FAO Roaken

[Killswitch]

Does anyone here (other than me) have any experience at all of localising (or internationalising) large applications? I worked for 14 months as a localisation (L10N) engineer for Sun Microsystems and worked on Solaris 9 and 10, OpenOffice and Java...

Anyone who thinks that taking essentially a non-internationalised codebase (like Mythic's DAOC) and producing it in multiple languages is an easy job is just dumb, frankly. The whole point about "Dyvet is English so it should be easy to move data to the US" is also dumb.

Mythic's DAOC codebase will be (probably) based on ASCII (1 byte per character) characters and strings. It will probably use fixed-length strings for names and other text. It will be highly optimised and probably have hard-coded strings compiled directly into it. The database structures it uses will again probably be using ASCII data and codepages.

To convert what is a very large, very complex and probably custom database to use UNICODE, to change the size of the fields, the indexing and everything else and then to make sure that this data can be moved without corruption, deletion, identical records, truncation would be a massive job. To make sure that the data sorts correctly, is indexed properly by Mythic and would appear correctly in-game, would also be a huge job.

The codebase that GOA work with is probably very, very highly modified...a lot more than most people seem to think. In order to allow translation to multiple languages, they will probably have had to redesign the databases and change lots and lots of the code in-game to use language-packs instead of hardcoded data. Both GOA and Mythic have said that the original code isn't properly internationalised, so I think it's incredible that FR/GE/ES versions of DAOC were EVER released and supported and that the patch cycles are so close to the US ones!!!

I say give 'em a break...I reckon that Mythic screwed up when they developed DAOC and that the amount of work, the difficulties and the success of DAOC in EU languages has been astounding. I'm just hoping that Mythic have learned from their mistakes (and WoW's success) and put some thought into WHO for their non-english-speaking clients.

I know that it could be done, but I think a lot of people (even our database experts here) might be underestimating the sheer amount of work there would be to do.

 

[Killswitch]

The Japanese server was run by Mythic already before the cluster wasn't it?

Yup...and does anyone know if it supported Kanji or Hiragana or Katakana or whether it was just done in Latin Alphabet characters (romaji?). If they didn't use proper Japanese glyphs, then there would be no localisation problems and therefore no cluster/transfer problems.

 

[Gahn]

I reckon if they invested 12 months to recover the Pryd Crash, they can manually type all the fucking records of 1 toon for let's say 30 euros don't they?
The only real chance is the char move to Usa, wake the fuck up tbh if u don't wanna lose even more people to Usa when War will be released.

 

[Gear]

Good stuff

OK, I understand that it might be a major task to transfer the data, but would you say that it is impossible or just that it falls under the "needs a lot of work to be done and we're not up for the task" category?

 

[Killswitch]

OK, I understand that it might be a major task to transfer the data, but would you say that it is impossible or just that it falls under the "needs a lot of work to be done and we're not up for the task" category?

It would need a lot of work, which they might not have the in-house skills to perform. It would give massive headaches to Mythic and be a risk to their data integrity. It would be prohibitively expensive and the work involved might well cost a lot more than 30 euros per character in terms of time spent, testing, integration.

Possible yes. It might be cheaper to just open up the game, let anyone who wants to play for free and run it at a loss though!!

 

[Gear]

It would need a lot of work, which they might not have the in-house skills to perform. It would give massive headaches to Mythic and be a risk to their data integrity. It would be prohibitively expensive and the work involved might well cost a lot more than 30 euros per character in terms of time spent, testing, integration.

Possible yes. It might be cheaper to just open up the game, let anyone who wants to play for free and run it at a loss though!!

I would pay 30 euros to play my shade in a US cluster. Even if that meant that it would require a brand new US account where it would be transfered.

If they implement chartransfer in a case to case basis I can't see how they'd run in a loss by doing so. They could probably figure out the cost required for the job to be done, then add on a little something, come to an agreement with Mythic on how to handle some of the things and there it is!

I understand that it might be difficult, but then again I can see how lots of people would be up for taking one of their toons and enjoying it on a normal cluster.

