Ah, probably another require I missed. Needle/Net::SSH bites again. :
( I'm am SO looking forward to rewriting that.
I'll get that require in there. Thanks for catching it.
- Jamis
On Jun 25, 2007, at 11:51 AM, [EMAIL PROTECTED] wrote:
>
> `initialize': uninitialized constant Net::SSH::Util::
>> install had no problems at all.. I'll check on the gem side to see if
>> i can track it down there. My gems were at 0.93 and then I got this
>> and figured I'd up it to latest but that didn't do it for me this
>> time..
>>
>> thanks again, at least
I certainly wouldn't pretend to speak for Mike, but I remember him
saying that he has plans for rewrite deprec for cap2. His current
work, I believe, is cap1-specific and will not work with cap2.
- Jamis
On 6/23/07, Faisal N Jawdat <[EMAIL PROTECTED]> wrote:
>
> Anyone put deprec and cap2 in a b
ch
> is why I was wondering if there was a gem package I could just
> download and install locally.
>
> Thanks again,
> Jack
>
> On Jun 25, 11:57 am, Jamis Buck <[EMAIL PROTECTED]> wrote:
>
> > Jack,
>
> > The beta gem is at:
>
> >gem install -
Thanks, Adrian. This has indeed been fixed in trunk. I've uploaded a
new gem to this mailing list (you can download it by going to
http://groups.google.com/group/capistrano and finding the files
section).
- Jamis
On 6/25/07, Adrian Danieli <[EMAIL PROTECTED]> wrote:
> Greetings,
>
> Not sure if
servers: ["server1", "server2"]
> Password: Password:
>
> Applying the patch works for me.
>
> Adrian -- -
>
>
> On Jun 25, 6:53 pm, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > Thanks, Adrian. This has indeed been fixed in trunk. I
age, but is over a year stale and
> pretty terse to begin with.
>
> Some of us can't choose what flavor of machine we deploy to (not in
> all cases at least), but we can choose to use a great deployment
> system like Capistrano. I'm hoping that where there's a will, there
It might be that the regex which Capistrano uses to recognize a sudo
prompt is not matching the prompt your remote host is using, for some
reason. If you want to dig in and see if you can figure out why that
is, that'd be great. Otherwise, I'll try and get to it this week
sometime, before I releas
gt;
> For example, this will fail:
>
> set(:foo) { fetch :user }
> fetch :foo
>
> Cheers,
> Adrian
>
>
> On Jun 25, 10:12 pm, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > Ah, I see it now. That's fixed now, too. :) Thanks!
> >
> > -
, gcnovus <[EMAIL PROTECTED]> wrote:
>
> CentOS, I believe. I'm not sure of the version.
>
> -Gaius
>
> On Jun 12, 11:23 am, Jamis Buck <[EMAIL PROTECTED]> wrote:
> > What OS are you deploying to?
> >
> > - Jamis
> >
> > On Jun 10,
Rob,
Many thanks for the excellent patch! Unfortunately, it doesn't apply
cleanly on the latest edge. Could you give it a makeover so it applies
to trunk? Thanks!
- Jamis
On 6/26/07, Rob Holland <[EMAIL PROTECTED]> wrote:
>
> I should note that I had to upgrade to a recent release of mocha in
>
Good catch. I've committed a fix for that. Thanks!
- Jamis
On 6/26/07, Tim Carey-Smith <[EMAIL PROTECTED]> wrote:
>
> Hello again,
>
> I have been using the deploy strategies outside of the common tasks. I
> noticed that the Strategy would fail to deploy if the :logger
> configuration value was
Note that the Capistrano::Shell.run method is not the same as
Capistrano::Configuration::Actions::Invocation#run method. Your
definition is resulting in an infinite recursive loop. What you need
to do is reimplement the Capistrano::Shell#exec_command method, to
extend the callback used there.
- J
Interesting idea. It seems you're proposing two things: one, a way to
define the username from the command line, and a way to define a set
of roles to use, overriding the defaults for subsequent tasks. Is that
right? If so, the following works today:
cap -s user=username ROLES=web ...
