You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Branko Čibej <br...@xbc.nu> on 2002/06/15 18:21:04 UTC
Re: svn commit: rev 2226 - trunk/subversion/libsvn_delta
brane@tigris.org wrote:
>Author: brane
>Date: 2002-06-15 17:50 GMT
>New Revision: 2226
>
>Modified:
> trunk/subversion/libsvn_delta/compose_delta.c
>Log:
>Ongoing work on the delta combiner (issue #531).
>
>* compose_delta.c: (buld_range_list): Fix bug in target range generation.
>
>
Whee, the delta combiner now actually _works_! Overlapping target copies
are the final hurdle (edge cases again).
--
Brane Čibej <br...@xbc.nu> http://www.xbc.nu/brane/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: svn commit: rev 2226 - trunk/subversion/libsvn_delta
Posted by Branko Čibej <br...@xbc.nu>.
Karl Fogel wrote:
>Branko Čibej <br...@xbc.nu> writes:
>
>
>>Whee, the delta combiner now actually _works_! Overlapping target
>>copies are the final hurdle (edge cases again).
>>
>>
>
>Wonderful!
>
>The big question is, how much does it improve performance? Our big
>problem case was undeltifying files larger than the delta window
>size.
>
I'll be able to tell you that as soon as I've tickled all the bugs out.
And reworked our deltification code a bit ...
--
Brane Čibej <br...@xbc.nu> http://www.xbc.nu/brane/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: svn commit: rev 2226 - trunk/subversion/libsvn_delta
Posted by Branko Čibej <br...@xbc.nu>.
Greg Hudson wrote:
>On Tue, 2002-06-18 at 12:15, Karl Fogel wrote:
>
>
>>Greg Hudson <gh...@MIT.EDU> writes:
>>
>>
>>>On Sun, 2002-06-16 at 19:28, Karl Fogel wrote:
>>>
>>>
>>>>The big question is, how much does it improve performance? Our big
>>>>problem case was undeltifying files larger than the delta window
>>>>size.
>>>>
>>>>
>>>Um, that problem case would have been due to a bug of some sort, not due
>>>to the lack of a delta combiner.
>>>
>>>
>>I'm not sure precisely what is being asserted or denied above :-).
>>
>>
>
>My understanding is that, at some time in the past, it was observed that
>if we deltified files above the window size, memory usage went through
>the roof. That's clearly a bug; windows are there to place a ceiling on
>memory usage, not make it bigger. I don't know if anyone has checked
>whether this bug still exists.
>
The memory problem was probably caused by overlapping source windows.
That, and and the fact that pools didn't know how to free memory the
last time we tested that.
--
Brane Čibej <br...@xbc.nu> http://www.xbc.nu/brane/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: svn commit: rev 2226 - trunk/subversion/libsvn_delta
Posted by Greg Hudson <gh...@MIT.EDU>.
On Tue, 2002-06-18 at 12:15, Karl Fogel wrote:
> Greg Hudson <gh...@MIT.EDU> writes:
> > On Sun, 2002-06-16 at 19:28, Karl Fogel wrote:
> > > The big question is, how much does it improve performance? Our big
> > > problem case was undeltifying files larger than the delta window
> > > size.
> >
> > Um, that problem case would have been due to a bug of some sort, not due
> > to the lack of a delta combiner.
>
> I'm not sure precisely what is being asserted or denied above :-).
My understanding is that, at some time in the past, it was observed that
if we deltified files above the window size, memory usage went through
the roof. That's clearly a bug; windows are there to place a ceiling on
memory usage, not make it bigger. I don't know if anyone has checked
whether this bug still exists.
A delta combiner shouldn't help much; the problem happened (I believe)
even if there was only one delta to apply.
> so my question can be
> interpreted as "Does it succeed in addressing the problems it is
> designed to address?"
It is designed to address performance problems when there are lots of
(smallish) deltas between the revision you want and a plaintext.
Without a delta combiner, an N-delta chain would take O(N*S) time to
undeltify, where S is the average size of the file. With a delta
combiner, it's O(N*D + S) time, where P is the average size of the
deltas.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: svn commit: rev 2226 - trunk/subversion/libsvn_delta
Posted by Karl Fogel <kf...@newton.ch.collab.net>.
