Publish to live site from Subversion

I’ve been using Subversion for a while to maintain the PHP web apps and
sites I develop in my Windows box. But when I want to publish a new
release I update the live site manually and I want to improve this process.
>
I can’t simply export to a fresh directory and upload all files because
that won’t handle renames and deletions (and the upload will take
forever anyway). I typically use WinMerge to find out what files changed
but there’re some drawbacks:
>
* It can’t compare files through FTP (and if it could, I guess it’d need
to download the whole site every time) so I need to maintain a local
copy of the live site and upload files manually.
>
* It can’t use the repository information so it won’t ignore
non-versioned items (settings, uploads, caches, data…) unless I maintain
a separate filter duplicating the svn:ignore data.
>
I could keep my current method and write a bunch of Subversion hook
scripts in PHP (I’m not proficient in any other language available) to
keep the data in synch but I looks like a overwhelming effort that can
easily break. So I wonder how is everyone else doing it. But according
to my googling the answer is: writing lots of custom scripts!
>
Is it the only way? Is there a tool to help with this common task? Can
Apache Ant be used for this? How do *you* do it?
>
>
>

I’ve been using Subversion for a while to maintain the PHP web apps and
sites I develop in my Windows box. But when I want to publish a new
release I update the live site manually and I want to improve this process.
>
I can’t simply export to a fresh directory and upload all files because
that won’t handle renames and deletions (and the upload will take
forever anyway). I typically use WinMerge to find out what files changed
but there’re some drawbacks:
>
* It can’t compare files through FTP (and if it could, I guess it’d need
to download the whole site every time) so I need to maintain a local
copy of the live site and upload files manually.
>
* It can’t use the repository information so it won’t ignore
non-versioned items (settings, uploads, caches, data…) unless I maintain
a separate filter duplicating the svn:ignore data.
>
I could keep my current method and write a bunch of Subversion hook
scripts in PHP (I’m not proficient in any other language available) to
keep the data in synch but I looks like a overwhelming effort that can
easily break. So I wonder how is everyone else doing it. But according
to my googling the answer is: writing lots of custom scripts!
>
Is it the only way? Is there a tool to help with this common task? Can
Apache Ant be used for this? How do *you* do it?
>
--
--http://alvaro.es- Álvaro G. Vicario - Burgos, Spain
-- Mi sitio sobre programación web:http://bits.demogracia.com
-- Mi web de humor al baño María:http://www.demogracia.com
--

sites I develop in my Windows box. But when I want to publish a new
release I update the live site manually and I want to improve this process.
>
I can’t simply export to a fresh directory and upload all files because
that won’t handle renames and deletions (and the upload will take
forever anyway). I typically use WinMerge ...

>
Why? Why don't you just do an "svn update" on the web server?

Failing that (because subversion is not installed, for example), you
could try to mount the site directory somehow to your computer and
use the subversion installation on your local PC to do the update.

Álvaro G. Vicario wrote:

I use svn or cvs here (depending on the project). I just keep a
development system then use one of the many FTP programs which compares
file date/time (ie. WinSCP, FileZilla, CoreFTP).

The only hassle is that I need to open my repositories to the outer
world.

world.

I recall having tried Windows' "Network Places" to mount an FTP site but
I discarded it because it was pretty useless (it would not allow command
line, drive mapping or shell extensions). But perhaps there's a
third-party tool I could use.

"�lvaro G. Vicario" wrote:>
As long as you're configuring apache properly, you'll open *only* the
subdirectories you want, not the enrite SVN tree.
>
Or am I missing something there?

What I meant is that, currently, all my repositories are limited to the
local network (or even with no network at all). To access them from the
internet I'd need to forward a port in the router and double-check
Apache configuration. Which is doable, but not done.