WebLoginPE v1.3.1 Bugfix (Released August 5th!)

Quote from: Soshite at Aug 06, 2008, 08:09 AM

Maybe you guys would know why this is happening...

For a while now, I’ve had an issue in WLPE where I need to set it up like this in the snippet file...

if (file_exists(MODX_BASE_PATH.'assets/snippets/webloginpe/lang/'.$lang.'.php'))
	{
		include MODX_BASE_PATH.'assets/snippets/webloginpe/lang/'.$lang.'.php';
	}

I have to change include_once to include, else the lang file never gets used, and I get a blank message.

(Which reminds me, I need to test again now to see if the "ghost" user bug still exists. Basically, WLPE wasn’t stopping accounts from being made, even if they had *badly* corrupted data. And users couldn’t register, because it gave errors like No username specified, when they has inserted one.)

I believe this is one of the possible outcomes (depending on code) of trying to pull the same include multiple times. It may very well have been being called from another included file.

If so, then it is a waste of system resources and we should make sure it is only included once at the top of the main snippet call and removed everywhere else. That’s probablly a code for the future.

I spent countless hours developing an international travel website utilizing WebloginPE 1.30.
I even made a 1.31 while Scotty was still directly involved in the project. For him to look over and make a 1.4 out of.

I pulled the zip I passed up to these forums, for three reasons: he disappeared ( I didn’t have time to directly take over the support), there may have been issues with the template subsystem (later found chunks were the best answer), it ended up causing confusion as it was too close to the original release.

Thank you Soshite for taking this on again.

In order to produce this site:
Here is what I did. Simply I used WebloginPE to create users and that is all.

I then created my own queries to manipulate the data, and displays, like multiple column users.

It appears the client is looking at expanding the website. I will install this version and run it parallel with the 1.30 I have currently working and we will see if I can help out with this new version a bit.

Quote from: BobRay at Aug 07, 2008, 12:37 AM

Quote from: Soshite at Aug 06, 2008, 08:50 PM

So I would create C:\Subversion\projects\WebLoginPE, and then click Create Repository here... with TortoiseSVN, right?

Then do I run the commands, get things set up, then upload the WLPE files?

Right. But it’s a little more complicated than that.

Before doing anything else, it’s important to first set the eol property in TSVN to "native" so the files will always have the proper end-of-line characters no matter what system the user is on
.
Go to TSVN menu | Settings | General | Edit

That will pop up the config file for editing.

Make sure this line is *not* commented out:
enable-auto-props = yes

Then add this at the end of the file:

*.php = svn:eol-style=native
*.txt = svn:eol-style=native
etc. for other file types of files in the collection

Also add (or uncomment if it’s there)

svn:keywords Date Author Revision

In case we want to use them later.

Now you’re done with the config file. Save it and exit TSVN.

Then go to an empty folder and create the repository ("create repository here").

Then create a temporary folder to hold the files. Make three subdirectories: trunk, tags, branches.

Put your wlpe files in the trunk subdirectory.

Go up one level to the temp directory.

Right click there and select "import" from the TSVN menu. Enter the path to the *repository* you created above (not where the files are). You should see the files being imported. BTW, this step failed for me because the wlpe php files have inconsistent line endings. I loaded them into my editor, PhpEd, which normalizes the eol characters and and is set to give Unix line endings (kind of a MODx standard). and saved them. That fixed it.

Now the files are in the repository. If you select "repo browser" and give the path to the repository, you should be able to see them.

You can go to another temp directory (empty) and select "checkout" to get version-controlled versions of the files back out of the repository. This isn’t necessary but it’s a comfort to see them. Select the trunk directory in the repository when you do this. Note that neither of these directories will be the official working directory you use in the future. To get that, you’ll go to the directory, click "checkout" and give the URL of the Google code repository (after you get the files there). If you right click on the files and select "properties" on the TSVN menu (not the windows menu if you’re on windows) you should see the properties you set earlier.

You can delete or empty this dir. now since you don’t need it.

Now, you want to do the dump of the repository using the directions here: http://www.shokhirev.com/nikolai/programs/SVN/backup_load.html

This creates the file you’ll upload to Google Code after setting the rev number to 0 at the bottom of the page.

Hope this all makes sense.

Bob

Sound good to me. I’ll try doing that later today.

webloginpe.class.php line 2162:
[tt]$url = $modx->makeURL($this->loHomeId);[/tt]
should be
[tt]$url = $modx->makeURL($this->liHomeId);[/tt]
as this is the ’LoginHomePage’ section.

Did SVN ever get imported for WLPE? It’d be great to get vhollo’s and the template fixes into SVN for a quick 1.3.2 release.

Quote from: rthrash at Aug 17, 2008, 08:23 PM

Did SVN ever get imported for WLPE? It’d be great to get vhollo’s and the template fixes into SVN for a quick 1.3.2 release.

I wasn’t able to do it due to various issues after trying to move over to this new notebook.

Uni’s starting up for me, so if anyone else wants to create the SVN, feel free.

The project owner has to set up the base SVN repository on the project. From there anyone that’s a team member can take care of importing the basic code for the first time; the documentation on that is a bit confusing if you’ve never done it before.

Quote from: rthrash at Aug 18, 2008, 10:22 AM

The project owner has to set up the base SVN repository on the project. From there anyone that’s a team member can take care of importing the basic code for the first time; the documentation on that is a bit confusing if you’ve never done it before.

I see. I’ll try getting it set up in a day or so (depending on if TS Fay knock off power around where I live).

Setting up the base doesn’t mean doing the initial import. I think it’s basically telling Google Code you want to have SVN and giving it a name. I’ve not done it myself, so I could be very wrong.

Quote from: rthrash at Aug 20, 2008, 08:19 AM

Setting up the base doesn’t mean doing the initial import. I think it’s basically telling Google Code you want to have SVN and giving it a name. I’ve not done it myself, so I could be very wrong.

The Google account is set up, so I think all that’s needed is to dump a local repository ( http://www.shokhirev.com/nikolai/programs/SVN/backup_load.html), set the rev number to 0 at Google, and upload the dump.

I think Soshite might be stuck at creating the local SVN repository. I outlined the steps for that earlier in this thread but may have left something out.