Greg Hudson <gh...@MIT.EDU> writes:
> On Sun, 2002-06-16 at 19:28, Karl Fogel wrote:
> > The big question is, how much does it improve performance? Our big
> > problem case was undeltifying files larger than the delta window
> > size.
>
> Um, that problem case would have been due to a bug of some sort, not due
> to the lack of a delta combiner.
I'm not sure precisely what is being asserted or denied above :-).
But in any case, the purpose of the delta combiner is obviously to
address certain performance problems (if we didn't have problems, why
would Brane spend all this effort on it?), so my question can be
interpreted as "Does it succeed in addressing the problems it is
designed to address?"
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: svn commit: rev 2226 - trunk/subversion/libsvn_delta
Posted by Greg Hudson <gh...@MIT.EDU>.
On Sun, 2002-06-16 at 19:28, Karl Fogel wrote:
> The big question is, how much does it improve performance? Our big
> problem case was undeltifying files larger than the delta window
> size.
Um, that problem case would have been due to a bug of some sort, not due
to the lack of a delta combiner.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: gui client status
Posted by Alan Langford <ja...@ambitonline.com>.
At 2002/06/17 10:17 +0200, Martijn Boekhorst wrote:
>Source is forthcoming once it's no more in this much flux and had a comfy
>hosting place somewhere on the web, and not to mention has some kind of
>licensing banner.
You could always create a project on sourc\\\\\ tigris.org...
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: gui client status
Posted by Martijn Boekhorst <Su...@boekhorst.net>.
> How can we make any suggestions if you don't publish the source, or even a
> binary?
Because it's not yet fit for distribution. But since you insist.
Source is forthcoming once it's no more in this much flux and had a comfy
hosting place somewhere on the web, and not to mention has some kind of
licensing banner.
For users running XP :
http://boekhorst.net/svn/binary/visor_exe.zip
(33,3KB - binary only)
For users running anything other than XP (eg. W2K) :
http://boekhorst.net/svn/binary/visor.zip
(984 KB Includes GDI+ redistributable from Microsoft - put DLL in same dir)
have not checked for virusses but my system should be clean (check it
bi-weekly or so).
Operation :
- Create a directory in C:\Dev\Visor as this is what will be displayed in the
"FileExplorer" view on the left.
- Open File loads you a text file.
- Open File second time starts a diff.
- Open File a third time starts a three-way diff, which currently crashes
horribly since the diff iface has it commented out / does not yet support it.
Cheers, Martijn.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: gui client status
Posted by Greg Stein <gs...@lyra.org>.
On Mon, Jun 17, 2002 at 04:01:47AM +0200, Martijn Boekhorst wrote:
>...
> Suggestions welcome.
How can we make any suggestions if you don't publish the source, or even a
binary?
Cheers,
-g
--
Greg Stein, http://www.lyra.org/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
gui client status
Posted by Martijn Boekhorst <Su...@boekhorst.net>.
Hi all,
Original plan was for monday to have a binary distributable, in the good
tradition of estimating low, I under-estimated.
Current state of affairs is at :
http://boekhorst.net/svn/Visor3.png
It's almost a full-fledged diff tool, but currently missing merging
functionality. Changes from last one are mostly internal, some visible ones
include proper (final) layout, functioning scrollbars and correlation hints
(the "gooey red muccus" in between the files).
Before sending it out as a merge & diff-tool, I still need to gut the MFC
document infrastructure, resolve some drawing bugs, get mousewheel support
("it matters to me"), and add some gui support to load and save files, not to
mention have fancy hilighting buttons to do the actual merging.
Say that takes me about a week(-end), then we have a merge/diff tool a la
many commercial ones but without the directory compare or built-in editor
capability.
And then svn client functions, which "should not be too hard".
Suggestions welcome.
Cheers, Martijn.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: svn commit: rev 2226 - trunk/subversion/libsvn_delta
Posted by Karl Fogel <kf...@newton.ch.collab.net>.
Branko Čibej <br...@xbc.nu> writes:
> Whee, the delta combiner now actually _works_! Overlapping target
> copies are the final hurdle (edge cases again).
Wonderful!
The big question is, how much does it improve performance? Our big
problem case was undeltifying files larger than the delta window
size.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org