Does tha
Yeah, it's gone. It was never accurate, especially when using it with
the deployment tasks, because some tasks would depend on the output of
a previous command to determine how to proceed, and the -P could never
handle that correctly.
- Jamis
On 6/26/07, Thomas Smith <[EMAIL PROTECTED]> wrote:
>
ugh.
- Jamis
On 6/26/07, Matt M. <[EMAIL PROTECTED]> wrote:
> I see. I realized that as soon as I pressed to "submit" button! How do you
> override a method outside of the class though?
>
> Matt
>
>
> On 6/26/07, Jamis Buck <[EMAIL PROTECTED]> wrote:
&
Is c:/test_neu a directory containing your keys, or is it a key file
itself? Note that the :keys option must point to an array of actual
key filenames, and not just a directory where they exist.
- Jamis
On 6/27/07, Feurio <[EMAIL PROTECTED]> wrote:
>
> Hello,
>
> I do have a problem with doing a
t;[EMAIL PROTECTED]> wrote:
> Hey thanks Jamis, unfortunately that doesn't seem to work. What do you
> recommend I do if I don't want the username and password to appear when the
> command gets executed?
>
> Matt
>
>
> On 6/27/07, Jamis Buck <[EMAIL PROTECTED]&g
oject-x/trunk/
> /usr/local/projects/project-x/releases/20070627152315 &&
> (echo 1168 >
> /usr/local/projects/project-x/releases/20070627152315/REVISION)"
> servers: [" some-domain.com"]
> [some-domain.com] executing command
>
>
> M
ename is test_neu.pub
> But I left the .pub because capistrano seems to add it automatically.
>
> Feurio
>
>
> On 27 Jun., 16:08, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > Is c:/test_neu a directory containing your keys, or is it a key file
> > itself?
oblem here. Has anyone else seen this behavior?
- Jamis
On 6/27/07, Feurio <[EMAIL PROTECTED]> wrote:
>
> It is a SSH-2 RSA key generated with Putty Key Generator.
>
> Feurio
>
> On 27 Jun., 18:17, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > What kin
If each of those DB boxes are separate, you could put :primary => true
on each of them, since each is a primary database server. You would
only omit :primary => true for db servers that are not primary (e.g.,
slaves, which replicate from some primary server).
- Jamis
On 6/27/07, dbalatero <[EMAI
they would be separate. Is that good
> enough?
>
> On Jun 27, 11:06 am, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > If each of those DB boxes are separate, you could put :primary => true
> > on each of them, since each is a primary database server. You would
Many editors will recognize a shebang line and set the syntax
appropriately. You should be able to put #!/usr/bin/ruby (or whatever)
at the top of your Capfile. Also, you should be able to (one way or
another) make your editor recognize Capfile as a ruby script (for
instance TextMate recognizes Ra
That's...really wierd. It's like test/unit is getting loaded somehow,
because those command-line options are the options for Ruby's
test/unit lib.
It all works as advertised for me. Anyone else?
- Jamis
On 6/27/07, dbalatero <[EMAIL PROTECTED]> wrote:
>
> I just upgraded to the new cap2 beta ge
puts "You passed in #{arg1}"
> exit
> end
>
> With that exit statement, it doesn't produce the Test::Unit error.
> Whatever is happening, happens after the method runs and Capistrano
> does whatever final teardown it needs.
>
> - David
>
> On Jun 27, 1
you know if this happens again, but
> > I guess consider this issue magically solved ;)
> >
> > - David
> >
> > On Jun 27, 12:22 pm, dbalatero <[EMAIL PROTECTED]> wrote:
> >
> > > I modified my task:
> >
> > > task :test_cap, :ro
Do you customizations need to go before the finalize_update task, or
can they go after the entire update_code task finishes?
Either way, you can put your customizations in their own task:
task :my_customizations do
# ...
end
And then instruct Capistrano to call it either before finalize
Excellent work, Rob. I've committed your patch, with a few minor
modifications. I went with :on_error instead of :on_errors, among
other things. Thanks, especially for including tests!
