[kepler-dev] [Ptolemy] Re: Conversion of Ptolemy II Tree to SVN this week

Christopher Brooks cxh at eecs.berkeley.edu
Thu Jun 12 10:02:20 PDT 2008


Hi Derik,
Thanks.  I'll either reconvert the repository or fix this by hand.

<rant>
This is further evidence that svn and most things associated with it
are of poor quality.

Why would the default value in cvs2svn be --default-eol be binary?  To
protect the user because they messed up the binary status of their CVS
repository?  This is another very poor decision.  I get punished
because my CVS repository was right?  Yep, I should have carefully
read the fine manual, but I think it is more reasonable to set the
default of the tool correctly.  I read skimmed the manual and
carefully cleaned my tree, but then the default setting of cvs2svn
ignores this work?

?

Also, why is it so hard to find out the status of the file?

Which is the better tool?

cxh at carson 45% cd ~/src/ptweb/img
cxh at carson 42% cvs status PtolemyIIIcon.gif
===================================================================
File: PtolemyIIIcon.gif	Status: Up-to-date

   Working revision:	1.1
   Repository revision:	1.1	/home/cvs/ptweb/image/PtolemyIIIcon.gif,v
   Sticky Tag:		(none)
   Sticky Date:		(none)
   Sticky Options:	-kb

cxh at carson 45% cd ~/ptII/doc/img
cxh at carson 46% svn status -v PtolemyII.jpg
            49777     6158 cxh          PtolemyII.jpg

Why doesn't svn status tell me that keyword substitution is not
occuring?

?

Another issue is that using the svn2cl script to create a ChangeLog
file takes 2-3 hours instead of the 10 minutes it took with CVS.

My guess is that the problem is svn2cl.

I'm starting to see the point of git.  Too bad the clients were
unusuable.

</rant>

Thanks, I feel better now. :-)

Many thanks also to Bert for looking into the repository size issue.

_Christopher




--------

    Hey Christopher,
    
    I'm not sure how to verify that cvs2svn treated all your files as  
    binary, but I think it did because the default to the --default-eol  
    switch is "binary" (and you didn't use the switch and so I assume it  
    used the default).
    
    I just found this section of the svn book:
    http://subversion.tigris.org/faq.html#auto-props
    
    I think (correct me if I'm wrong svn experts) that section illustrates  
    one limitation of your svn repo as it is now: keywords like $Id$ will  
    not be automatically changed in your ascii files because the  
    svn:keywords property is not set on these files.
    
    You don't *have* to use --default-eol=native switch when you convert,  
    but it has the benefit of setting the svn:keywords property for ascii  
    files. Also beware: the manual warns to *not* use this switch if you  
    haven't ensured all binary files in your cvs are indeed marked as such  
    (-kb), your svn repo may be corrupt otherwise.
    
    Using other switches, cvs2svn can also try to deduce which files are  
    binary for you during conversion. I didn't try that, I manually looked  
    at every kepler file (in eclipse) and changed the cvs ascii/binary  
    property on those that were set incorrectly so that I could use -- 
    default-eol=native ie "Full conversion" method described on the  
    cvs2svn page.
    
    More info here:
    http://cvs2svn.tigris.org/cvs2svn.html#convert
    http://cvs2svn.tigris.org/cvs2svn.html#prep
    
    Derik
    
    On Jun 11, 2008, at 6:38 PM, Christopher Brooks wrote:
    
