The new site seems to suffer from the current idea that there needs to be lots of color and pictures and shiny, when it ends up being a lot of distraction and harder see what you are actually looking for and to navigate.
Part of my job is helping companies assess the impact of site updates.
They never learn that these types of changes are more detrimental than helpful.
Never.
I pestered them about the usability issues already on Twitter, and was invited to post something on their support site.
I'm staring at it right now.
When you pull up an individual card, you can click on the set logo and it will show you all cards from that set.
Alternately, "Show Advanced Search" (whenever you see it) will turn up the original Puca's Mischief card face with all the searchable boxes as per usual.
I should have been more specific.
I meant the search engine in "Send Cards" section. It was able to search with set abbreviations but now it can't, even full set names.
I do agree that the package finder thing is a very good addition.
Package trader is great IF you actually keep an inventory on Puca, which I never did. And even then, it still shows you packages of ten cards even if people only have points for one. Their one good feature is still somewhat annoying :/
I've been on Puca since late 2013 and would like to share a few comments I already posted to the subreddit.
The site doesn't have the information density or efficiency the old one had. But once I turned on the Lightweight mode, I was able to use it without any blocking issues.
The package trading is efficient and has gotten me more trades already. The condition support and language support has facilitated lots of trades. The MTGO support, while still experiencing some issues, is a major step forward in trading tech.
Yes, the UI needs improvement to clear some bugs and meet the standards for power traders. But the site is usable, more advanced, and a better platform than it was before the change.
The best measure for me is how many cards I can send and receive. Since the new site overhaul, I've been sent 4 Japanese fetchlands and have sent out two packages. I'm receiving more and sending more, so its an improvement in the most basic measure.
Is there a way to see how often a card is traded in the new interface? A card's page used to have a line "This card is traded 14.2 times per day", "this card is traded 1.7 times per month". I found that useful ... and now I find it missing.
It would appear the site adding lower condition has netted me one card I've had on my list a while - someone has staked a claim to send me a SP Phyrexian Dreadnought (a kinda key card for the Varolz, the Scar-Striped EDH I have almost built). I've had it listed on my wants for quite a while, but as one of what appears to be 172 want'ers, it was... unlikely, to say the least. Well, the majority of want'ers haven't added tolerances for lower conditions yet, so I guess someone who had an SP one on the shelf they thought they couldn't move on the old site (my profile said I'd happily accept SP but when you're in a sea of want'ers you can't expect your profile to get read) finally saw an opportunity to move it.
I do agree that the package finder thing is a very good addition.
Package trader is great IF you actually keep an inventory on Puca, which I never did. And even then, it still shows you packages of ten cards even if people only have points for one. Their one good feature is still somewhat annoying :/
True, but it's actually a reason TO keep a running inventory now. Before, taking the time to add every card I own to my "Have List" was a big ol' fat waste of time, because even if someone did want my 20 cent rare, that was likely all they wanted from me and it simply wasn't worth it to try and look through everyone's pages and try to put together a package worth sending. Now that that is done for me automatically, it becomes much more worthwhile to list lower-value cards on the chance I can easily package them to someone.
I wasn't aware of the package pricing problem (alliteration for the win!) so that's something to keep an eye on if it becomes a hassle, but that's probably something PucaTrade can fix very easily, such as graying out or minimizing results if the receiver doesn't have enough points for the whole package, or just adding a filter to only show people who can afford a whole package.
One thing I was curious about as someone who doesn't currently use Pucatrade but certainly keeps an eye on the goings on there via this thread mostly... The addition of mtgo trading and such, has me a tad curious, how does that work within the pucapoints system and what sorts of conformation exist for the online trading to help avoid the potential for scammers and such that something like package tracking and such for physical cards can help avoid?
Is it just me or the send package displays wrong figures as the total amount of point in a given package?
I see people wanting for 3000 worth of PP, then when I open the package, I realize they only have like 320 points worth of cards.
