On Wed, Jul 16, 2008 at 10:38:41PM -0500, Dean Roehrich wrote:
> I wonder if you can improve your time by using ssh compression?
All the changesets are already compressed; it shouldn't make much of a
difference, and on loaded machines, might actually make it worse.
Danek
http://cr.opensolaris.org/~mjnelson/webrev.6725705/
Cleaning up nightly.sh and the findunref Makefile, so that SCM_TYPE is set
appropriately, and ignored appropriately.
Testing: I took the modified nightly and added echo/exit after the
bringover, so I could validate the SCM_TYPE setting/resett
On Wed, Jul 16, 2008 at 05:45:24PM -0700, Danek Duvall wrote:
> On Wed, Jul 16, 2008 at 04:35:15PM -0700, Eric Schrock wrote:
>
> > hg status: 3.6 seconds
> > hg list:8.1 seconds
> > hg pdiffs: 11.7 seconds
>
> I don't recall what pdiffs actually does, but
Sorry, it appears I wasn't clear
For the record, I looked at this and had one question, which I think Rich
incorporated, and now I'm happy.
--Mark
On Tue, 15 Jul 2008, Richard Lowe wrote:
> Date: Tue, 15 Jul 2008 21:29:19 -0400
> From: Richard Lowe
> To: scm-migration-dev at opensolaris.org
> Subject: [scm-migration-dev] Ple
Please review:
4758439 some files use "current date" sccs keywords
Webrev:
http://cr.opensolaris.org/~richlowe/onnv_4758439
Again, these changes are largely incidental to output, so I haven't
bumped the copyright. The only result on output from the Java files
would be if javadoc were gene
usr/src/lib/libslp/javalib/com/sun/slp/SLPTemplateRegistry.java
- Why did you update this copyright?
Should JStyle.py check for @version and warn?
I assumed that our keywords checks would catch any remaining SCCS
keywords, if you had missed any.
--Mark
On Wed, 16 Jul 2008, Richard Lowe wrot
On Wed, Jul 16, 2008 at 01:59:55PM -0700, Bill Sommerfeld wrote:
> > You can also put:
> > [trusted]
> > groups=other
> >
> > in your ~/.hgrc to get that noise to go away
>
> please don't.
>
> in my ~/.hgrc, and I do an "hg log" on someone's workspace that's owned
> by group other, mercurial co
On Wed, Jul 16, 2008 at 11:59:28AM -0500, Will Fiveash wrote:
> In regards to point #3, will elpaso.eng/onnv.sfbay support an ssh/hg
> service as hg.opensolaris.org does? I ask this because the time to
> clone a gate to my system in Austin, TX varies greatly depending on the
> protocol used to tra
part --
An embedded message was scrubbed...
From: Stephen Kennedy
Subject: Meld 1.2 released
Date: Fri, 27 Jun 2008 00:16:52 +0100
Size: 7664
URL:
<http://mail.opensolaris.org/pipermail/scm-migration-dev/attachments/20080716/6b4b806d/attachment.nws>
>>> It's still impressive that we can burn alsost 12 seconds of CPU time
>>> to find one character difference in my workspace :-(
Really? How long does it take "wx init" to check your entire workspace
and find that same change?
Presumably your development habits aren't changing dramatically,
Alan Burlison writes:
> Mark J. Nelson wrote:
>
>> You'll want an hgignore list, something like
>>
>> syntax=regexp
>> \#.+#
>> .+\.pyc
>> proto/.+
>> .+.o
>> /debug32/
>> /debug64/
>> webrev/
>> \.make\.state
>>
>> ...and referenced in your .hgrc.
>
> Shouldn't hgsetup do this for you?
We hav
Will Fiveash writes:
> So hg status in a repository where nightly has run is displaying a lot
> output like:
>
> ? usr/src/uts/intel/zfs/debug64/txg.o
> ? usr/src/uts/intel/zfs/debug64/uberblock.o
> ? usr/src/uts/intel/zfs/debug64/unique.o
> .
> .
> .
>
> The teamware to hg guides I've seen say t
On Wed, Jul 16, 2008 at 04:10:49PM -0700, Danek Duvall wrote:
>
> I'd say "hg outgoing" is more like putback -n, except that you have to have
> your commit complete.