Thanks for the input btw

 

[GReaper]

It should be in their interests to get the game to support multiple languages on a single server so that both WAR and DAoC can do it. This is assuming they're reusing existing code from DAoC for WAR, and assuming that improvements will filter back to DAoC.

Will it ever happen? Probably not.

This game might die before it should all because of the language difficulties, with an English speaking playerbase on EU servers stuck with their characters on servers they don't want to lose.

 

[Imgormiel]

I don't think anybody has mentioned the fact that Mythic managed to cluster the Japanese server with the rest of the Devon cluster.

GoA are just lieing to you lot, and its getting stupid now.

You missed reading one of my last posts then

 

[Azathrim]

Stuff

I agree, transfering the data can very well be a huge job. But it can be done.

What is NOT a huge job is keeping the costumers in the loop.

GoA constantly fails to keep the costumers informed. Ergo, GoA is incompetent.

 

[Himse]

The Japanese server was run by Mythic already before the cluster wasn't it?

Don't think so.

But it remains the fact, they still translated a different language server.

I can't imagine the databases would differ that much within the SAME game.

 

[Penlid]

Don't think so.

But it remains the fact, they still translated a different language server.

I can't imagine the databases would differ that much within the SAME game.

These japanese servers were hosted by Mythic before they were clustered, anything that GoA doesnt run, Mythic runs

 

[Eeben]

Does anyone here (other than me) have any experience at all of localising (or internationalising) large applications? I worked for 14 months as a localisation (L10N) engineer for Sun Microsystems and worked on Solaris 9 and 10, OpenOffice and Java...

Anyone who thinks that taking essentially a non-internationalised codebase (like Mythic's DAOC) and producing it in multiple languages is an easy job is just dumb, frankly. The whole point about "Dyvet is English so it should be easy to move data to the US" is also dumb.

Mythic's DAOC codebase will be (probably) based on ASCII (1 byte per character) characters and strings. It will probably use fixed-length strings for names and other text. It will be highly optimised and probably have hard-coded strings compiled directly into it. The database structures it uses will again probably be using ASCII data and codepages.

To convert what is a very large, very complex and probably custom database to use UNICODE, to change the size of the fields, the indexing and everything else and then to make sure that this data can be moved without corruption, deletion, identical records, truncation would be a massive job. To make sure that the data sorts correctly, is indexed properly by Mythic and would appear correctly in-game, would also be a huge job.

The codebase that GOA work with is probably very, very highly modified...a lot more than most people seem to think. In order to allow translation to multiple languages, they will probably have had to redesign the databases and change lots and lots of the code in-game to use language-packs instead of hardcoded data. Both GOA and Mythic have said that the original code isn't properly internationalised, so I think it's incredible that FR/GE/ES versions of DAOC were EVER released and supported and that the patch cycles are so close to the US ones!!!

I say give 'em a break...I reckon that Mythic screwed up when they developed DAOC and that the amount of work, the difficulties and the success of DAOC in EU languages has been astounding. I'm just hoping that Mythic have learned from their mistakes (and WoW's success) and put some thought into WHO for their non-english-speaking clients.

I know that it could be done, but I think a lot of people (even our database experts here) might be underestimating the sheer amount of work there would be to do.

tbh i agree but they had to know that this day would come sooner or later, them not thinking about doing anything about it before it was tolate is just fucking stupid , they could see when wow was released there was gonna be a problem one day .
An not giveing any real information about this problem is just bad managment imo .

really dont see how anybody can stick up for goa anymore.

 

[Himse]

These japanese servers were hosted by Mythic before they were clustered, anything that GoA doesnt run, Mythic runs

BUT, the databases cant differ that much right?

 

[Sharkith]

I feel sorry for you guys I really do

I do believe GOA here tbh to be fair to them, its really sad it has taken this long and to have such a sad final point.

However, one solution might be to do the following:

Offer all active accounts on Dyvet a /level 50 on an existing server - say Avalon and give the players enough gold and BP's to fully ML and TOA their characters once. End of problem.

