not run "clean" in between that could have deleted the .o files. Here's one more try below to show it. $ ls test_8/lib/ libpetsc-obj $ ls test_8/lib/libpetsc-obj/ | wc -l 560 $ PETSC_DIR=/cygdrive/d/petsc-dev-2012-06-26 PETSC_ARCH=test_8 ./config/builder.py (The same happens (existing .o are deleted) if ./config/builder2.py build is used) This is not what I wrote. You used builder.py, which is python 2.4 compatible, but has to do everything without options, so yes it deletes all the objects and rebuilds everything. That is why I did not suggest you use it. I said to use builder2.py, which does proper dependency tracking. Matt Rebuilding Dependencies Building in /cygdrive/d/petsc-dev-2012-06-26/src/sys Building in /cygdrive/d/petsc-dev-2012-06-26/src/sys/bag Building in /cygdrive/d/petsc-dev-2012-06-26/src/sys/comm Building in /cygdrive/d/petsc-dev-2012-06-26/src/sys/dll I stopped the build using control-c. It would have deleted all files first. Traceback (most recent call last): File "./config/builder.py", line 1534, in <module> PETScMaker().run() File "./config/builder.py", line 1526, in run if self.buildLibraries('libpetsc', self.rootDir): File "./config/builder.py", line 1331, in buildLibraries objects += self.buildDir(root, files, objDir) File "./config/builder.py", line 1251, in buildDir objects.extend(getattr(self, 'compile'+language)(sourceMap[language], objDir)) File "./config/builder.py", line 1025, in compileC return self.compile(self.configInfo.languages.clanguage, source, objDir) File "./config/builder.py", line 1012, in compile (output, error, status) = self.executeShellCommand(cmd, checkCommand = noCheckCommand, log=self.log) File "/cygdrive/d/petsc-dev-2012-06-26/config/BuildSystem/script.py", line 254, in executeShellCommand (output, error, status) = runInShell(command, log, cwd) File "/cygdrive/d/petsc-dev-2012-06-26/config/BuildSystem/script.py", line 243, in runInShell thread.join(timeout) File "/usr/lib/python2.6/threading.py", line 655, in join self.__block.wait(delay) File "/usr/lib/python2.6/threading.py", line 258, in wait _sleep(delay) KeyboardInterrupt $ ls test_8/lib/libpetsc-obj/ bag.o All the previous .o were deleted. Perhaps I'm not doing something right? Chetan From: petsc-dev-bounces(a)mcs.anl.gov [mailto:[email protected]] On Behalf Of Matthew Knepley Sent: Thursday, June 28, 2012 2:46 PM To: For users of the development version of PETSc Subject: Re: [petsc-dev] slow "ar Scq" on cygwin On Thu, Jun 28, 2012 at 3:31 PM, Chetan Jhurani <chetan.jhurani(a)gmail.com> wrote: Thanks Matt, Satish. A few issues though. Like you said, the latest cygwin has python 2.6.8 which does not have argparse. When I place the following in path http://argparse.googlecode.com/svn/trunk/argparse.py I can run the python build with 2.6.8. However, the "ar" based build had worked fine and the python based build fails for these two files. src/snes/impls/MULTIB~1/multiblock.c src/snes/impls/vi/rsaug/virsaug.c Yep, it fails, but the build succeeds, so no problem. It looks like these two files were NOT compiled in the ar based build. The two make.log files are attached. One more question. I tried to run the python based build the second time (after it was unsuccessful the first time) and then it deleted all the previous .o files it had created in ARCH/lib/libpetsc-obj. How can that be avoided? It does not do that unless you run 'clean'. Matt Chetan From: petsc-dev-bounces(a)mcs.anl.gov [mailto:[email protected]] On Behalf Of Matthew Knepley Sent: Thursday, June 28, 2012 1:35 PM To: For users of the development version of PETSc Subject: Re: [petsc-dev] slow "ar Scq" on cygwin On Thu, Jun 28, 2012 at 2:23 PM, Chetan Jhurani <chetan.jhurani(a)gmail.com> wrote: Hi devs, Is there a way to avoid running the archiver (ar) every time a petsc directory is compiled? The reason I'm asking this is that when libpetsc.lib becomes large enough (say 100MB+, esp. for debug build), "ar Sqc" on cygwin takes very long in each directory. Depending on OS caching, very long can be 10 seconds per PETSc directory just to run ar. It creates a new temp file from scratch, puts stuff from libpetsc.lib and new stuff in it and then renames the temp file. As you can imagine, the compilation process is fast in the beginning but crawls by the end. I've seen the same IO sequence on Linux (with Sqc flags to ar), but since cygwin IO is slower the effect is more clearly visible. Building within a ramdisk leads to 25% faster compilation but it was more of a curiosity. Build with Python. It never uses antique things like ar :) python2.7 ./config/builder2.py build Matt Chetan -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener

1 0

Re: [petsc-dev] Patches like this make me want to cry
by Paul Mullowney 28 Jun '12

28 Jun '12

I moved all CUSPARSE functions into seqcusparse/aijcusparse.cu (and a similar mpi version). I was asked to do this some months ago. I did my best. I think it is significantly improved from before. I was also asked by multiple people to attempt to get Complex GPU capabilities working. I succeeded by making changes to petscmath.h. As Jed pointed out, some of the changes may have not been C99 compliant. There is one commented out function in aijcusparse.cu. That can easily be fixed. Many of the other changes (especially to .cu) files were made to protect against builds breaking because many of the CUSP preconditioners do not support complex arithmetic. I have asked for input from multiple people on several occasions and not gotten responses. What am I to do if I ask for input/feedback and don't get it????? -Paul > > There are essentially no comments on this massive push. I was just > reading it, getting mad. Jed mailed first. > Pushes need enough comments that I can see exactly what functionality > is being added and how it works. > You will not be the maintainer of this functionality, we will. If we > can't do it, it will be thrown out, and all your > time will be wasted. > > Matt > > > This is non-portable: > > http://petsc.cs.iit.edu/petsc/petsc-dev/rev/66ca8db0d5f8 > > > Can we please institute some sort of policy on patch > quality/reviewability? This one patch is going to take a > significant amount of fix-up (not made easier by the > several merges since) and/or generate several build > failures and user inconvenience (petsc-maints). No doubt > the functionality is important, but we just don't have > time to fix these things line-by-line after they are pushed. > > > > > > > -- > What most experimenters take for granted before they begin their > experiments is infinitely more interesting than any results to which > their experiments lead. > -- Norbert Wiener

1 0

Re: [petsc-dev] Patches like this make me want to cry
by Matthew Knepley 28 Jun '12

28 Jun '12

On Thu, Jun 28, 2012 at 4:33 PM, Jed Brown <jedbrown(a)mcs.anl.gov> wrote: > On Thu, Jun 28, 2012 at 2:24 PM, Paul Mullowney <paulm(a)txcorp.com> wrote: > >> VecTransplantPlaceArray >> >> http://petsc.cs.iit.edu/petsc/**petsc-dev/rev/d2f118b395b2<http://petsc.cs.iit.edu/petsc/petsc-dev/rev/d2f118b395b2> >>> >>> This thing is way too big to review, includes huge swaths of >>> commented-out code, breaks coding conventions and portability, and >>> introduces strange new APIs (like VecTransplantPlaceArray) that haven't >>> really been explained and seem to produce questionable semantics. >>> >>> >>> Why is the non-portable? > > > It assumes C99 to begin with. > Barry has yelled at many people, including me, about // comments and declarations in the middle of a block. Being able to push means LISTENING when these things happen on the mailing list. > I asked for your input on this 2 days ago? > > > Sorry, I'm at a conference, finishing a proposal, and trying to make > progress on my research. I was hoping someone else would comment because I > think it's leaking implementation details. > > Also, you didn't provide the whole patch series to comment on, just some > bits of code. > There are essentially no comments on this massive push. I was just reading it, getting mad. Jed mailed first. Pushes need enough comments that I can see exactly what functionality is being added and how it works. You will not be the maintainer of this functionality, we will. If we can't do it, it will be thrown out, and all your time will be wasted. Matt > >> This is non-portable: >>> >>> http://petsc.cs.iit.edu/petsc/**petsc-dev/rev/66ca8db0d5f8<http://petsc.cs.iit.edu/petsc/petsc-dev/rev/66ca8db0d5f8> >>> >>> >>> Can we please institute some sort of policy on patch >>> quality/reviewability? This one patch is going to take a significant amount >>> of fix-up (not made easier by the several merges since) and/or generate >>> several build failures and user inconvenience (petsc-maints). No doubt the >>> functionality is important, but we just don't have time to fix these things >>> line-by-line after they are pushed. >>> >> >> > -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener

1 0

Re: [petsc-dev] Patches like this make me want to cry
by Jed Brown 28 Jun '12

28 Jun '12

On Thu, Jun 28, 2012 at 2:24 PM, Paul Mullowney <paulm(a)txcorp.com> wrote: > VecTransplantPlaceArray > > http://petsc.cs.iit.edu/petsc/**petsc-dev/rev/d2f118b395b2<http://petsc.cs.iit.edu/petsc/petsc-dev/rev/d2f118b395b2> >> >> This thing is way too big to review, includes huge swaths of >> commented-out code, breaks coding conventions and portability, and >> introduces strange new APIs (like VecTransplantPlaceArray) that haven't >> really been explained and seem to produce questionable semantics. >> >> >> Why is the non-portable? It assumes C99 to begin with. > I asked for your input on this 2 days ago? Sorry, I'm at a conference, finishing a proposal, and trying to make progress on my research. I was hoping someone else would comment because I think it's leaking implementation details. Also, you didn't provide the whole patch series to comment on, just some bits of code. > > This is non-portable: >> >> http://petsc.cs.iit.edu/petsc/**petsc-dev/rev/66ca8db0d5f8<http://petsc.cs.iit.edu/petsc/petsc-dev/rev/66ca8db0d5f8> >> >> >> Can we please institute some sort of policy on patch >> quality/reviewability? This one patch is going to take a significant amount >> of fix-up (not made easier by the several merges since) and/or generate >> several build failures and user inconvenience (petsc-maints). No doubt the >> functionality is important, but we just don't have time to fix these things >> line-by-line after they are pushed. >> > >

1 0

Re: [petsc-dev] Patches like this make me want to cry
by Sean Farley 28 Jun '12

28 Jun '12

> Can we please institute some sort of policy on patch quality/reviewability? > This one patch is going to take a significant amount of fix-up (not made > easier by the several merges since) and/or generate several build failures > and user inconvenience (petsc-maints). No doubt the functionality is > important, but we just don't have time to fix these things line-by-line > after they are pushed. Revoke all write access (except for a few trusted people) and move to a pull-request-type system?

1 0

Re: [petsc-dev] Xcode
by Barry Smith 28 Jun '12

28 Jun '12

Yes, but with limited experience. See the section in the users manual: http://www.mcs.anl.gov/petsc/petsc-current/docs/manual.pdf section 13.13 page 177 Feed back on improving the documentation on this etc would be appreciated. Barry On Jun 28, 2012, at 4:35 PM, Cai Durbin wrote: > Hey, > > I was wondering if anyone has had any success using PETSc with Xcode? > > Cheers, > > Cai

1 0

Patches like this make me want to cry
by Jed Brown 28 Jun '12

28 Jun '12

http://petsc.cs.iit.edu/petsc/petsc-dev/rev/d2f118b395b2 This thing is way too big to review, includes huge swaths of commented-out code, breaks coding conventions and portability, and introduces strange new APIs (like VecTransplantPlaceArray) that haven't really been explained and seem to produce questionable semantics. This is non-portable: http://petsc.cs.iit.edu/petsc/petsc-dev/rev/66ca8db0d5f8 Can we please institute some sort of policy on patch quality/reviewability? This one patch is going to take a significant amount of fix-up (not made easier by the several merges since) and/or generate several build failures and user inconvenience (petsc-maints). No doubt the functionality is important, but we just don't have time to fix these things line-by-line after they are pushed.

