WoW Community API Updates - Patch 7.3.5

API Discussion
Mok'ra community API developers!

We wanted to take a moment and provide an update on the current status concerning World of Warcraft Community API updates for patch 7.3.5.

At the moment, we have completed all scheduled updates to Community API related services, apart from the below listed services:

Web Auction House
Mobile Auction House
Community API Auction House Dumps

We are actively working on bringing the above services back online.

Thank you for your patience.

~~~~~~~~~~~~ Update 2018-02-14 ~~~~~~~~~~~~

Greetings!

It sounds like some of you are already aware, but to make it official; the Auction House API has been restored, along with the Mobile and Web Auction Houses.

In efforts to ensure that new issues are posted in new threads, and so that we can more closely stay on-top of any new issues as they arise, I will likely lock this thread sometime next week. So that the conversation can continue as needed, the thread will remain open until that time.
Thank you! Does Blizzard have a timeline or an ETA for the Community API Auction House Dumps?
Hey cuddlywookie,

At this time, we are not able to offer any timelines or estimates on when these services will be restored.
Please fix this issue because it is affecting game's economy a lot, i cannot work with tsm and other services based on the API and it gets really frustrating >.<

4 days now. *sigh*
01/21/2018 12:24 AMPosted by Luram
Please fix this issue because it is affecting game's economy a lot, i cannot work with tsm and other services based on the API and it gets really frustrating >.<

4 days now. *sigh*


This - I can understand how bringing things back up and running takes time, but 4 days seems a tad long.

Currently the only way i have of grabbing data on a new server is to manually search the entire AH - I don't like sitting around for an hour waiting for that, and i can only assume that with me and pressumably others doing this, AH queries slow down. Is a bad time :(

Hope you get it back up soon, thanks!
Yup as of 16th Of january -
the auction data hasn't been updated. Please keep us posted. :) tyvm
Last auc update was at Tue, 16 Jan 2018 14:12:42 GMT
We need dataaa
^^ We do. Please update us. Thanks!
Please, any news of an ETA?
An estimate is just a guess. You aren't even able to guess if it will be fixed in 2018? That's a bit scary.
01/22/2018 07:52 AMPosted by Azramon
An estimate is just a guess. You aren't even able to guess if it will be fixed in 2018? That's a bit scary.


When you say not able to provide an estimate (aka guess).. what definition of able are you meaning? I think you must be using the word wrong. I doubt the dev team lacks the ability based on the definitions below.

able
adjective
1. having the power, skill, means, or opportunity to do something.
2. having considerable skill, proficiency, or intelligence.
01/17/2018 12:57 PMPosted by Maguthul
Mok'ra community API developers!

We wanted to take a moment and provide an update on the current status concerning World of Warcraft Community API updates for patch 7.3.5.

At the moment, we have completed all scheduled updates to Community API related services, apart from the below listed services:

Web Auction House
Mobile Auction House
Community API Auction House Dumps

We are actively working on bringing the above services back online.

Thank you for your patience.

Still no updates?
01/22/2018 07:58 AMPosted by Azramon
01/22/2018 07:52 AMPosted by Azramon
An estimate is just a guess. You aren't even able to guess if it will be fixed in 2018? That's a bit scary.


When you say not able to provide an estimate (aka guess).. what definition of able are you meaning? I think you must be using the word wrong. I doubt the dev team lacks the ability based on the definitions below.

able
adjective
1. having the power, skill, means, or opportunity to do something.
2. having considerable skill, proficiency, or intelligence.


I am not sure why the down votes on my comments. I copied that meaning from the online dictionary. I didn't make it up. On my other comment, people seem to not want it fixed this year. I would like it fixed this year please. It's not fair if you don't want it fixed to have an impact on others :( They won't be taking away time from other projects for a long time to fix it if that is the concern.
There's no doubt the new scaling method involves a lot of changes; undoubtedly affecting items. At the very least, this involves new fields for items in the dB; perhaps making a final assessment and addressing the iLevel of items issue that's been a problem for quite some time. If they're choosing to address this once and for all, this requires a game plan for the upgrade path for the API, as I'm sure the Built-in auction is of primary consideration. So, I imagine that's a primary focus, and the code for it as well. Once that path and changes are stablized, work can begin on the external API access. This is just a logical analysis based on my experience with dB changes.

We know that making sure the majority of Players don't have issues with game play is a primary consideration. So, practically everything other than what was mentioned all of which revolve around the API.

Even when that is finished, there's no saying what that will entail for the client side of the API. It really depends on what their end goal is with the API, no doubt as a 'semi' supported action, the Mobile will have some priority, as it's 'technically' a Blizzard offering. The others, Addon related, and web related clients(sic) will need more extensive information on the chnges to implement them.

I don't recall how long the API was down last time, it seems at least two weeks. We're on week 2 as of today. So, the earlierst I'm expecting even a clue as to a "status" update is sometime before next week's update push, or garbage collection, maybe, Tuesday. As a developer, this is just part of development life. Taking one's time to ensure it works takes less time than cleaning up a mess on a failed dB update.

All this said, there's nothing worse than waiting for a response; apparently PR on this is lacking or even possibly a non-delegated responsibility. If patience is a virtue, we all need to be virtuous. I'm certain they are well aware of how important this is to all of us; although I really have no confirmation on that. ;-)