    > Hi Derik,
    >
    > I'm not familiar with svn property sets.
    >
    > At about 5:30 PM PDT, I did end up making the ptII svn repository
    > writable by the cvs Unix group, so it might not have been writable
    > before by anyone but myself before that time:
    >
    > I was just able to set a foo property on $PTII/makefile
    >
    > bash-3.00$ svn propset foo bar makefile
    > property 'foo' set on 'makefile'
    > bash-3.00$ svn proplist makefile
    > Properties on 'makefile':
    >  foo
    > bash-3.00$
    >
    > I did not use --default-eol=native when I did the conversion using
    > cvs2svn.  If this is required, then the docs of cvs2svn need to be
    > updated and I may need to rerun the conversion.
    >
    > How would I tell if all the files are binary?
    >
    > I think we are ok, $PTII/makefile does not have a mime-type,
    > but a binary jpg does.
    >
    > bash-3.00$ cd $PTII
    > bash-3.00$ svn propget svn:mime-type makefile
    > bash-3.00$ svn propget svn:mime-type doc/img/PtolemyII.jpg
    > application/octet-stream
    > bash-3.00$
    >
    > _Christopher
    > --------
    >
    >    Hi Christopher and Kepler developers,
    >
    >    We're delaying a bit with our transition to SVN to do a little more
    >    testing.
    >    If we don't hit any problems tonight we'll lock kepler and kepler- 
    > docs
    >    CVS and do the move tomorrow morning.
    >
    >    So if everyone could plan on *not* using kepler and kepler-docs  
    > CVS in
    >    the morning, that would be appreciated.
    >
    >
    >    One note Christopher: I notice files in the new ptII SVN don't have
    >    properties viewable with e.g the svn proplist command. My guess is
    >    this is because you didn't use --default-eol=native in your
    >    conversion, thus treating all files as binary(?). Which is a safer
    >    route for conversion to avoid repository corruption, but now the
    >    properties are not available. I haven't thought through what the
    >    ramifications of this are, but I wanted to bring it up to get your
    >    thoughts.
    >
    >    Derik
    >
    >    On Jun 9, 2008, at 9:58 AM, Matthew Jones wrote:
    >
    >> Hi Christopher,
    >>
    >> We're ready to go with the kepler conversion to SVN as well, and
    >> Wednesday sounds like a good day to do it.  Here's our plan:
    >>
    >> 1) Monday: run test conversion of kepler-docs and kepler modules
    >> 2) Tuesday: inspect test repositories, and test builds
    >> 3) Wednesday: Lock CVS, do the conversion, test build, update nightly
    >> build, update build documentation
    >>
    >> Right now our current plan is for Derik to do the move with support
    >> from
    >> me, Chad, and Ben.  Coordinating the move with ptolemy will minimize
    >> downtime.
    >>
    >> Matt
    >>
    >> Christopher Brooks wrote:
    >>> I'm thinking about converting the Ptolemy II cvs repository to
    >>> subversion this week.
    >>>
    >>> To do this, what I would do is:
    >>> 1) Update the Eclipse installation instructions
    >>> 2) Disable the ptII cvs repository
    >>> 3) Convert the ptII cvs repository to svn, which would take a few
    >>> hours
    >>>
    >>> I'm thinking of doing this on Wednesday.
    >>>
    >>> Any comments, questions or concerns?
    >>>
    >>> _Christopher
    >>> _______________________________________________
    >>> Kepler-dev mailing list
    >>> Kepler-dev at ecoinformatics.org
    >>> http://mercury.nceas.ucsb.edu/ecoinformatics/mailman/listinfo/kepler-de
   v
    >>
    >> -- 
    >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    >> Matthew B. Jones
    >> Director of Informatics Research and Development
    >> National Center for Ecological Analysis and Synthesis (NCEAS)
    >> UC Santa Barbara
    >> jones at nceas.ucsb.edu                       Ph: 1-907-523-1960
    >> http://www.nceas.ucsb.edu/ecoinfo
    >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    >> _______________________________________________
    >> Kepler-dev mailing list
    >> Kepler-dev at ecoinformatics.org
    >> http://mercury.nceas.ucsb.edu/ecoinformatics/mailman/listinfo/kepler-dev
    > --------
    
    _______________________________________________
    Ptolemy maillist  -  Ptolemy at chess.eecs.berkeley.edu
    http://chess.eecs.berkeley.edu/ptolemy/listinfo/ptolemy
--------


More information about the Kepler-dev mailing list