Well, it appears to me that hg outgoing isn't close to putback -n either. The
reason I say that is that putback -n shows you the w
On Wed, Jul 16, 2008 at 05:10:57PM -0600, Mark J. Nelson wrote:
>
>> So hg status in a repository where nightly has run is displaying a lot
>> output like:
>>
>> ? usr/src/uts/intel/zfs/debug64/txg.o
>> ? usr/src/uts/intel/zfs/debug64/uberblock.o
>> ? usr/src/uts/intel/zfs/debug64/unique.o
>> .
>>
On Wed, Jul 16, 2008 at 06:55:04PM -0500, Will Fiveash wrote:
> So how does one display all the changes that a push will perform on the
> parent repository?
Every hg subcommand that displays changesets (log, tip, outgoing, incoming)
uses -v to display a bit more information, which normally includ
On Wed, Jul 16, 2008 at 04:10:35PM -0700, Bill Sommerfeld wrote:
> On Wed, 2008-07-16 at 17:58 -0500, Will Fiveash wrote:
> > The teamware to hg guides I've seen say that hg status is similar to
> > putback -n but they do not indicate how the developer is to deal with
> > the ? status lines.
>
> w
So hg status in a repository where nightly has run is displaying a lot
output like:
? usr/src/uts/intel/zfs/debug64/txg.o
? usr/src/uts/intel/zfs/debug64/uberblock.o
? usr/src/uts/intel/zfs/debug64/unique.o
.
.
.
The teamware to hg guides I've seen say that hg status is similar to
putback -n but
Eric Schrock writes:
> On Wed, Jul 16, 2008 at 01:31:53PM -0700, Eric Schrock wrote:
>>
>> I also tried 'hg pbchk' and instantly hit an exception:
>>
>> ** unknown exception encountered, details follow
>> ** report bug details to http://www.selenic.com/mercurial/bts
>> ** or mercurial at seleni
On Wed, Jul 16, 2008 at 04:35:15PM -0700, Eric Schrock wrote:
> hg status:3.6 seconds
> hg list: 8.1 seconds
> hg pdiffs:11.7 seconds
I don't recall what pdiffs actually does, but
hg diff
should get you that information in little more time than "hg status"; or,
if you've alread
Eric Schrock writes:
> > AttributeError: 'NoneType' object has no attribute 'branch'
> >
>
> After I edited one file this exception went away - perhaps 'pbchk' has a
> problem handling zero edited files?
Yes; it's a known problem. Rich Lowe is looking into it.
--
James Carlson, Solaris Network
> So hg status in a repository where nightly has run is displaying a lot
> output like:
>
> ? usr/src/uts/intel/zfs/debug64/txg.o
> ? usr/src/uts/intel/zfs/debug64/uberblock.o
> ? usr/src/uts/intel/zfs/debug64/unique.o
> .
> .
> .
>
> The teamware to hg guides I've seen say that hg status is simil
Danek Duvall writes:
> On Wed, Jul 16, 2008 at 01:57:27PM -0700, Eric Schrock wrote:
>> It's also annoying that simple actions like 'hg status' on an untouched
>> workspace take over 5 seconds, but I assume that's just an artifact of
>> the mercurial architecture.
>
> It's gotta go stat the enti
On Wed, Jul 16, 2008 at 04:58:23PM -0700, Bill Sommerfeld wrote:
> On Wed, 2008-07-16 at 16:35 -0700, Eric Schrock wrote:
> > It's still
> > impressive that we can burn alsost 12 seconds of CPU time to find one
> > character difference in my workspace :-(
>
> if anyone wanted to do something about
On Wed, 2008-07-16 at 16:35 -0700, Eric Schrock wrote:
> It's still
> impressive that we can burn alsost 12 seconds of CPU time to find one
> character difference in my workspace :-(
if anyone wanted to do something about this, they could port this:
http://www.selenic.com/mercurial/wiki/index.cgi
On Wed, Jul 16, 2008 at 04:30:55PM -0700, Eric Schrock wrote:
> On Wed, Jul 16, 2008 at 04:16:34PM -0700, Stephen Hahn wrote:
> > * Eric Schrock [2008-07-16 21:30]:
> > > After editing a single file and switching to SSH, I now get the
> > > following times:
> > >
> > > hg status: 14 second
On Wed, Jul 16, 2008 at 04:16:34PM -0700, Stephen Hahn wrote:
> * Eric Schrock [2008-07-16 21:30]:
> > After editing a single file and switching to SSH, I now get the
> > following times:
> >
> > hg status: 14 seconds
> > hg list:29 seconds
> > hg pdiffs: 41 seconds
Bill Sommerfeld wrote:
> On Wed, 2008-07-16 at 13:37 -0700, Stephen Lau wrote:
>> I would really recommend using ssh instead of NFS.