Last week is the most up to date response I've seen; or 5 days ago today, so I'm referring to the one here in these posts.
I'm actually impressed that this 'scaling' update has gone as smooth as it has. Only one 'major' issue that I've seen, and even that's not horrid, and I don't know if that's been handled; I think it was some dungeon or raid group issue being too hard.

So, patience is the byword, and with hope, Blizzard will deign to keep us updated on progress. Let's hope we hear from them soon.
How sad that the blizzard api is still down for the AH...manually scanning on a high pop server takes so long that tsm actually crashes o/
What's up with this lack of communication though? Normally I get to gush about how awesome blizzard support etc is that they make it seem like we're valued customers :D unlike some other game developing companies..
Sure, these are forums and not GM-tickets, but you'd think they'd let us know?
ps: Thanks StalePeeps for explaining a bit :) Being a blonde idiot I don't get much further than 'this is the on button' when its computer related ;) so all I'm thinking is "why is it taking this long? how hard can it be? what's going on? D; "
01/24/2018 06:05 AMPosted by Domenica
"why is it taking this long? how hard can it be? what's going on? D; "

First I'd like to start, and end with:
Kudos to the development team and operators, with much appreciation. Please update us on progress, your time is truly appreciated!

Thanks for asking, although perhaps, for some, my explanation might be a bit tech oriented, I'll try to simplify my response and not get 'too' wordy.

So, you ask 3 questions, let me address "how hard can it be?" first. When it comes to difficulty, this has a moderate difficulty. It doesn't require anything that hasn't been done before; and with the exception of incorporating new things (iLevel, scaling), that just brings it from easy, to moderate. When people outside the development community use these 'new' things, we don't have to create all the references and calls to use or view the items (they're prepackaged tools); that's what will be created/is being created. Adding new objects and the methods to handle them is the most difficult as you have to ensure that the additions adhere to your naming conventions, no duplication and coding is perfect or perfected.

With so many objects, having a convention enables 'everyone' to find that name you gave it, so it can be used; creating new items means you have to strictly adhere to that naming convention, with no room for error in this process.

So, a low to moderate difficulty, just routine technical requirements that are tedious, and time consuming; machine are literal, a typo is a no-no and no different to a machine than a complete miss on doing it right, and of course the machine is also clueless to what went wrong; but, may complain and create an error! But, this is all routine, just time consuming, new objects and methods are added all the time.

2nd Question: Why is it taking this long. Well, I answered most of that in the first question; but, creating the references to the new parameters of the object, say iLevel, and other pertinent information regarding the new 'scaling' method require one to use the items mentioned above in answer to the first question, and the method to say what the object does is created. In this case, most of these things already exist, we're just adding additional functionality to the object itself. That is the answer to the second question. The code has to be written, often by another team. The information on the new additions need to be documented, passed around to the team, and eventually made available as required, but only if there are no other pressing issues with a higher priority can work on other problems begin.

This requires a bit of explanation: This process is strictly regulated with versioning, meaning keeping track of where everyone is in the development process and ensuring everyone is on the same page while developing. So, these are the development of the methods that will be used on the new objects, otherwise known a coding. And modifying the parts of the game that will use it! That last part being the point to everything.

The new 'scaling' is a critical component, and already done with the release. All this had to be done first; and testing done by publishing and collecting issues/tickets... and comments ;-) like this and other comments of people having problems. So, in a way, most of the database exists for the functionality in the game for individuals, the game itself and such; but, bug reports will often point to a situation where the object itself has something missing; or the method to make it do what it does may not address the mathematics or logic of this behavior correctly. This too can affect what and how the data is stored.

But, this then gets to the bottom line, the API database, and just leaves the collection process. The first priority to getting the API to do it's job, requires that all of the above is doing everything it needs to do before we can have the confidence needed to ensure that the static part of the data is perfect (zero tolerance for error), and now the database needs to work in and with the Auction House for each of these new objects and their methods, including how and where the Auction House stores and accesses this information for updating. This can be ongoing, and finalizing can take the time needed after major changes have been added required for functionality in the game.

Again, computers are literal, and the zero tolerance for errors conditions apply to every step along the path to completion, especially if you want to reduce bugs. Internal error collection and reporting is done behind the scenes giving feedback to developers about emerging issues. One can come close to this standard of perfection; but, being human, is never perfect. This zero-tolerance goal is the ultimate goal, often never ever reached, if ever. ;-) The database structure must be functional, stable, and everything that creates and relies on that information must be working. This gives us the confidence we can proceed to what ultimately is a static version, over time, for all that information, which we know is correct, or stable enough to use! Knowing that is also part of the process, which I'll not go into.