1 0

Re: [petsc-dev] slow "ar Scq" on cygwin
by Matthew Knepley 28 Jun '12

28 Jun '12

On Thu, Jun 28, 2012 at 4:10 PM, Chetan Jhurani <chetan.jhurani(a)gmail.com>wrote: > Matt,**** > > ** ** > > I was surprised too on both counts. Couldn’t find any libpetsc.lib after > the python build**** > > and I did not run “clean” in between that could have deleted the .o files. > **** > > ** ** > > Here’s one more try below to show it.**** > > ** ** > > *$ ls test_8/lib/* > > libpetsc-obj**** > > ** ** > > *$ ls test_8/lib/libpetsc-obj/ | wc -l* > > *560* > > ** ** > > *$ PETSC_DIR=/cygdrive/d/petsc-dev-2012-06-26 PETSC_ARCH=test_8 > ./config/builder.py* > > *(The same happens (existing .o are deleted) if ./config/builder2.py > build is used)* > This is not what I wrote. You used builder.py, which is python 2.4 compatible, but has to do everything without options, so yes it deletes all the objects and rebuilds everything. That is why I did not suggest you use it. I said to use builder2.py, which does proper dependency tracking. Matt > * * > > Rebuilding Dependencies**** > > Building in /cygdrive/d/petsc-dev-2012-06-26/src/sys**** > > Building in /cygdrive/d/petsc-dev-2012-06-26/src/sys/bag**** > > Building in /cygdrive/d/petsc-dev-2012-06-26/src/sys/comm**** > > Building in /cygdrive/d/petsc-dev-2012-06-26/src/sys/dll**** > > ** ** > > *I stopped the build using control-c. It would have deleted all files > first.* > > ** ** > > Traceback (most recent call last):**** > > File "./config/builder.py", line 1534, in <module>**** > > PETScMaker().run()**** > > File "./config/builder.py", line 1526, in run**** > > if self.buildLibraries('libpetsc', self.rootDir):**** > > File "./config/builder.py", line 1331, in buildLibraries**** > > objects += self.buildDir(root, files, objDir)**** > > File "./config/builder.py", line 1251, in buildDir**** > > objects.extend(getattr(self, 'compile'+language)(sourceMap[language], > objDir))**** > > File "./config/builder.py", line 1025, in compileC**** > > return self.compile(self.configInfo.languages.clanguage, source, > objDir)**** > > File "./config/builder.py", line 1012, in compile**** > > (output, error, status) = self.executeShellCommand(cmd, checkCommand = > noCheckCommand, log=self.log)**** > > File "/cygdrive/d/petsc-dev-2012-06-26/config/BuildSystem/script.py", > line 254, in executeShellCommand**** > > (output, error, status) = runInShell(command, log, cwd)**** > > File "/cygdrive/d/petsc-dev-2012-06-26/config/BuildSystem/script.py", > line 243, in runInShell**** > > thread.join(timeout)**** > > File "/usr/lib/python2.6/threading.py", line 655, in join**** > > self.__block.wait(delay)**** > > File "/usr/lib/python2.6/threading.py", line 258, in wait**** > > _sleep(delay)**** > > KeyboardInterrupt**** > > ** ** > > *$ ls test_8/lib/libpetsc-obj/* > > *bag.o* > > ** ** > > All the previous .o were deleted. Perhaps I’m not doing something right?* > *** > > ** ** > > Chetan**** > > ** ** > > ** ** > > *From:* petsc-dev-bounces(a)mcs.anl.gov [mailto: > petsc-dev-bounces(a)mcs.anl.gov] *On Behalf Of *Matthew Knepley > *Sent:* Thursday, June 28, 2012 2:46 PM > *To:* For users of the development version of PETSc > *Subject:* Re: [petsc-dev] slow "ar Scq" on cygwin**** > > ** ** > > On Thu, Jun 28, 2012 at 3:31 PM, Chetan Jhurani <chetan.jhurani(a)gmail.com> > wrote:**** > > Thanks Matt, Satish. A few issues though.**** > > **** > > Like you said, the latest cygwin has python 2.6.8 which**** > > does not have argparse. When I place the following**** > > in path http://argparse.googlecode.com/svn/trunk/argparse.py**** > > I can run the python build with 2.6.8.**** > > **** > > However, the “ar” based build had worked fine and the python**** > > based build fails for these two files.**** > > **** > > src/snes/impls/MULTIB~1/multiblock.c**** > > src/snes/impls/vi/rsaug/virsaug.c**** > > ** ** > > Yep, it fails, but the build succeeds, so no problem.**** > > **** > > It looks like these two files were NOT compiled in the ar based**** > > build. The two make.log files are attached.**** > > **** > > One more question. I tried to run the python based build**** > > the second time (after it was unsuccessful the first time)**** > > and then it deleted all the previous .o files it had created in **** > > ARCH/lib/libpetsc-obj. How can that be avoided?**** > > ** ** > > It does not do that unless you run 'clean'.**** > > ** ** > > Matt**** > > **** > > Chetan**** > > **** > > **** > > *From:* petsc-dev-bounces(a)mcs.anl.gov [mailto: > petsc-dev-bounces(a)mcs.anl.gov] *On Behalf Of *Matthew Knepley > *Sent:* Thursday, June 28, 2012 1:35 PM > *To:* For users of the development version of PETSc > *Subject:* Re: [petsc-dev] slow "ar Scq" on cygwin**** > > **** > > On Thu, Jun 28, 2012 at 2:23 PM, Chetan Jhurani <chetan.jhurani(a)gmail.com> > wrote:**** > > Hi devs, > > Is there a way to avoid running the archiver (ar) > every time a petsc directory is compiled? > > The reason I'm asking this is that when libpetsc.lib > becomes large enough (say 100MB+, esp. for debug build), > "ar Sqc" on cygwin takes very long in each directory. > Depending on OS caching, very long can be 10 seconds per > PETSc directory just to run ar. > > It creates a new temp file from scratch, puts stuff from > libpetsc.lib and new stuff in it and then renames the > temp file. As you can imagine, the compilation process is > fast in the beginning but crawls by the end. I've seen > the same IO sequence on Linux (with Sqc flags to ar), but > since cygwin IO is slower the effect is more clearly visible. > Building within a ramdisk leads to 25% faster compilation > but it was more of a curiosity.**** > > **** > > Build with Python. It never uses antique things like ar :)**** > > **** > > python2.7 ./config/builder2.py build**** > > **** > > Matt**** > > **** > > > Chetan**** > > > > **** > > **** > > -- > What most experimenters take for granted before they begin their > experiments is infinitely more interesting than any results to which their > experiments lead. > -- Norbert Wiener**** > > > > **** > > ** ** > > -- > What most experimenters take for granted before they begin their > experiments is infinitely more interesting than any results to which their > experiments lead. > -- Norbert Wiener**** > -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener

1 0