>>
>> You can also put:
>> [trusted]
>> groups=other
>>
>> in your ~/.hgrc to get that noise to go away
>
> please don't.
>
> this is, IMHO, a really bad idea.
>
* Eric Schrock [2008-07-16 21:30]:
> After editing a single file and switching to SSH, I now get the
> following times:
>
> hg status: 14 seconds
> hg list:29 seconds
> hg pdiffs: 41 seconds
>
> I don't see how this is viable as an everyday development tool.
On Wed, Jul 16, 2008 at 03:56:06PM -0700, Stephen Hahn wrote:
>
> I believe we're in a better understood situation than these paragraphs
> suggest.
>
> For SSH, you can either use a captive shell environment, as Bill
> writes and is already implemented for hosting on opensolaris.org, or
>
On Wed, Jul 16, 2008 at 05:58:39PM -0500, Will Fiveash wrote:
> So hg status in a repository where nightly has run is displaying a lot
> output like:
>
> ? usr/src/uts/intel/zfs/debug64/txg.o
> ? usr/src/uts/intel/zfs/debug64/uberblock.o
> ? usr/src/uts/intel/zfs/debug64/unique.o
> .
> .
> .
>
>
On Wed, 2008-07-16 at 17:58 -0500, Will Fiveash wrote:
> The teamware to hg guides I've seen say that hg status is similar to
> putback -n but they do not indicate how the developer is to deal with
> the ? status lines.
what I've discovered is that:
hg status -mard
shows only changes to tracked
* Eric Schrock [2008-07-16 22:46]:
> On Wed, Jul 16, 2008 at 03:38:39PM -0700, Bill Sommerfeld wrote:
> >
> > > In the meantime, without this setting, *none of the hooks run*. That's
> > > your gate check hooks, your generate a webrev hooks, etc.
> > >
> > > We had way too many accidentally-hidd
On Wed, Jul 16, 2008 at 01:22:53PM -0500, Will Fiveash wrote:
> On Wed, Jul 16, 2008 at 01:02:46PM -0500, Will Fiveash wrote:
> > On Wed, Jul 16, 2008 at 10:56:46AM -0700, Stephen Lau wrote:
> > > Will Fiveash wrote:
> > >> On Tue, Jul 15, 2008 at 06:17:00PM -0400, Richard Lowe wrote:
> > >>
> >
...and the source code, though your mileage may vary.
The raw source code is a hidden child of the scm-migration project page.
It won't display correctly there, but if you edit the page, the asciidoc
source is intact.
The html doc is a child of the "How to transition" page in the tools
commun
On Wed, Jul 16, 2008 at 03:38:39PM -0700, Bill Sommerfeld wrote:
>
> > In the meantime, without this setting, *none of the hooks run*. That's
> > your gate check hooks, your generate a webrev hooks, etc.
> >
> > We had way too many accidentally-hidden putbacks happen due to this
> > setting not b
On Wed, 2008-07-16 at 23:18 +0100, John Levon wrote:
> On Wed, Jul 16, 2008 at 01:59:55PM -0700, Bill Sommerfeld wrote:
>
> > > You can also put:
> > > [trusted]
> > > groups=other
> > >
> > > in your ~/.hgrc to get that noise to go away
> >
> > please don't.
> >
> > in my ~/.hgrc, and I do an
Eric Schrock wrote:
>> This definitel shouldn't be mandatory, and it really shouldn't even be
>> suggested except as a hacky workaround. It's dangerous -- you can end up
>> running untrusted code as yourself just by accessing someone else's
>> mercurial workspace.
>
> So what is the internal pol
On Wed, Jul 16, 2008 at 02:24:18PM -0600, Mark J. Nelson wrote:
>
>> Thanks, that works great. Another question; where can the cscope DBs be
>> found for the gate-hg?
>
> Not currently being built.
>
> A "make cscope.out" will work correctly in a child of either repository,
> and should get all o
On Wed, Jul 16, 2008 at 11:05:19AM -0600, Mark J. Nelson wrote:
>
> It does now.