Would that not be possible and fair?

 

[preacherboy]

I feel sorry for you guys I really do

I do believe GOA here tbh to be fair to them, its really sad it has taken this long and to have such a sad final point.

However, one solution might be to do the following:

Offer all active accounts on Dyvet a /level 50 on an existing server - say Avalon and give the players enough gold and BP's to fully ML and TOA their characters once. End of problem.

Would that not be possible and fair?

How about the LGM crafters they had and there RR LVL?

 

[Sharkith]

it should be obvious that what your looking at is the end here. So something that involves a good head start in a new home ought to be considered good. Otherwise basically its time to close subs and move on altogether...

 

[Bluesky]

Offer all active accounts on Dyvet a /level 50 on an existing server - say Avalon and give the players enough gold and BP's to fully ML and TOA their characters once. End of problem.

Would that not be possible and fair?

Not really no, but it depends.... what happens to my 18 mil rp's on my SM, my 4 lgm capped crafters, the hundreds of quests I've already done, several mithrils of equipment/items, my houses (and trophies), etc, etc the list goes on and on.

Additional clustering is the only solution that I think most ppl would be happy with. I suspect the only way to do this is to create a new server then COPYING all ppl's (those that actually want to move and there could be a charge for the service etc) chars/belongings/etc to it and then adding it to an existing cluster either EU or US depending on license/contract etc. Ofc there will be problems with this which would need planning for etc but none which i think would be impossible to sort out.

I very much doubt this will happen tho purely because GOA dont seem to care in the slightest what happens to Dyvet.

 

[Scrinthe]

Offer all active accounts on Dyvet a /level 50 on an existing server - say Avalon and give the players enough gold and BP's to fully ML and TOA their characters once. End of problem.

Would that not be possible and fair?

I was an active player ...without missing more than a weeks subs once in a while ...My MAIN PC has blew up ...going to cost £500 - £600 to get it up again ( upgrades rather than try retain the specs )

Do I ( and people in simular possition ) not deserve things a Active account does ....considering I am going to re-sub once I find the cost to repair

( will not be any time soon btw )

 

[Manisch Depressiv]

Does anyone here (other than me) have any experience at all of localising (or internationalising) large applications? I worked for 14 months as a localisation (L10N) engineer for Sun Microsystems and worked on Solaris 9 and 10, OpenOffice and Java...

Anyone who thinks that taking essentially a non-internationalised codebase (like Mythic's DAOC) and producing it in multiple languages is an easy job is just dumb, frankly. The whole point about "Dyvet is English so it should be easy to move data to the US" is also dumb.

Mythic's DAOC codebase will be (probably) based on ASCII (1 byte per character) characters and strings. It will probably use fixed-length strings for names and other text. It will be highly optimised and probably have hard-coded strings compiled directly into it. The database structures it uses will again probably be using ASCII data and codepages.

To convert what is a very large, very complex and probably custom database to use UNICODE, to change the size of the fields, the indexing and everything else and then to make sure that this data can be moved without corruption, deletion, identical records, truncation would be a massive job. To make sure that the data sorts correctly, is indexed properly by Mythic and would appear correctly in-game, would also be a huge job.

The codebase that GOA work with is probably very, very highly modified...a lot more than most people seem to think. In order to allow translation to multiple languages, they will probably have had to redesign the databases and change lots and lots of the code in-game to use language-packs instead of hardcoded data. Both GOA and Mythic have said that the original code isn't properly internationalised, so I think it's incredible that FR/GE/ES versions of DAOC were EVER released and supported and that the patch cycles are so close to the US ones!!!

I say give 'em a break...I reckon that Mythic screwed up when they developed DAOC and that the amount of work, the difficulties and the success of DAOC in EU languages has been astounding. I'm just hoping that Mythic have learned from their mistakes (and WoW's success) and put some thought into WHO for their non-english-speaking clients.

I know that it could be done, but I think a lot of people (even our database experts here) might be underestimating the sheer amount of work there would be to do.

