Upgrade is needlessly difficult

I am now upgrading from 2.3 to 2.3.1. Here are the steps involved, which is a far cry from the 4 upgrade steps in the readme file:

1. Back up my database.
2. Delete all themes except for the one I modified.
3. Delete all plugins except for the ones I added.
4. Delete wp-includes directory.
5. Delete all files in main directory except for .htaccess and wp-config.php (missed that one–see next message).
6. Delete the wp-admin directory. (At this point, I am left with my .htaccess, my theme, my uploads directory, and a couple of plugins.
7. Upload all new files.
8. Run /wp-admin/upgrade.php.

Things I could have risked:

• Deleting uploads (wp-content/uploads).
• Deleting themes I modified.
• Deleting plugins I did not mean to delete.
• Deleting or overwriting files I modified, which may not necessarily include WordPress-specific files. (E.g., I could have deleted .htaccess when I removed all the files in the main directory.)

Advantages of WordPress only giving me the delta files:

• Far less risk of me deleting or overwriting files I meant to keep.
• Fewer installation steps.
• Much less downtime. It takes a while to delete the old files and re-upload every file using WinSCP over an ADSL connection.
• Less bandwidth required to run www.remarpro.com.
• Less thought required.
• Easier for newbies or non-pros to deal with.

Given all this, I just don’t understand the defensiveness and support of the current method. Who does it really hurt to make it easier to manage WordPress upgrades?

Crap! I deleted my wp-config.php file! Case in point that this is overly hard!

I don’t think the WordPress upgrade process is “hard” per se, but it’s certainly not “one click easy” – and I think a product as mature as WordPress should be striving toward a “one click” upgrade. Say, an upgrade function inside WordPress that sucks the appropriate files into the site and does an in-place upgrade.

People keep talking about not messing with the core files but isn’t it true that some of these upgrades effect themes? Should that not be our one safe haven? For some reason code has not been kept separate from design.

otto42- this comment is a bit dated, but I feel I had to respond. You said:
“If all you want is to post on a blog and you want to pay somebody else to maintain it, then do that. Why did you buy your own hosting? Why not find somebody who will manage your website for you and will leave running the blog content to you?
WordPress.com offers just that service, in fact. You can set up a site, get your own domain name, etc, etc. No technical knowledge needed.”

I’m one of those people who is almost happy with that, but wordpress.com has some pretty severe limits, e.g. no plugins, no PHP or HTML templates, etc. that make wordpress.com largely worthless to anyone who wants even a tiny bit more than the stripped down functionality provided.

That means we need to run something on our own. What I’d like to see is an upgrade process that checks the modified files, makes copies of them in a separate location, and tells you that you need to reapply changes to those files. It can still be left to the user to work out whether to overlay them, edit them, etc. The wp upgrade process is never very clean, with always some hitch in a php file somewhere, or a database schema change, or a change to one of the base templates that breaks people’s themes.

That’s why people are complaining. Telling them that they shouldn’t be using wp and should go to wp.com is either underestimating what they’re doing or overestimating what you think wp.com is good for.

One of the things you get use to in dealing with wordpress is the endless uploading and re-uploading of version after version of the software. 17 or 18 different versions in a year and change. If you are maintaining hundreds of instances, then maintaining WP software is almost a full time job.