- Jamis
On 6/27/07, Rob Holland <[EMAIL PROTECTED]> wrote:
> > Many thanks for the excellent patch! Unfortunat
Ken,
Capistrano doesn't provide any way, out of the box, to run multiple
processes in parallel on a single box. However, it isn't hard at all
to do with Capistrano: just write a shell script on each server, which
spawns the processes and waits for them to finish. Then, invoke that
shell script vi
Yeah, that's a known issue that I keep forgetting about. I'll try to
make that smarter, so it only breaks on a period if there is
white-space after it.
Bugs may be reported at any time via the Rails bug tracker at
http://dev.rubyonrails.com. Just specify the product as "Capistrano"
when filing th
I am such a chicken. I very much wanted the next release of Capistrano
to be the official "Capistrano 2.0" release. But as I watched the
changelog grow, I started to get cold feet.
Thus, tonight I announce the fourth (and final, hopefully!) preview
release of Capistrano 2.0. As before, you can gr
Because some tasks use sudo, and :runner is used (in cap1) to indicate
who to sudo as. If you are simply setting :runner to yourself, you
should probably just disable sudo by setting :use_sudo to false.
- Jamis
On 7/2/07, Andy <[EMAIL PROTECTED]> wrote:
>
> Can someone please explain why i have
If you aren't using a spin file, you'll need to reimplement the
deploy:start task, even if it's just to make the start task do
nothing:
deploy.task :start do
# nothing
end
- Jamis
On 7/2/07, Andy <[EMAIL PROTECTED]> wrote:
>
> I am a little confused about cap2's need for a spin file wh
o at times (such as run "sudo /usr/sbin/
> apachectl stop")
>
> Thanks
>
> On Jul 2, 5:12 pm, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > Because some tasks use sudo, and :runner is used (in cap1) to indicate
> > who to sudo as. If you are s
with OpenSSH(1:4.3p2-5~bpo.1) on the server(from
> > backports.org), Capistrano (v1.99.2) and Net:stp(1.1.0) on the client.
> > Client is running debian etch(stable), server debian sarge(oldstable).
> >
> > Server is configured with public key auth.
> >
> > non-ruby sftp comm
1.8/gems/net-ssh-1.0.10/lib/net/ssh/session.rb:
> 134:in `ini
> tialize': app (Net::SSH::AuthenticationFailed)
> from c:/ruby/lib/ruby/gems/1.8/gems/net-ssh-1.0.10/lib/net/
> ssh.rb:47:in
> `start'
> from utils/ssh-client.rb:46
>
> On Jun 27, 10:03
The problem is that you have a task named "deploy", which (as the
error message states) conflicts with an existing namespace with the
same name. Mostly likely you've got a custom implementation of the
deploy task, which is named deploy:default in cap2. To override the
default task in cap2:
depl
Which version of capistrano are you using?
- Jamis
On 7/5/07, Nikolay Kurtov <[EMAIL PROTECTED]> wrote:
>
> I'm new to Capistrano.
>
> When I do "cap deploy" it connects to the server, begins to connect to
> CVS, but asks for password and freezes :(
> What can i do with it?
>
>
> >
>
--~--~
Good point. I'll tweak the task so that it aborts when the dependency
check fails.
- Jamis
On 7/6/07, Chris Bailey <[EMAIL PROTECTED]> wrote:
> I've wondered about this as well, as I had the same setup. I haven't had a
> chance to dig in to see what a solution would be, and will be interested t
to scope the depend to a role?
>
>
> On 7/6/07, Jamis Buck <[EMAIL PROTECTED] > wrote:
> >
> > Good point. I'll tweak the task so that it aborts when the dependency
> > check fails.
> >
> > - Jamis
> >
> > On 7/6/07, Chris Bailey <[EMAIL
Capistrano does not have any notion of local vs. remote. It's purely a
remote administration tool. You'll need to handle that as a special
case (e.g., not add the local host to a role and instead do the copy
manually in the task, using something like FileUtils).
- Jamis
On 7/5/07, ry <[EMAIL PRO
a password", prefix
> ch.send_data "#{actor.password}\n"
> elsif out =~ %r{^Enter passphrase}
> message = "CVS needs your key's passphrase and cannot
> proceed"
> actor.logger.info message, prefix
>
Wayne,
What version of cap were you using before upgrading to 1.99.3?