Yeah, I do have experience in programming large applications that switch language on the fly. That is also from the pre DAoC era. Maybe living in a country other than Dumbfuckistan, knowing that other languages than American English exist and native support is a must for the customers made me get the experience.

It's almost 2008, there is no solution to the language problems with DAoC, the patches for GoA servers are way too late in my opinion, and we're only dumb for sticking with GoA.

Even if someone stil really has hard coded strings and he hasn't be shot yet for it, it's EASY to "move" them out for further processing like translation and have a meta data system while leaving the original language strings in the code. You can use a system like gettext (gettext - Wikipedia, the free encyclopedia) and still stay ASCII. With a script it's a day's work, your "change lots and lots of the code in-game to use language-packs instead of hardcoded data" is actually no work at all. Even if the text is/was originally hard coded in the Mythic client, I doubt that GoA has several code bases (one for each language) and that the translators are allowed to look at it. They will most likely get auto-generated ASCII text files. If there are really some slaves at GoA extracting the text manually and putting it back in to compile a specific language client then I can't help it.

Take a look at the translated texts or GoA's websites and you'll see that amateurs are at work, I doubt that they are able to highly customize a MMORPG code base.

I don't know why it should be a massive job to move a database from ASCII to Unicode, even if the text-fields are fixed length, it's a job of a few hours.

But maybe Mythic isn't "programming" at all, maybe they just create content against the Gamebryo engine and don't want to update it for DAoC. I assume today's Gambryo engine and the server module has no problems multiple languages.

Offered to solve the language problem on my own, pay me a trip to Mythic/GoA and give me a month time and I'll do it. Just don't call anyone dumb, software engineering makes a lot of possible and correcting errors that have been made, even desing errors, should be top priority (principle of a broken window). It's possible, it has been done before many times.

As long as the Dyvet players are happily rerolling on GoA's servers or moving to US, nothing will change.

Mythic has a new product coming, they simply do not give a shit about the language problem, it has nothing to do with the sheer amout of work needed to solve the problem.

 

[Gear]

Stuff

Convince GOA to give you access to the database and arrange a move over to the US servers and I'm contributing 100 euro to the "Manisch's month in Paris sorting out GOA's shite" project.

If we get a confirmation from a GM in some sort of official way, open up a paypall account and I'll send you the money as soon as I get the details.

 

[Ctuchik]

and people aren't happy then WTF are you doing on an English server?

no americans on a english server?

 

[Imgormiel]

no americans on a english server?

I wasn't directing it in that way, ofc there are some Canadian's on an English server too, despite their main language being French ^_^

 

[Bugz]

Sorry to say but the GoA team COULD do a char copy from Dyvet to the other servers - if only to use GM powers to 'copy a replica of the dyvet' character. I refusr to believe that there is no way they can copy characters from one server to another.

They cant' use the 'that's not professional bullshit' as GoA are a bunch of unprofessional gits anyway.

Roaken's activity on this board shows he fails at GM and was most likely a new recruit who got shafted into playing in a position that his employee friends don't give two tosses about.

 

[Puri]

Run a poll, see how many would like to see their characters copied over to other servers (euro ones of course, you wouldn't want to lose paying customers by transfering to us, would you ?), if it comes back positive.. get your team together, start copying.. and yeah, for free.I find it hard to believe that something like charcopy is not feasible, defies all rules of logic.

This cluster was kept alive by the players solely up to some point, the same players who were left in the dark about the fate of the server and were not given any information until some topic sprang up that numbered 200 posts.

Or come straight with an official post on your webpage that says "nothing can be done", so we know what to do with our chars.And be fast for a change.

 

[Cromcruaich]

I'd be very interested in seeing the timesheets showing all those long long hours that those devs have put into investigating solutions for dyvet.

Decision is made, it was made a year ago. Solution is obviously possible, it's just not cost effective, and you know, the longer they left it, the lest cost effective it was going to be.

