Idea: Non-English EU realms without history?

[Erorus]

There are two big problems with supporting Non-English EU realms, and a few smaller problems. The two big problems are:
1) I have no room on the SSD to maintain an auction history for the rest of the EU realms
2) Web pages and contact responses will remain in English, as I don't have the structure in place to support multiple languages

The small problems, concerns really, are the increase in bandwidth going to my auction database, and the increased site load in general from supporting more visitors.

I'm told that big problem #2 isn't really that big, that many non-English speakers (or, at least, not native English) wouldn't mind using the site in English. Also, if required, I could set up the database to store and return other languages' tooltips, and search by other languages' item names.

Big problem #1 is hard to get around, though. I don't see the sense in dropping a good chunk of cash and a bunch of time on what would end up being a whole new server (new SSD, and a new motherboard to support more RAM for the database, which means a new CPU...), when I couldn't support the other languages properly.

However, maybe with a few omissions, we can maintain a lot of the site functionality and include the rest of the EU realms. The biggest data hog is the auction history. Of the 40.1 GB on the auction database's SSD right now, 36.3 GB is the ~275 million auctions over the past 14 days from 350 realms. I really can't optimize that table any further. If we didn't have auction history for those other realms, we'd lose some really neat parts, but keep most of the rest of the site functional for those other realms.

Just want to make it clear: the current English-language realms wouldn't be changed at all. The omissions would only be for non-English realms.

So, with the following features, would it be worth it to those users to support those other realms? (Included features are marked as "Yes", features not available are marked as "No")

Front Page
Syndicated Editorials - Yes (English only, natch)
Hot Items list - Yes
Great Deals list - Yes
Busy Sellers list - No
Market API - Yes
Addon - Yes

Item Page
Top Stats table (Available, Market Price, Mean, Std dev, 95% CI, Last Seen, etc) - Yes
Price and Availability History - No
Post Activity Heat Map - No
Sold Activity Heat Map - No
Market Notifications - Yes
Other Markets scatter chart - Yes
Component Parts - Yes
Recent Sellers - No
Sold Auctions - No
Current Auctions - No

Seller Page
Armory Lookup - Yes
Auction Listings by Hour - No
Post Activity Heat Map - No
Sold Activity Heat Map - No
Posts by Item Class - No
Seller Notifications - Yes
Sold Items - No
Current Items - No
(Basically, the only seller page function left is the notifications, which may or may not be worth having a seller page at all.)

Category Pages
Deals category - Yes
All other categories - Yes, with every feature currently on every page

So, would that be enough to be worth it? Or would it be a useless embarrassment not worth putting together?

[Sinshroud]

This is quite an interesting idea @Erorus . I think a big concern is also for the amount of money and effort involved, how much will the service be used? For example you don't want to go through all this trouble to set it up for only say 10 people to use it.

Your ideas above sound like a good compromise. You can implement the easiest and cheapest options first, and see how much use comes out of it. If there is enough interest you could start adding more features such as other language tooltips or searches, maybe EVENTUALLY fully support like the rest of the realms - if there is enough demand to say cover the costs via donations / subscriptions.

I also think this could be a good experiment to see which features of TUJ could effectively be removed in times of great need (emergencies, bad host performance etc). Perhaps even being able to implement a system where realms that don't get used very frequently get some of the less important features removed until they are more in demand again.

[AceSigma]

Great idea, i play on a non-english eu-realm.

You have seen the many requests to support eu-realms in the forum, so i think, there is a demand.

Two questions:
- Will it be possible to export the ah-data to use in a spreadsheet?
- Why it has to be a ssd; doesn't work a "cheaper" server-quality hd?

Greetings

[Mikasha]

Hello Erorus,

I believe that right there is the most awesome idea i have heard in a while. I play on a non-english eu-realm as well, but my game is installed in english. I hope i'm not completely wrong in saying that i believe that most people that visit this website and speak english probably even play the game in english especially since most translations that Blizzard has done in other languages (for titles, quest names, mobs) sound completely ridiculous.

Now to the second part. I have been using for a while now wow-trader and sincerely imo the only really important thing that it is missing would be the market notifications. I understood that the owner of wow-trader was planning on adding it but got rl problems and couldn't get to it yet. From my side i understand perfectly your issue in adding the history, but i know for sure that if you decide to do this i will use tuj for market notifications and just go to wow-trader for market history and you will have my life-long gratitude for it.

[Sinshroud]

Also 1 thing, I didn't see you mention whether or not XML downloads - I think that is a big thing for a lot of people to be able to plug in that data into spreadsheets.

[TraX]

I think that is a big thing for a lot of people to be able to plug in that data into spreadsheets.

Eeeeeeexactly

Would be cool if I would be able to use the shuffling-sheet the way it's meant to be used

[Erorus]

Your ideas above sound like a good compromise. You can implement the easiest and cheapest options first, and see how much use comes out of it. If there is enough interest you could start adding more features such as other language tooltips or searches, maybe EVENTUALLY fully support like the rest of the realms - if there is enough demand to say cover the costs via donations / subscriptions.

Yeah, it's a good way to measure demand.

I also think this could be a good experiment to see which features of TUJ could effectively be removed in times of great need (emergencies, bad host performance etc). Perhaps even being able to implement a system where realms that don't get used very frequently get some of the less important features removed until they are more in demand again.

Well, in this case, it's a matter of data collection. Either I collect it all for a realm, or I forget the auctions after the stats are updated and notifications are sent out. It's not something you turn on and off, really.

Also 1 thing, I didn't see you mention whether or not XML downloads - I think that is a big thing for a lot of people to be able to plug in that data into spreadsheets.

- Will it be possible to export the ah-data to use in a spreadsheet?

That's called the "Market API" on the site, and it's marked above as "Yes."

- Why it has to be a ssd; doesn't work a "cheaper" server-quality hd?

When I had just the US realms, I started the server with a 10k RPM HD, and ran into I/O wait delays. Since I'm updating every realm every hour, and pruning old data about as often, I need very fast access and write times to keep up with the updates. Here's the devblog from when the SSD was purchased for more info.

[Xsinthis]

Can't you just, err, add on a cheaper hard drive for the new non-English realms?

[Erorus]

Can't you just, err, add on a cheaper hard drive for the new non-English realms?

Err, no?

When I had just the US realms, I started the server with a 10k RPM HD, and ran into I/O wait delays. Since I'm updating every realm every hour, and pruning old data about as often, I need very fast access and write times to keep up with the updates.

[Xsinthis]

Well how many realms are we talking about adding? Can't be nearly as many as there are US realms