I find myself with a few moments to spare, so I thought I'd just write a quick update on progress with vB3.
This has been a very busy week, in which Freddie, Scott and I have all put in some very long hours towards what I consider to be the final sprint for the finish line in the race to get vB3 into the hands of you, the customer. As an added bonus, long time vB developer Mike 'Ed' Sullivan, who has been absent from the development team for several months, approached me this week about coming back and doing some more work for Jelsoft, and of course I was happy to oblige.
The last remaining task before we can release vBulletin 3.0.0 Beta 3 to the public is to get the language system into shape. While the back end of the language system has been in place for a long time now, a lot of the text that you see in the control panel and the front end interface is still hard-coded into the PHP and the templates. We are now working feverishly to get all hard-coded text into the language system, and to improve on previously defined phrases.
With an application the size of vBulletin 3, especially considering the huge control panel, there is an awful lot of text to manage so the methodology behind putting all that text into phrases is critical. If we were lazy and wanted to get the task done in the shortest possible time, we would simply create a unique phrase for every single text element in vBulletin. However, this would result in an extortionate number of phrases, which would be a nightmare for people to translate. Rather, what we are doing is identifying text elements that can be re-used and rewriting a lot of the text to be less verbose.
For example, we could create the following phrases as unique entities:
Using this approach, it is important to find a balance between putting enough description into the text to make its function intuitive, and making the text so terse that although we have a small number of phrases this is to the detriment of the ease of use of the system.
In order to do this, I created a set of tools in order to make sure that text is re-used wherever possible and that new phrases are not created when it's not specifically necessary. Meanwhile, Freddie wrote an extension to the tools to allow us to check the scripts for text and verify that all text has been successfully put into the language system.
As far as I can see, the vast majority of the control panel text is now complete, so that's the biggest hurdle overcome. Now we need to decide if we are going to give the front-end interface the same treatment before we release the public beta, or wait until the new templates are complete so as to avoid creating new phrases unneccessarily.
I think this will be the last status update before we release the public beta, barring unforseen problems. In the meantime, look out for a patch to vBulletin 2.3.0 in order to fix a problem that seems to have been introduced with MySQL 4.0.13. We will release this patch some time next week.
This has been a very busy week, in which Freddie, Scott and I have all put in some very long hours towards what I consider to be the final sprint for the finish line in the race to get vB3 into the hands of you, the customer. As an added bonus, long time vB developer Mike 'Ed' Sullivan, who has been absent from the development team for several months, approached me this week about coming back and doing some more work for Jelsoft, and of course I was happy to oblige.
The last remaining task before we can release vBulletin 3.0.0 Beta 3 to the public is to get the language system into shape. While the back end of the language system has been in place for a long time now, a lot of the text that you see in the control panel and the front end interface is still hard-coded into the PHP and the templates. We are now working feverishly to get all hard-coded text into the language system, and to improve on previously defined phrases.
With an application the size of vBulletin 3, especially considering the huge control panel, there is an awful lot of text to manage so the methodology behind putting all that text into phrases is critical. If we were lazy and wanted to get the task done in the shortest possible time, we would simply create a unique phrase for every single text element in vBulletin. However, this would result in an extortionate number of phrases, which would be a nightmare for people to translate. Rather, what we are doing is identifying text elements that can be re-used and rewriting a lot of the text to be less verbose.
For example, we could create the following phrases as unique entities:
- Modify this forum
- Edit this post
- Change your avatar
- Alter this event
Using this approach, it is important to find a balance between putting enough description into the text to make its function intuitive, and making the text so terse that although we have a small number of phrases this is to the detriment of the ease of use of the system.
In order to do this, I created a set of tools in order to make sure that text is re-used wherever possible and that new phrases are not created when it's not specifically necessary. Meanwhile, Freddie wrote an extension to the tools to allow us to check the scripts for text and verify that all text has been successfully put into the language system.
As far as I can see, the vast majority of the control panel text is now complete, so that's the biggest hurdle overcome. Now we need to decide if we are going to give the front-end interface the same treatment before we release the public beta, or wait until the new templates are complete so as to avoid creating new phrases unneccessarily.
I think this will be the last status update before we release the public beta, barring unforseen problems. In the meantime, look out for a patch to vBulletin 2.3.0 in order to fix a problem that seems to have been introduced with MySQL 4.0.13. We will release this patch some time next week.
Comment