Cant be arsed reading people epeen fencing over how much technical knowledge and experience they have. Though I enjoyed manisch's initial post. But really, ive come to question people motives for continuing a detailed technical discussion.

Additionally, nothing lasts for ever - no point looking at it as 5 years of lost play. Just look back at the good times and the enjoyment that was of the moment - you weren't studying to be a medical doctor, or training to be a fighter pilot - mmorpg's are huge timesinks that will always look like a complete waste of time once theyve had their day - and believe me, they all will have their day.

NFD back trying the Limors anyway.

For people considering a move, I could selfishly tell them to come over to avalon and give it a crack - and actually mid/avalon is pretty healthy (practically as many people on the Limors Mid NFD/TT alliance as were on Duvet/Hib lastnight ) - but if I were doing it from scratch again, i'd personally go US - the language barrier is a great pain if you havent got friends etc to move with you. IMO ofc.

 

[Cromcruaich]

Roaken's activity on this board shows he fails at GM and was most likely a new recruit who got shafted into playing in a position that his employee friends don't give two tosses about.

That is totally unfair to say he fails as a GM. A football team doesnt win games with just a goalie!

 

[Manisch Depressiv]

Not planing to pay GoA any more, they easily have the money to put some effort into this problem. They don't want. 300 players don't make someone who is responsible (not Requiel or Roaken) care.

That's the true disaster. Once upon a time I worked at a computer store, sold a notebook to a lady who was working in coutries with humid climate, that notbook had to be sealed properly and so on so this took some time. After a week or so she called and asked to have a modem built in but that delayed the delivery from the middleman as he needed to reseal the notebook after putting the modem in. She complained about the long delivery and said that she is paying 10,000 DM (old German concurrency, ~ 5,1k Euros) and that she can expect some service for such a price.

I told her that she is getting the same (good) service/treatment as a customer who buys something for 1 DM. That silenced her.

Some time later my boss told me that she spoke to him about me and that she mentioned me/our shop in a good way, it impressed her that we not only care about the big orders/customers but treat everyone with the same respect.

If I would run GoA I would care about every single subscriber, but we're really not in the position to make them care. We either still play on Dyvet or on the German servers or moved to Mythic's servers.

What I suggest is: Register a domain and put up a web site and state Dyvet user stories which express what they think of GoA and their service in regard of Dyvet and that they are not going to play Warhammer on a GoA server and ask everyone to do the same. Then after the site is populated with some information (level-headed written entries not some "ololol GoA suxx" stuff) the word needs to be spreaded. One can post the domain to various forums, send it to friends, gaming sites/magazines and so on.

Maybe then someone will listen. I'll gladly pay the costs for the domain and the traffic. Just need a good domain name. Maybe somthing like www.goa-no-thanks.org?

We could be constructive and offer Mythic and GoA help from the community with programmers/data base experts if they are willing to pay the trips to the office where this would be made.

I know that every game has a lifetime, but realisticly looking a cluster could still somehow work with 1000-1500 players on it and to achieve that every cluster with same ruleset will have to be merged at some point. M. Jacobs said that DAoC is his baby and that he is not going to let it die, did he expect us to reroll from a dying cluster to the next one when he said it? In-game friendships break, people spread all over the servers, some do not make new friends on a new server and stop top play for good.

Anyway, what do you think of this idea?

 

[Cromcruaich]

Have a rep for a good suggestion, but my life is too short.

Write to you MP about beam trawlers and over exploitation of commercial salt water fish species instead!

 

[aika]

spread some of the whines on the official EU warhammer forums, burn down GoA offices!!!

 

[Sorin]

I belive that it is possible to cluster a cluster.
Even with different languages.

Its a game ffs,
you can write/code whatever you want. /jump, /morph, /path /code, /plvl
/gmrelicpad etc.

Everything with a game is possible.

It just takes time and money.

Goa got alot of both.

Instead of the ingame text/interface laying on the servers they should be on ones own pc. That would pretty much solve the problems.

And you might say : Oh Noes! that will take ages.

Well, we've been paying you people for ages.
Time to see some results if you ask me