Page 1 of 1
DB2 Installation
Posted: Thu Oct 12, 2006 4:32 pm
by 73blazer
OK, I got "subversion" finally compiled and working. That was a chore. The checkout thing works, so I'm pretty sure it's working ok.
But, how do I do a diff? ANytime I try to use it against my install it keeps complaining
about: Can't open file 'serendipity/.svn/entries
Can i not do a diff against a "production" install? Am I going to have to make all my changes manually to a "checked out" version?
I'm new to all this, I've never used cvs or or subversion, so I don't really know what I'm doing with it.
I've tried:
svn diff serendipity
svn diff --new svn://svn.berlios.de/serendipity/trunk --old ./serendipity
both produce:
svn: 'serendipity' is not a working copy
svn: Can't open file 'serendipity/.svn/entries': A file or directory in the path name does not exist.
Re: DB2 Installation
Posted: Thu Oct 12, 2006 5:14 pm
by garvinhicking
Hi!
You must do a "svn checkout", not a "svn export", did you do that? Do you have any ".svn" directories at all? There should be one for each usual directory you have.
Usually you should do this:
1. Checkout an SVN trunk
2. Copy your modified files into that directory - do NOT copy your previous .svn directories over; just the .php files and other files you manually modified
3. Make "svn diff" and that should work
So, I suppose what your problem is, is that you simply have not checked out a SVN repository to put your changes into and diff against.
Best regards,
Garvin
Posted: Thu Oct 12, 2006 5:29 pm
by 73blazer
Ahh..ok
I was trying to do a diff against my current installation, which was a tarball from the main (not svn) download s9y page + my modifications for DB2 over the last many weeks..
So I need to do a checkout, then put my files I've modified in there, then do a diff...
ok...I'll try that..
Thanks...
Posted: Thu Oct 12, 2006 6:39 pm
by 73blazer
Ok, I got it.
It is clear that many changes have been made since 1.0.1 (not including my changes).
Should I attempt a merge, or should I make my changes manually to the trunk and then do a diff?
What's the best way to produce a solid working latest tree?
Thanks for the coaching...
Posted: Thu Oct 12, 2006 6:50 pm
by garvinhicking
Hi!
If you use a testinstall, I'd suggest this:
1. Create the SVN checkout of trunk
2. Copy your changes into the checkout
3. Run "svn update" and hope that SVN will merge in the 1.1 changes that happened into your files
I'm quite sure there must be a way to merge in the 1.1 changes into your 1.0 version, but this is hitting my experience with SVN. Maybe judebert or someone else has more solid SVN knowledge to get a way for this? Maybe we'd also need to consult some SVN forums for that.
Best regards,
Garvin
Posted: Thu Oct 12, 2006 11:43 pm
by 73blazer
Well, it looks like the update worked.
Ok, so I'm back to editing files on the latest trunk now, I've got some time in the next couple days to hammer this down, so I should have a somewhat fully working tree pretty soon. I've probably broke some things for the other databases though. Starting with the original patch from Dan, I've tried to take as much if/then about databases out of the code and into the various db include files, anything db specific is contained within those. That required reformatting alot of statements so they could be passed thru functions for any database, most notably the syntax of the LIMIT/OFFSET clauses because the DB2 is drastically different for those. I thought about trying to change the whole method for needing limit/offset in the first place,but that could be for another day.
Anyhow, I don't know how adventurous you guys are to these changes, but you can see what I've got and see if you want to move forward in this direction. The good part about it is it should be alot easier to add other database backends.
Posted: Fri Oct 13, 2006 5:32 am
by judebert
SW33T.
I have to say, this sounds like a lot of work. If it makes later database integration easier, though, I'm all for it.
When you're done, I suppose we'll want to browse the differences as a kind of peer review to avoid unexpected problems. Then we can let others test it out.
And then we can say we support DB2.