>
> Use the following paths:
>
> ssh://elpaso.eng//export/gate-hg
> ssh://elpaso.eng//export/gate-hg/usr/closed
>
> ...and it should work correctly.
>
> You may prepend "username@" to "elpaso.eng," but by d
On Wed, Jul 16, 2008 at 02:28:47PM -0700, Eric Schrock wrote:
> Should we be exposing these gates over NFS at all?
My suggestion would be to just export the working directory, which would be
kept up-to-date via one of the gate hooks. That way mercurial simply won't
recognize the directory as a r
On Wed, Jul 16, 2008 at 01:31:53PM -0700, Eric Schrock wrote:
>
> I also tried 'hg pbchk' and instantly hit an exception:
>
> ** unknown exception encountered, details follow
> ** report bug details to http://www.selenic.com/mercurial/bts
> ** or mercurial at selenic.com
> ** Mercurial Distribute
On Wed, Jul 16, 2008 at 05:10:50PM -0400, Richard Lowe wrote:
> Danek Duvall writes:
>
> > If you use cadmium, it does as wx did, and keeps track of the files
> > you're intentionally changing, and will give you that list very
> > quickly.
>
> No, that's not true.
Sorry, I don't know why I tho
On Wed, Jul 16, 2008 at 02:05:14PM -0700, Danek Duvall wrote:
>
> You need two slashes after the hostname. This is weird, but well
> established, ssh syntax. If you specify one slash, then it's relative to
> your home directory. If you specify two slashes, then it's relative to /.
> Obviously,
>> It does now.
>>
>> Use the following paths:
>>
>> ssh://elpaso.eng//export/gate-hg
>> ssh://elpaso.eng//export/gate-hg/usr/closed
>>
>> ...and it should work correctly.
>>
>> You may prepend "username@" to "elpaso.eng," but by default, it should
>> connect as you.
>
> Thanks, that wo
Will Fiveash writes:
> I don't think so. It would be good to also explain to ON peeps that
> admin a build system what needs to be done to install Cadmium.
It comes for free as part of SUNWonbld, and you'll have it in your hg
environment if you run the SUNWonbld-supplied (and one-time)
/opt/onbld
Hi Eric,
On Wed, 2008-07-16 at 13:57 -0700, Eric Schrock wrote:
> On Wed, Jul 16, 2008 at 01:37:33PM -0700, Stephen Lau wrote:
> > I would really recommend using ssh instead of NFS.
>
> The flag day only mentioned NFS access. I tried switching
> '/net/elapso.eng/export/gate-hg' to 'ssh://elpaso
On Wed, 2008-07-16 at 13:57 -0700, Eric Schrock wrote:
> On Wed, Jul 16, 2008 at 01:37:33PM -0700, Stephen Lau wrote:
> > I would really recommend using ssh instead of NFS.
>
> The flag day only mentioned NFS access. I tried switching
> '/net/elapso.eng/export/gate-hg' to 'ssh://elpaso.eng/expor
On Wed, Jul 16, 2008 at 01:57:27PM -0700, Eric Schrock wrote:
> On Wed, Jul 16, 2008 at 01:37:33PM -0700, Stephen Lau wrote:
> > I would really recommend using ssh instead of NFS.
>
> The flag day only mentioned NFS access.
It really probably shouldn't. I think it'll be more problematic than
a
On Wed, 2008-07-16 at 13:37 -0700, Stephen Lau wrote:
> I would really recommend using ssh instead of NFS.
>
> You can also put:
> [trusted]
> groups=other
>
> in your ~/.hgrc to get that noise to go away
please don't.
this is, IMHO, a really bad idea.
a trusted hgrc in a workspace can spec
On Wed, Jul 16, 2008 at 01:37:33PM -0700, Stephen Lau wrote:
> I would really recommend using ssh instead of NFS.
The flag day only mentioned NFS access. I tried switching
'/net/elapso.eng/export/gate-hg' to 'ssh://elpaso.eng/export/gate-hg',
but now I gate:
remote: abort: There is no Mercurial
"Mark J. Nelson" writes:
What is the story regarding snapshot gates like the ones found in
/net/onnv.sfbay/export/snapshot? Will there be a snapshot-hg directory
similar to the gate/gate-hg dirs?
>>> Haven't checked in a while, so my info might be stale - but at least
>>
I would really recommend using ssh instead of NFS.
