Re: [python-committers] Survey about DVCSs compared to svn
Btw, one of my concerns is that a move away from Svn breaks the process for people who pull sources from Svn to build their own Pythons. I know a few teams that do that, and switching to another system isn't just an apt-get away for them. Do we have enough log info to be able to determine how common it is that people pull down source kits using Svn?
</F>
On Feb 28, 2009 1:11 AM, <jnoller@gmail.com> wrote:
The survey isn't biased. You have a value "the same /worse than the status quo" - wherein the status quo is subversion. If you hate DVCes, you mark it as "same/worse than the status quo" and we move on.
No one is suggesting we accept *less* functionality than subversion: in fact we're looking at these for *more* functionality.
On Feb 27, 2009 7:04pm, Fredrik Lundh <fredrik@pythonware.com> wrote: > > No need for a long answe...
On Feb 27, 2009 9:24 PM, "Brett Cannon" brett@python.org> wrote: > > I had a long reply all writt...
Fredrik Lundh wrote:
Btw, one of my concerns is that a move away from Svn breaks the process for people who pull sources from Svn to build their own Pythons. I know a few teams that do that, and switching to another system isn't just an apt-get away for them. Do we have enough log info to be able to determine how common it is that people pull down source kits using Svn?
With the limitation that svn urls are difficult to process, see
http://svn.python.org/webstats/
The actual logs themselves roll over after four days.
Regards, Martin
On Sat, Feb 28, 2009 at 1:35 AM, "Martin v. Löwis" <martin@v.loewis.de> wrote:
Fredrik Lundh wrote:
Btw, one of my concerns is that a move away from Svn breaks the process for people who pull sources from Svn to build their own Pythons. I know a few teams that do that, and switching to another system isn't just an apt-get away for them. Do we have enough log info to be able to determine how common it is that people pull down source kits using Svn?
With the limitation that svn urls are difficult to process, see
http://svn.python.org/webstats/
The actual logs themselves roll over after four days.
What on earth happened between May and June 2008? The number of "visits" went down dramatically (at least 3x), and stayed at that level ever since.
-- --Guido van Rossum (home page: http://www.python.org/~guido/)
What on earth happened between May and June 2008? The number of "visits" went down dramatically (at least 3x), and stayed at that level ever since.
As we don't have the log files anymore, it's hard to tell. My guess is that it is some automated procedure that completed or got turned off. For example, on PyPI, a single system was producing 25% of the hits for some time, trying to keep PyPI mirrored (and the operator didn't even know it was running).
For svn.python.org, here are some guesses:
- webalizer only counts accesses to PageType URLs as potential visits (and as a new visit only if there was no access within 30min from that IP). In our configuration, htm*, cgi, and py are PageType URLs. So regular svn accesses don't trigger a visit.
- in March 2008, there were 6810 accesses to /webstats/usage_200705.html (for whatever reason, from whatever hosts)
So the number of visits is probably irrelevant for that site.
Regards, Martin
Why not add a requirement to the PEP such that if a DVCS is chosen then a read-only SVN mirror or gateway should be maintained? Would that be a reasonable compromise?
I have more or less the same opinion as Guido regarding svn merge. It sucks. We bump up against problems with svn merge tracking on a regular basis at work. We'd have switched to a DVCS by now if it wasn't for tool support (trac mostly) and the fact that we use a lot of svn:externals in our repository.
On Sat, Feb 28, 2009 at 1:16 AM, Fredrik Lundh <fredrik@pythonware.com> wrote:
Btw, one of my concerns is that a move away from Svn breaks the process for people who pull sources from Svn to build their own Pythons. I know a few teams that do that, and switching to another system isn't just an apt-get away for them. Do we have enough log info to be able to determine how common it is that people pull down source kits using Svn?
</F>
On Feb 28, 2009 1:11 AM, <jnoller@gmail.com> wrote:
The survey isn't biased. You have a value "the same /worse than the status quo" - wherein the status quo is subversion. If you hate DVCes, you mark it as "same/worse than the status quo" and we move on.
No one is suggesting we accept *less* functionality than subversion: in fact we're looking at these for *more* functionality.
On Feb 27, 2009 7:04pm, Fredrik Lundh <fredrik@pythonware.com> wrote: > > No need for a long answe...
On Feb 27, 2009 9:24 PM, "Brett Cannon" brett@python.org> wrote: > > I had a long reply all writt...
python-committers mailing list python-committers@python.org http://mail.python.org/mailman/listinfo/python-committers
Bob Ippolito wrote:
I have more or less the same opinion as Guido regarding svn merge. It sucks. We bump up against problems with svn merge tracking on a regular basis at work. We'd have switched to a DVCS by now if it wasn't for tool support (trac mostly) and the fact that we use a lot of svn:externals in our repository.
We have a few of those (svn:externals) as well... Brett, another question for your DVCS champions!
Cheers, Nick.
-- Nick Coghlan | ncoghlan@gmail.com | Brisbane, Australia
Nick Coghlan wrote:
Bob Ippolito wrote:
I have more or less the same opinion as Guido regarding svn merge. It sucks. We bump up against problems with svn merge tracking on a regular basis at work. We'd have switched to a DVCS by now if it wasn't for tool support (trac mostly) and the fact that we use a lot of svn:externals in our repository.
We have a few of those (svn:externals) as well... Brett, another question for your DVCS champions!
As far as I can recall the PEP doesn't cover metadata at all. Subversion supports metadata in properties via svn prop[set|get|del]. Do the DVCS support properties, too? What about svn:ignore (ignore files and directories via glob style patterns), svn:keywords (expand keywords like $ID$), svn:executable (chmods a file a+x) and svn:eol-style (forces a line ending style).
Christian
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Nick Coghlan wrote:
We have a few of those (svn:externals) as well... Brett, another question for your DVCS champions!
Sun uses a Mercurial extension, Forest, to manage JDK.
Mercurial forest: http://www.selenic.com/mercurial/wiki/index.cgi/ForestExtension
And a proposal for the core Mercurial: http://www.selenic.com/mercurial/wiki/index.cgi/NestedRepositories
Jesus Cea Avion _/_/ _/_/_/ _/_/_/ jcea@jcea.es - http://www.jcea.es/ _/_/ _/_/ _/_/ _/_/ _/_/ jabber / xmpp:jcea@jabber.org _/_/ _/_/ _/_/_/_/_/ . _/_/ _/_/ _/_/ _/_/ _/_/ "Things are not so easy" _/_/ _/_/ _/_/ _/_/ _/_/ _/_/ "My name is Dump, Core Dump" _/_/_/ _/_/_/ _/_/ _/_/ "El amor es poner tu felicidad en la felicidad de otro" - Leibniz -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.8 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iQCVAwUBSbZ2LJlgi5GaxT1NAQL0hQP/YE2zvORq+b7320CoSWmnHqolGE/zWtcC jGfqcaSq9J1FwCeDSCjbYDAZTxuAescyCB2IZJEZtHTRgsgNnKkyk0YUGN/wf+// hWUR/QXpU68K54PAXkX8oA8RkYmFtI0rFBxOTnYbPcIauSik87ImWesLPzyMFxc4 TEpENXfYqzg= =evgj -----END PGP SIGNATURE-----
participants (7)
-
"Martin v. Löwis" -
Bob Ippolito -
Christian Heimes -
Fredrik Lundh -
Guido van Rossum -
Jesus Cea -
Nick Coghlan