- Jamis
On 7/2/07, Wayne E. Seguin <[EMAIL PROTECTED]> wrote:
>
> Jamis,
>
> After upgrading to 1.99.3 (PR 4) several people on my team noticed
> that deploy hangs on the first command which attempts to connect to
> the serve
Yes, it was added in preview release #3.
- Jamis
On 7/6/07, Wes Gamble <[EMAIL PROTECTED]> wrote:
>
> There's a CVS module for cap2 now?
>
> Wes
>
> Jamis Buck wrote:
> > Thanks Nikolay. It looks like this is already fixed in the CVS module for
> >
Well, I've spent a couple hours this afternoon trying to duplicate
this. I can't duplicate the "hang" effect, but I have been able to
consistently trigger another failure: "Corrupted MAC on input".
I've only been able to trigger that when uploading a larger file over
at least two connections, via
e file in and then using put instead of
> using put_file?
>
> On Jul 7, 4:43 pm, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > Well, I've spent a couple hours this afternoon trying to duplicate
> > this. I can't duplicate the "hang" effect, but I ha
If you have the directory under the shared directory, and then symlink
from your deployed version to shared after deployment, the files will
be preserved between deploys. It's only when your files are being
stored directly under the deployed revision that you run into problems
with files going awa
ustom_picture"
> servers: ["domain.com"]
> [domain.com] executing command
> command finished
>
> yet i'm not seeing those symbolic arrows pointing properly, obviously
> not working..
>
>
> On Jul 7, 7:22 pm, "Jamis Buck" <[EMAIL
If you're using an ssh-agent, make sure you do NOT have
ssh_options[:keys] specified anywhere in your deploy.rb. (It looks
like Net::SSH is trying to load the key directly, rather than through
the running SSH agent.)
How did you generate that key originally? Also, is it an RSA key, or
an RSA1 key
Thanks for all the data, that's great stuff. I'm still at a loss as to
what the difference is... especially if it works in cap1 but not cap2,
against the same Net::SSH version. If it isn't too much to ask, could
you try comparing the sftp dumps against a cap1 session, to see if
there is any differ
your UI, which will be served via
the image_tag helper and friends.
- Jamis
On 7/10/07, Eugenol <[EMAIL PROTECTED]> wrote:
>
>
>
> On Jul 8, 4:22 am, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > If you have the directory under the shared directory, and then s
;
> Hello Jamis,
>
> Of course they aren't stored in public/images. But it is still curious
> that capistrano halts so much time at that step. Any idea ?
>
> Regards
>
>
> On Jul 10, 3:57 pm, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> >
Lee,
One technique I use when something does seem to quite "fit" anywhere
is to put it in a module, in the lib/ directory. Then I just include
that module in (e.g.) the application controller, or a model.
If I don't even know where to include it, though, another option is to
break it out into it
Sorry about that. That was a bug that crept in with the introduction
of the :on_error => :continue feature. I've just committed a fix. In
the meantime, you can apply this patch to get things working for you
again:
Index: lib/capistrano/configuration/connections.rb
Nope, that's about the way I'd do it, too. You can clean it up a bit
by using default values for the Hash:
all_data = Hash.new { |h,k| h[k] = "" }
Then you don't have to test for whether any particular value in the
hash is nil. You can just append to it blindly:
all_data[channel[:host]] <<
7/10/07, Eugenol <[EMAIL PROTECTED]> wrote:
>
>
>
> On Jul 10, 5:24 pm, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > How many files have you got in public/images, public/stylesheets, and
> > public/javascripts? That line just takes a few seconds to e
You might find some enlightenment here:
http://groups.google.com/group/capistrano/browse_thread/thread/389fa347ab955fbd/013e03ccfc7ac8bb?lnk=gst&q=capistrano%3Ahost&rnum=1#013e03ccfc7ac8bb
Or, if google totally munges that url:
http://tinyurl.com/2q6xuq
The short summary answer is that, n
t, it's running that check on my DB machines. I get this back on a
> deploy:check:
>
> --> `gem' command could not be found. Try setting :gem_command
> (vip-db.mydomain.com)
>
> On 7/7/07, Chris Bailey <[EMAIL PROTECTED]> wrote:
> > Jamis, that's entirely
Note that 100% CPU consumption is a known issue with Net::SFTP, and is
not necessarily related to the upload hanging. Even when everything
works correctly, you will still see huge CPU usage by Net::SFTP.