So, that's why it's taking time. The question is related to the number of changes made in the already pushed update to the game. Gathering the data about errors, and incidents are the last act, after everything is "working". The time required for this is a percentage of the the time used to create everything AND ensuring it all works. Determining what is close enough is part of this.

That's the phase we're in now. Troubleshooting the game. The good news is that no additional updates were pushed out during the down time. Or, it could mean that any updates haven't been solidified yet; and they'll come out. We expect minor revisions. Work on the API database will begin in earnest and updating and creation ongoing. Updating the API and database, ensuring it too makes available all the changes in the database. Nevertheless, it can't be released until everything else is working within expected limits. We're already using, in game, the first commit of the data, rigorously being tested in situ, with hope, no bugs or fixes needed. BTW, again that never happens, we're human.

Summary, there is a lot to do. This is why it takes so long. Things have to be tested in stages... Release, stage one is complete, and Testing in progress. Any updates will probably be pushed next week; hopefully no longer, and it's possible the database will be stable, and a first release of the API will come concurrently, or shortly after. If the API update is not available with the next commit and release on the fixes, and no more critical issues appear, then the API, a work in progress will soon reach commit stage, and be released.

I hope all this at least clears up a little of what's going on. I will, and would expect that a lack of response from the team is due to how busy everyone is on the project.

As a programmer/developer, no one other than one who does or has done this kind of work can appreciate how much deep thinking goes on in a project like this. Interruptions break concentration and extend the time required to do the work needed. No one, other than the development team leaders can respond to where we're at, and then whoever is overseeing the progress can report on it, and they're very busy overseeing everything, and making sure deadlines are met, and that every team stays in sync with other teams, and it's all coming together. Each team leader is busy answering questions overcoming problem areas, and the expectation is that there will be no issues when they're ready to commit. The end point at this stage is a moving target, getting closer, and moving away depending upon problem areas.

As you can see it's a complicated process. With the #1 rules, computers have a zero tolerance for errors! I can't over state that! In fact, to state that trapping errors, creating the error reports that users never see is one of the largest parts of any project. Adding this is fairly straightforward, and doesn't require much thought initially; it's the reaction to the error reports where the work on perfection gets started, along with user feedback.

Yes, I'm a bit wordy, and for that I apologize; I hoped to make myself as clear and positive as possible and some things bear repeating, like the zero tolerance for errors, and I'll add trapping them hopefully without the need for user intervention. ;-) The very first concept that coders learn is garbage in/garbage out. One misstep can create a never ending loop that freezes the system; takes too much background time causing lag; and/or and this is important, fails to eliminate what is collectively referred to as garbage, data that is no longer needed, but still exists in memory or is written out, to be handled. Doing all this is often as much a process of destruction, as it is creation; lest the whole system collapse under it's own weight! Never fear it's what they do!

Even with all these processes, the weekly down time is used for minor fixes, and clearing the system of garbage collected over time. When this happens it shows up for the user in quests failing to complete. NPCs and mobs failing to respond perperly, items going missing or misplaced not reaching the final destination, etc. etc. Garbage can be things like items destroyed, not getting completely destroyed. Duplicate activities due to timing and synchronizing, failing to be destroyed and horror of horros timing out and left unhandled. Code must be made to handle all this. The economy books not balancing on both sides because garbage collection fails, or the calculations are wrong, ultimatly because somting is missing.

Clearing garbage from dropped connections (DC's), queries getting dropped for various reasons. Transactions getting timed out, stalling and failing to respond all of which compound and affect attempts to ensure a static record (the API database) is created properly. Loads have to be balanced between servers, the list seems endless at times.

Kudos to the development team, with much appreciation. Please update us on progress, your time is truly appreciated!
Updates?
This would be a much different story if what StalePeeps said came from Blizzard. That is what most software companies would provide (at least after the fact). The problem here is that there has been zero communication from Blizzard (other than acknowledging that things are down), and despite this being a recurring issue, they have never given any insight into why things go down in the first place, or what they are doing to improve moving forward. So, I'd be very hesitant to try to guess at what the underlying issue is, or how long it's going to take to fix, considering Blizzard has never provided any information of this sort to go off of in the past.

Keep in mind that the level of communication we're dealing with here is one which has still not provided a reason or even real acknowledgement for an entire region worth of APIs randomly disappearing one day and never coming back.
Still awaiting a response from Blizzard. Thanks.
It'd be nice to find out even a rough estimate of when the APIs will be back up.

Join the Conversation

Return to Forum