UnknownFreak.2805

Crash is due to nullpointer :))--> Crash <--Exception: c0000005Memory at address 00000000`00000000 could not be read

Agree, please fix...

Same problem, EU here.

Qa pls...Every patch there is something...

Like, not to sound rude or anything, but really... second time... please change that validation, like.Try parse the build - it fails -> promt the user or use the in game mail system/ I do not know, that their build is no longer working and allow them to update it.Or make proper migration scripts that totally will not cause any issues what so ever.

Not the only one that gets triggered. Trust me, but anet busy pushing out those gemstore skins instead of fixing quality of life bugs.

I dont even get why the should happen crash in the first place... like any reasonable developer would make it so that, if it failed to validate, prompt the user the build(s) does not work and allow them to change/remove it...How it works today;Try save an invalid -> prompts it's not possible //OK;Try load an existing invalid build that (somehow) got saved into the template system -> Crash // NOK;two simple test cases and this could have been prevented...

How? Qa pls?

Yes, Chronojail.

Was confused at first, but prefer the smaller font. But font size should be change-able... but guess that's too hard.

Like this is unplayable when pugging people that leave all the time... how am I supposed to play the game...Right raids are dead content forgot about that... :joy:

That works, but why not use accept headers, where the version is specified?Maybe there is a reason not to do that... Edit: Should probably read before asking question with the answer in the post you linked

Nice, but how to actually use the codeand skills_by_palette in v2/professions?if I just get guardian https://api.guildwars2.com/v2/professions/Guardianno codefield nor skills_by_palette field in the data returned. (ctrl + f palette gives no results)Unless I'm missing something obvious...

Never had a problem decoding the link, it's the fact there is a pallete id and a different skill id, however I can now see that the wiki has been updated with those pallete id's but they are not available from an endpoint... bleh.I'd like something like the followinghttps://api.guildwars2.com/v2/skills?palleteId=id, to get the actual skill id(s) since there is the possibility that multiple results are returned... but that would have been too easy...Guess I have to fetch that wiki page now...

Cool, I knew I was missing something. And in order to unpack the profession spec, like the wiki states, the bits after the profession is 3 pairs of two bytes for each traitline. My example string DQcBHRc7KC4jD24BkxIAAIcSgwGAAYEBvAG8AQAAAAAAAAAAAAAAAAAAAAA=So for an example the following string of bytes (Not doing the entire unpacking here)Gives the following byte array(13, 7, 1, 29, 23, 59, 40, 46, 35, ..., 0) # the two bytes (1, 29) is the bytes describing what specialization and the trait choises made for that specialization.the 1 is pretty simple (binary 0b000 0001) - the specialization id (in this case defined in the api as mesmer duelling)the 29 is a little bit more tricky. It has the following binary value (0b0001 1101) Each pair of two bits (first unused (00)) (trait choise3 (01)) (trait choise2 (11)) (trait choise1 (01))The bits are extracted like the following pseudo code. bitfield extraction magic. To pack the extracted bytes back into one single byte, there is just some bitshifting required.return (choise3 << 4) + (choise2 << 2) + choise1 Hope this helps :)

Not really api related So I'm trying to decode a build template link, however I am stuck on parsing out the skill info and cannot understand how I should proceed...The link I am decoding is the following (Stripped the link start and end):Which is a chrono template with only the heal skills selected (It was generated from gw2skills)DQcAAAAAAAAjD24BAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA=Which gives the following bytes13 7 0 0 0 0 0 0 35 15 110 1 0 0 . . . 0The interesting bytes are 35 15 110 1, which correspond to the terrestial heal followed by the underwater, which is when I generated the link, Signet of ether (skill ID from Api is 21750) but no matter how I reorder those two bytes I am unable to get that number. 35 = (bin 100011), 15 (bin 1111), and 21750 (bin 101010011110110)So what am I doing wrong?I have read on the wiki and this confuses me as well:After that, there are 20 bytes of skill information. The IDs here are provided in little-endian format and refer to internal skill palette IDs. In order, each 16-bit value represents the following skills: (https://wiki.guildwars2.com/wiki/Chat_link_format)It mentions internal skill palette ID:s, are those different from the ID I get from the api?, and if so, how do I convert the internal skill palette id into the actual skill ID?Or do I have to manually map these somehow? Using the following simple python code:s = base64.decodebytes(b'DQcAAAAAAAAjD24BAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA=')struct.unpack('BBBBBBBB', s[0:8]), struct.unpack('<HHHHHHHHHHHHHHHHHH', s[8:])doc for python struct the '<' character says to unpack using little-endian. and H stands for unsigned short, which is last time I checked 16 bitsI get the following when I run0 = {tuple} <class 'tuple'>: (13, 7, 0, 0, 0, 0, 0, 0) # Template id, profession id, empty trait lines1 = {tuple} <class 'tuple'>: (3875, 366, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0), terrestial heal, aquatic heal, ... acording to wiki...Am I missing something obvious? 3875 != 21750

It's obv not game breaking, but like I said, makes the game look unpolished. Probably low prio anyways, because there are other (more) important bugs to fix.Still, the bugfix shouldnt take more than 2 seconds to implement... (Assuming* proper design here.) Where the only thing in need of change (should) be the icon name string.... But if it was that easy to fix.... It should have been done by now. *(Assumption is the mother of all fuckups).

Like seriously... makes the game look so unpolished.https://cdn.discordapp.com/attachments/107112447755726848/623588323574939687/unknown.png How wasnt this discovered before deployment?

If only this was qa tested...

So... you phase qadim at like 60% and he decides to do his auto attack just before you get lifted up.Now have fun getting bombarded with no blocks whatsoever... Please fix?

Apparently feasts cannot be placed inside raid instances... so the list of problems is clearly growing.The lack of QA is pretty sad... it would have taken like 2 seconds to actually go in an raid instance and verify they can be placed there as well. The lack of finding these kinds of issues before deploying is killing the game for me. They should never happen in the first place... Apparently feasts cannot be placed inside raid instances... so the list of problems is clearly growing. That's crazy! :o I hope it will be addressed soon, along with the lack of individual ascended food (which should be added as well, at a lower cost than feasts). Yeah, even if it has the same ascended material cost as feasts but produces, say, 5 servings that would be OK. As it is I don't see how ascended food being feasts is really all that useful to small groups. I mean, in a raid group you have tanks, healers, power dps and condi dps and all of those roles would use different consumables. Is there any real benefit to putting down a feast if it's only of use to a few of the 10 man group? Totally agree on this, and if you are pugging people, they will leave to left and right. So the lack of a non feast version is pretty dumb.This feature is not properly implemented in my eyes. And as a main tank in our guild raids, feasts that only I will (and the off tank if it's required) use feels like a huge waste to me. (Not to mention I cannot benefit from them anyways, because they cannot be used in raid instances in the first place)Concept - Okay.Implementation - Awful

Thanks for ruining my old food items that lasted 1h.... Now they feel pretty underwhelming... not to mention the economy crash on those items...Really happy about this one.

Now we only need proper QA testing... ;)

Also happens on CA... Is there even a QA(Quality assurance) team anymore?This shouldnt have happen in the first place.