- Jamis
On 7/11/07, Andrew Beacock <[EMAIL PROTECTED]> wrote:
>
> Just a little more informa
Well, you could install your own logger instance, which logged to a
StringIO or something. Unfortunately 1.99.3 doesn't let you set the
logger directly, but you can do something like this:
output = StringIO.new
config.instance_variable_set(:@logger,
Capistrano::Logger.new(:output => outp
Can you show me your deploy.rb file?
- Jamis
On 7/11/07, sol <[EMAIL PROTECTED]> wrote:
>
> Hey,
>
> Shouldn't deploy_to affect the current_path var?
>
> I set deploy_to to /var/apps, and if I try to cap restart, it still
> uses the /u/ prefix.
> release_path uses the correct /var/apps prefix.
>
ECTED]> wrote:
> jamis, if it would help.
> I could see if my hosting provider can get you access to an instance on
> the machine.
>
>
>
>
>
> On 7/11/07, Jamis Buck <[EMAIL PROTECTED]> wrote:
> >
> > Note that 100% CPU consumption is a known issue wit
p a bit, but I don't have the time right
> now, so it might seem a bit confusing :)
>
> http://pastie.caboo.se/78103
>
> On Jul 11, 8:54 pm, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > Can you show me your deploy.rb file?
>
>
> >
>
--~--~---
but it still uses /u/ for current_path
>
> On Jul 11, 9:20 pm, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > I wonder if something in mongrel_cluster/recipes is grabbing
> > current_path before you set the deploy_to variable... you might try
> > moving tha
> I don't think that this affects the path though.
>
> On Jul 11, 9:42 pm, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > In the pastie, there is a line that says:
> >
> > # some custom app vars here
> >
> > Is that literally in your fil
Yup, that set :mongrel_conf is the culprit. Try this instead:
set(:mongrel_conf) { "#{current_path}/config/mongrel_cluster.yml" }
Putting it in braces forces it to be evaluated lazily, when it is
first needed. Which will hopefully be after you've changed the
:deploy_to variable.
- Jamis
On 7/
You can pass the block implicitly:
set(:user) { Capistrano::CLI.ui.ask("...") }
- Jamis
On 7/13/07, Edward Ocampo-Gooding <[EMAIL PROTECTED]> wrote:
>
> I'm trying out the 1.99.3 preview and was wondering if there's a cleaner
> way to write the following:
>
> set :user, Proc.new { Capist
>
> Any way I could get this down to
>
> set(:user) { prompt("...") }
>
> ?
>
> Jamis Buck wrote:
> > You can pass the block implicitly:
> >
> > set(:user) { Capistrano::CLI.ui.ask("...") }
> >
> > - Jamis
> &g
This is because you are using cap1, and cap1 uses "svn log" to
determine the latest revision, and "svn log" won't pick up any new
revisions since the last "svn up".
In other words:
1. svn up, note the revision number.
2. svn commit any number of times.
3. svn log, note the last revision number
ommand I came across, "cap update_current", likewise fails because
> "Expected 'log' to be a directory but found a file"
>
> Is there an svn command that helps it follow symlinks?
>
> I really appreciate the help I'm getting here!
> S
>
>
solaris is just wierd. I'm sure it is something with solaris' tools
using a slightly different syntax than what cap is feeding it. Beyond
that, though, I have no idea why you're seeing what you're seeing, but
if you figure it out, I'd be curious to know.
- Jamis
On 7/16/07, [EMAIL PROTECTED] <[E
Currently, no, there isn't a way to specify a per-role gateway, but
that might make a nifty patch if anyone is feeling adventurous. As
things stand, I'm swamped just trying to get cap2 out the door...