You can also put:
[trusted]
groups=other
in your ~/.hgrc to get that noise to go away
cheers,
steve
Eric Schrock wrote:
> Just for fun, I tried this out today. I was able to create a mercurial
> clone, as well as a usr/closed clone underneath
Just for fun, I tried this out today. I was able to create a mercurial
clone, as well as a usr/closed clone underneath it. However, I couldn't
get it to shut up about:
---
Not trusting file /net/elpaso.eng/export/gate-hg/.hg/hgrc from untrusted
user daemon, group other
---
Spitting this out on
Mark J. Nelson writes:
> If a development team is affected by a hideous problem that caused a
> respin, it would be much more worth their time to pull up to the needed
> fix from the main branch and merge, than to deal with divergence.
I suspect that the most useful thing the gatekeepers could p
On Wed, Jul 16, 2008 at 01:02:46PM -0500, Will Fiveash wrote:
> On Wed, Jul 16, 2008 at 10:56:46AM -0700, Stephen Lau wrote:
> > Will Fiveash wrote:
> >> On Tue, Jul 15, 2008 at 06:17:00PM -0400, Richard Lowe wrote:
> >>
> > reparent is provided by Cadmium. Do you have the cdm extension loaded?
On Wed, Jul 16, 2008 at 10:56:46AM -0700, Stephen Lau wrote:
> Will Fiveash wrote:
>> On Tue, Jul 15, 2008 at 06:17:00PM -0400, Richard Lowe wrote:
>>
>>> Will Fiveash writes:
>>>
>>>
On Tue, Jul 15, 2008 at 12:31:09PM -0700, Stephen Lau wrote:
> Will Fiveash wrote:
>>
On Tue, Jul 15, 2008 at 06:17:00PM -0400, Richard Lowe wrote:
> Will Fiveash writes:
>
> > On Tue, Jul 15, 2008 at 12:31:09PM -0700, Stephen Lau wrote:
> >> Will Fiveash wrote:
> >>>
> >>> What is the story regarding snapshot gates like the ones found in
> >>> /net/onnv.sfbay/export/snapshot? Wi
>>> Okay but now I have another question. I want to use zfs snapshot/clone to
>>> clone a ON build level. Here's what I'm trying to do:
>>>
>>> zfs create pool/willf/hg/i386/onnv_94
>>> g clone -r onnv_94 /net/elpaso.eng/export/gate-hg onnv_94
>>> abort: destination 'onnv_94' already exists
>>
On Mon, Jul 14, 2008 at 06:59:29PM -0600, Mark J. Nelson wrote:
> -- Developer Responsibilities, aka What You Need To Do NOW --
>
> 1. Follow the tools flag day from last week:
>
> http://www.opensolaris.org/os/community/on/flag-days/pages/2008071001/
>
> 2. Unless you're integrating
Author: Richard Lowe
Repository: /hg/scm-migration/onnv-transition
Latest revision: 77b142ca3ea77df3d5075004bfd6c28e2cd7620d
Total changesets: 1
Log message:
6560958 Solaris:: perl modules should not use SCCS keywords in version
information
Files:
update: usr/src/cmd/perl/5.8.4/utils/lib
>>> What is the story regarding snapshot gates like the ones found in
>>> /net/onnv.sfbay/export/snapshot? Will there be a snapshot-hg directory
>>> similar to the gate/gate-hg dirs?
>>>
>>>
>> Haven't checked in a while, so my info might be stale - but at least
>> when I was there we tagged the
It does now.
Use the following paths:
ssh://elpaso.eng//export/gate-hg
ssh://elpaso.eng//export/gate-hg/usr/closed
...and it should work correctly.
You may prepend "username@" to "elpaso.eng," but by default, it should
connect as you.
--Mark
On Wed, 16 Jul 2008, Will Fivea
Will Fiveash wrote:
> On Tue, Jul 15, 2008 at 06:17:00PM -0400, Richard Lowe wrote:
>
>> Will Fiveash writes:
>>
>>
>>> On Tue, Jul 15, 2008 at 12:31:09PM -0700, Stephen Lau wrote:
>>>
Will Fiveash wrote:
> What is the story regarding snapshot gates like the o
Mark J. Nelson wrote:
> hg pull -r onnv_97
>
> Whether or not the gk's keep a separate snapshot workspace is really
> up to them, mostly for their convenience. I believe that it makes
> sense to do so, and a divergent snapshot becomes a mercurial
> repository with multiple heads.
Minor nit.
64 matches
Mail list logo
