We are using "svn update" to deploy. It usually happens that you may find bugs or add new features after the system is deployed for the first time. Then you will appreciate how convenient "svn update" is. To make it safer, you could use svn branches to make sure you only "svn update" on the production server when you think the code is stable enough.
Ah yes this solution did cross my mind. Is this possible on a shared host however?
Just did some googling and it seems to get it working I first need to learn how to connect to the server with SSH. So it is possible. So I’m going to do some playing around with that. You pointed me in the right direction. Thanks!
EDIT: just got putty/ssh working on dreamhost. This is cool stuff… svn is already installed
EDIT2: OK it seems like its gonna work on my other host too. This is gonna save a lot of time
I’ve never heard of mounting a server as a remote drive… just googled it and it sounds pretty nifty/cool. I like using svn update as apposed to copying the files over as it takes less time for the server to do, as it only sends the file changes.
I just checked if the .svn folders were accessible and luckily they are not (for whatever reason).
I use plink (command line version of putty) to deploy the application via a .bat file (one click!). I am also planing on creating .bat files to bring up and down the server for public use (by replacing index.php probably).
I hear git has a lot of improvements over svn so I hope to learn it someday when I have time…
For updating the database, I have always just done it manually on the server phpMyAdmin script. Not the cleanest method… Whenever I made a change on my local database for testing, I would just jot down the sql that phpMyAdmin executed, then executed the same thing on the remote server phpMyAdmin once deploying. Haven’t needed to convert any data yet. So I’m going to try out this MySQL Workbench and see what it’s about…
I know, that everyone is knowing it, but i want to mention rsync for every non-svn/git/googlecode project around. Initial sync takes long (since it copies up ALL data) but after that, only the changes are ‘commited’ which makes it very fast ( way < 1 second )…
The beta of MySQL workbench seems to have a lot of bugs right now. For instance, after altering a table, the changes don’t show until I refresh my object explorer, and when I do that, I have to re-navigate to where I was. Besides the bugs it’s perty good tho
On all of my past projects I simply uploaded new/updated files via FTP without putting site offline or something. It worked well since the overall amount of files were always small. I’m looking forward to use a version control system to do the updates of new projects.
You guys know if there’s an offline svn/git/bazaar solution (maybe a bundle with webserver or something)? Because for now I work alone but I would still like to use versioning and then later put everything online so others can access.