- Jamis
On 7/16/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
>
> Hello,
> I was wounder if
You've got the scm_password value in quotes...don't do that. :) Do:
set(:scm_password) { Capistrano::CLI.password_prompt("...") }
- Jamis
On 7/16/07, gcnovus <[EMAIL PROTECTED]> wrote:
>
> I'm running preview release 2, and I'm getting the problem Merg
> mentioned a while back with the follow
Assuming you are using rubygems to install both 1.4.1 and 2, you can do:
cap deploy # uses cap2
cap _1.4.1_ deploy # uses cap 1.4.1
Some have set up aliases (e.g. "cap1" and "cap2") which use the
appropriate command-line configuration.
Furthermore, if you want to make sure that a project is
You need to get tricky to work around the confusion with the mysql
namespace. There's no collision going on there, but rather Capistrano
is having a hard time realizing that you want the mysql namespace at
the root level, rather than the one within the calling scope.
What you'll need to do is gra
Yup, good catch, needs to be a task.
- Jamis
On 7/18/07, Brian Smith <[EMAIL PROTECTED]> wrote:
>
>
>
> On Jul 18, 3:42 am, peterBlahutka <[EMAIL PROTECTED]> wrote:
> > I got file mysql.rb with methods
> >
> > Capistrano::Configuration.instance(:must_exist).load do
> > namespace :mysql do
> >
It's actually been on the upgrade FAQ for awhile :)
http://www.capify.org/upgrade/faq
- Jamis
On 7/18/07, Thibaut Barrère <[EMAIL PROTECTED]> wrote:
>
> I had exactly the same question; thanks! Should it go to a FAQ section
> on the site ?
>
> -- Thibaut
>
>
> >
>
--~--~-~--~~---
I just committed a fix for this. In the meantime, you can add the
following lines to your deploy.rb to work around it:
require 'net/ssh/service/shell/services'
require 'net/ssh/service/shell/driver'
Once cap2 is done, I am SO going to rewrite Net::SSH and Net::SFTP. :(
- Jamis
On 7/18/07,
Not deleted no: that faq is for Cap 1.4.1, and is still valuable.
Probably, though, a note should be added that it is 1.4.1-specific, if
nothing else.
- Jamis
On 7/18/07, Ray Baxter <[EMAIL PROTECTED]> wrote:
>
> Jamis Buck wrote:
> > It's actually been on the
wrote:
>
> Jamis,
>
> Have you been able to investigate this further? I'm still getting the
> issue (even when running with just one role). "require 'fastthread"
> seems to help but not solve the problem.
>
> Andy
>
> --
> Andrew Beacock
> b
Neat idea, Mat. I don't know of anything that currently does that, but
if you find a nice way to implement it, I'd certainly consider adding
it to cap. Bonus points if it works on Windows, too. :)
- Jamis
On 7/19/07, Mat Schaffer <[EMAIL PROTECTED]> wrote:
>
> Just had an idea flash, that might
t; Andy
>
> --
> Andrew Beacock
> blog: http://blog.andrewbeacock.com
>
>
> On Jul 19, 2:48 pm, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> > Andy,
> >
> > I'm investigating even as we speak. I was set up with an account that
> > exhibited th
Sorry about that. The page isn't ready yet. Somehow it got linked up
anyway. I've removed the link.
- Jamis
On 7/20/07, Jacob Atzen <[EMAIL PROTECTED]> wrote:
>
> Hi,
>
> It seems the community page of capify.org is gone awol. Maybe it was
> never there in the first place? Anyways, I recently r
Try:
role :other, "server2", :no_deploy => true
- Jamis
On 7/21/07, jemminger <[EMAIL PROTECTED]> wrote:
>
> hello,
>
> using cap 1.99.3
>
> i have a few roles defined:
>
> role :web, 'server1'
> role :app, 'server1'
> role :db, 'server1'
> role :other, 'server2'
>
> when i do a deploy, cap a
Thanks David, I can duplicate it now with that sequence of commands.
I'll investigate and see what I can find out.
