You down with OCB?


log in or register to remove this ad

Just gotta say this: while most of what you guys are saying would make perfect sense to most of my group- 80% are programmers- its been mostly impenetrable to me. Then again, I'm a "turn it on & lets go" guy. The more work you make me do to make the product a comfy fit for me, the less happy I am.

Yeah - I'm in the same boat (not the 80% of my group are programmers just the incomprehension of the tech-talk here). I don't care how or why WotC did this or that - I care about the usability of the product. What I have read is not convincing me the online CB is worth the time or bother (or money) to use it. The old CB was actually fun to use. The online one was a chore.
 

I purchased a month subscription on Friday, mostly to get the Compendium again.

1) My offline Character Builder stopped working right away after I got my new subscription. It just crashes now. That TOTALLY sucked. :rant:

2) The organization of feats is kind of lame and non-intuitive. For example, Wand Expertise is not in the Implement Expertise section. It's in the Other Heroic Tier section. WT???

3) They have a super slow interface. So, why do they have all of the menus wide open? For example, if one goes to Portraits or Feats, every item in the super long list is visible. That means that you cannot find most things quickly because you are either too busy slowly scrolling through the extremely long window, or you are shutting each menu item so that you can get to the one you want.

If the menus were defaulted closed instead, then you can open up any of the top 19 feat menus with a single button click, and you can open up any of the 7 bottom feat menus with two button clicks.

One or two button clicks to get where you want to go, not a ton of scrolling through a list with ~450 things in it (and more as they add more feats).

This is programming 101. I don't understand why a developer wouldn't see this right away. In the case of portraits, don't download 200 portraits slowly one by one. Instead, download the list of race menus and open the Dwarf portraits when the user clicks on the closed Dwarf menu. One button click to get to the Dwarve Portraits, not a ton of scrolling while every single picture they have slowly downloads from the server.

Where is the Human Factors expert on the development team? Taking a nap???

4) And the marketplace is kind of painful to use. If I'm building a PC, I want to compare similar weapons. I don't want a dropdown of every weapon where I have to select each one, click on it, read it, and do this for every single weapon I am interested in. The designers of the offline Character Builder understood that.

This design is INSANE for what it is meant for. If you want to check out every single Heavy Thrown weapon to find one you like, you either have to know which ones are Heavy Thrown ahead of time, or you have to go through every weapon in the game system to find them all. The Search can only be used to limit the list AFTER you select a Category and a single Item. So if you search on heavy thrown, you have to select Simple Melee One Handed Category and Javelin Item to see Javelin in the list, and Military Melee One Handed Category and Handaxe Item to see Hand Axe in the list.

New players would be totally clueless that certain weapon options even exist.

5) And, there are a lot of minor user interface foobards. For example, I have a Wizard. I go to the marketplace. I click on Weapons. It shows Implements and Simple Weapons. I fart around for a while and finally figure out that the checkbox phrase "Show Suggested Items" actually means "Show Proficient Items". Is English really that hard of a language to understand for a company with dozens of writers and editors based in the United States?

And why does it default to levels 1 through 30 for a first level PC? The program designers are totally clueless here.


I'm sorry, but even though they have cleaned up some bugs, this is the second worse user interface that I've seen in the last decade (Hero Builder was worse).

It has a lot of promise, especially in look and feel. But one of the first things that they need to do is to clean up the speed and to make some reasonable improvements.
 

Yeah, there are plenty of bad rough human factors considerations in there. OTOH I argue that the very fact that 4e requires a CB with 26 different feat categories is telling us something about the design of the game itself...
 

Yeah, there are plenty of bad rough human factors considerations in there. OTOH I argue that the very fact that 4e requires a CB with 26 different feat categories is telling us something about the design of the game itself...

They should simply consider the categories as if they were tags -- any given feat could have more than one tag and therefore fit into more than one category. They do that now with Suggestions.
 

Totally avoidable by using AJAX to round trip the important bits of data behind the scenes. To much 'AJAX' ends up as cludgy solutions where huge chunks of data are passed backwards and forwards, and additional rendering is required at the front end again too, but between intelligent cacheing and careful use of AJAX, there is no reason why an HTML/js app would see that kind of lag (unless the server(s) are not being kept up to spec for the usage, of course!)

But intelligent and careful use of AJAX is hard. Whereas roundtriping back to the server or using something that's already there in jQuery or the AJAX toolkit (which will pass around a lot of data if you're doing something complex) is pretty simple.
 

But intelligent and careful use of AJAX is hard. Whereas roundtriping back to the server or using something that's already there in jQuery or the AJAX toolkit (which will pass around a lot of data if you're doing something complex) is pretty simple.

Eh, my design philosophy would be to keep the client as dumb as possible. Just send it data to display, lists of choices for the user to select from, and have it send back requests for data and what the user picked. Smart use of training together requests and batching response data should keep the number of round trips fairly low while still allowing the client to be purely View and keeping model and control on the server side. Given that most data is static a decent caching strategy will pay off well too, but noting that how the client does all of that should be relatively transparent to the server side. Once you have a good framework it really isn't that hard. There just have to be a couple of clever people in on the design from day one to architect that.
 

One thing that would help a lot would be improving the logic of their Suggested / Suggestions lists. For example, Human Perseverence should NOT be suggested to someone who has already taken Resilient Focus. Conversely, someone who has taken Two-Weapon Fighting is likely to want Two-Weapon Defense.

The really interesting thing will be to see how well they use the feedback from CB to improve their suggestions. Just as Pandora uses the data from the Music Genome project to determine what type of music I like, WotC may eventually be able to predict what options I am likely to want for a character. This could be based either on clusters related to my previous choices for the character currently being edited, or even on all of my characters. For example, it's possible that I take Melee Training (X) for all of my low-Strength characters, where X is my highest ability score.
 

One thing that would help a lot would be improving the logic of their Suggested / Suggestions lists. For example, Human Perseverence should NOT be suggested to someone who has already taken Resilient Focus. Conversely, someone who has taken Two-Weapon Fighting is likely to want Two-Weapon Defense.

The really interesting thing will be to see how well they use the feedback from CB to improve their suggestions. Just as Pandora uses the data from the Music Genome project to determine what type of music I like, WotC may eventually be able to predict what options I am likely to want for a character. This could be based either on clusters related to my previous choices for the character currently being edited, or even on all of my characters. For example, it's possible that I take Melee Training (X) for all of my low-Strength characters, where X is my highest ability score.

Yeah, I could imagine them being able to say take all the choices that are made in CB and by correlating them with other choices made for similar characters they might be able to predict the 10 most likely feats you're going to want to look at next. It is a lot of fairly advanced number crunching to do it well enough to be useful, but it does sound feasible. Mostly a question of resources I imagine. Optimize suggestions, fix that pesky bug, or add that new material from last month's Dragon? I hope they will have the resources to do all of the above, but I imagine the optimizing things task is usually lowest priority...
 

I didn't mean to derail the thread with my question earlier in the thread. I'm also of the 'just make it work' mindset, mostly, despite my technical preferences.

The whole idea of online tools and a rule/monster/item/stuff database appeals to me. If my return to D&D pans out into a regular thing, I prefer DDI (or something similar) to books. I just don't have the time, money or space for an RPG collection.

But anyhoo:

In my 3e days I never much used crunch from outside the core books, because it was just too much of a bother (and I had ample time, back then). I did buy a lot of the books, though, but mostly for the fluff.

But does DDI offer any kind of fluff? Or is it all crunch, all the time?


EDIT: Having just posted this, I realize the fallacy of apologizing for derailment, then asking a question that might do the same.
 
Last edited:

Remove ads

Top