- Jamis
On 7/21/07, David Lemstra <[EMAIL PROTECTED]> wrote:
>
> It seems this problem is popping up more than once.
> See groups.google.com/group/capistrano/browse_thread/thread/
Thanks for the prod, Neil. Mostly I'm just fighting my way through
some suffocating frustration over Net::SSH and Net::SFTP's ABOMINABLE
design. I so hate myself for writing that code. I'm going to suppress
my gag reflex long enough to release cap2 (with some caveats, since
certain things are brok
In general, I'd recommend doing what Brian demonstrates with variable
interpolation. However, there are times when an ERB template is
preferable, and it looks like you're doing that just fine. The upload
error you're seeing is unrelated to the template; it looks like you
might be trying to store t
d_servers ={}
> servers.each{|x| gated_server.merge(which_gateway(x))}
> gated_servers.each{|name,server_array|
>
> Thread.new
> #from establish_connections_to
> connection_factory
>
> threads = Array(server_array).map { |server|
> establish_connection_t
>
>
> On Jul 19, 2007, at 6:57 PM, Tim Carey-Smith wrote:
>
> >
> > * Jamis Buck <[EMAIL PROTECTED]> [2007-07-19 15:41:42 -0600]:
> >
> >> Neat idea, Mat. I don't know of anything that currently does that,
> >> but
> >> if you find a
The short answer is this: if you have a file that you want under
version control, but which you do not want to deploy with your
application, then you need to make sure the file is stored in your
source repository somewhere other than under your project root. This
may mean you need to refactor your
I'm not 100% sure why this happens, but I have figured out how to work
around it. I've committed a fix to Capistrano, which you can test
yourself; just add this to the top of your recipe file:
HighLine.track_eof = false
That fixed the problem for me.
- Jamis
On 7/13/07, Edward Ocampo-Gooding
You can set global SCM command line options this by setting the
scm_command variable:
set :scm_command, "svn --ignore-externals"
That won't let you tweak based on the specific svn command being
issued, though. If someone wants to investigate this and submit a
patch, I'd consider a cross-SCM, t
Sorry I took so long to get to those patches, Tim. Thanks for putting
them together!
- Jamis
On 7/1/07, Tim Carey-Smith <[EMAIL PROTECTED]> wrote:
>
> Hey all,
>
> Did some hacking on the capistrano source.
> Results are here.
> http://halorgium.net/cap-hacks
>
> Thoughts are welcome.
>
> Cheers
I fully expect this announcement to be eclipsed by the far more
world-shattering event of the release of Harry Potter #7, but I
suppose it runs in the family. (I have a stagecoach robber in my
family several generations back, whose pièce de résistance happened to
coincide with the sinking of the T
s. I swear 98% of what I've
> > written in that time was written by some rank amateur posing under my
> > name.
> >
> > :-)
> >
> > NeilW
> >
> > On 21 Jul, 21:49, "Jamis Buck" <[EMAIL PROTECTED]> wrote:
> >
> > > T
EMAIL PROTECTED]> wrote:
> >
> >
> >
> > > Thanks Jamis for the update, if you need any help testing out patches
> > > or alternative suggestions please let me know - I'd be happy to help
> > > get to the bottom of this one :)
> >
>
Mike Bailey suggested a more robust way to detect the sudo password
prompt, using the "-p" switch to sudo to specify a standard prompt
that Capistrano can recognize. However, before I implement that, I'd
like to know how wide-spread the availability of that switch is. I've
verified that it works o
If you do "cap -h", you'll see the following included in the output:
-q, --quiet Make the output as quiet as
possible (default)
-v, --verboseBe more verbose. May be given more
than once.
It says that "-q" is the default, but it isn't really. I'll fi
Yes, the deploy:symlink task _only_ adds the current symlink. The
other symlinks are handled in deploy:finalize_update.
- Jamis
On 7/23/07, Lei <[EMAIL PROTECTED]> wrote:
>
> Hi,
>
> The Symlink function in cap 2.0 is not linking the public and shared
> directory.
>
> on line 197 of /usr/lib/rub
401 - 500 of 1553 matches
Mail list logo