I'm having the exact same issue now. None of the packages at the top (I organize by pp) show up as the advertised pp amount when i click them, it's much less. They need to address this as it's making the package send (the only real benefit right now) worthless.
Private Mod Note
():
Rollback Post to RevisionRollBack
BUDGET DECK BUILDER SINCE 93'
Modern ------------------------- WU Azorius Titan Midrange
Commander ------------------- WUG Derevi, Empyrial Tactician
***Are you looking for cards priced below TCG NM "LOW" and with FREE SHIPPING? - Click HERE! for my sales thread - 485 Feedback 100% positive!***
Has anyone else found that certain cards in their wantlist show a point value of 0?
Mine was doing that a couple days ago, but then inexplicably fixed itself 6-12 hours after I first noticed. I was never certain if prospective traders were seeing my wants as 0 either. I'd recommend just putting yourself on "vacation" mode for a few hours, and check back in at the site later tonight or tomorrow morning.
One thing I was curious about as someone who doesn't currently use Pucatrade but certainly keeps an eye on the goings on there via this thread mostly... The addition of mtgo trading and such, has me a tad curious, how does that work within the pucapoints system and what sorts of conformation exist for the online trading to help avoid the potential for scammers and such that something like package tracking and such for physical cards can help avoid?
They have bots set up. You trade your cards to the bot and then I presume get your points even if its a couple days before they pickup.
Private Mod Note
():
Rollback Post to RevisionRollBack
Out of the blackness and stench of the engulfing swamp emerged a shimmering figure. Only the splattered armor and ichor-stained sword hinted at the unfathomable evil the knight had just laid waste.
One thing I was curious about as someone who doesn't currently use Pucatrade but certainly keeps an eye on the goings on there via this thread mostly... The addition of mtgo trading and such, has me a tad curious, how does that work within the pucapoints system and what sorts of conformation exist for the online trading to help avoid the potential for scammers and such that something like package tracking and such for physical cards can help avoid?
They have bots set up. You trade your cards to the bot and then I presume get your points even if its a couple days before they pickup.
Ohhh, okay, that makes 100% sense then. That's awesome of them to be willing to act as intermediaries between mtgo traders to allow for a free flow of mtgo cards that can be converted into pucapoints and then into physical cards. Sounds like a win win for a lot of people (or for physical cards into points and then into mtgo cards).
One thing I was curious about as someone who doesn't currently use Pucatrade but certainly keeps an eye on the goings on there via this thread mostly... The addition of mtgo trading and such, has me a tad curious, how does that work within the pucapoints system and what sorts of conformation exist for the online trading to help avoid the potential for scammers and such that something like package tracking and such for physical cards can help avoid?
They have bots set up. You trade your cards to the bot and then I presume get your points even if its a couple days before they pickup.
Ohhh, okay, that makes 100% sense then. That's awesome of them to be willing to act as intermediaries between mtgo traders to allow for a free flow of mtgo cards that can be converted into pucapoints and then into physical cards. Sounds like a win win for a lot of people (or for physical cards into points and then into mtgo cards).
I'm testing the waters for this right now. i had an open order for tix for about a week, but no bites since I don't pay bonus. I just uploaded a bunch of commons and uncommons from standard and modern so we'll see how that goes.
Private Mod Note
():
Rollback Post to RevisionRollBack
Out of the blackness and stench of the engulfing swamp emerged a shimmering figure. Only the splattered armor and ichor-stained sword hinted at the unfathomable evil the knight had just laid waste.
The new site seems to suffer from the current idea that there needs to be lots of color and pictures and shiny, when it ends up being a lot of distraction and harder see what you are actually looking for and to navigate.
Part of my job is helping companies assess the impact of site updates.
They never learn that these types of changes are more detrimental than helpful.
Never.
I pestered them about the usability issues already on Twitter, and was invited to post something on their support site.
I'm hoping they take me seriously and give me an opportunity to help out.
Agreed. Moreover, I'm aghast at how they thought a 48ish hour downtime to some users was an even remotely acceptable user experience. I mentioned this in their chat during the transition and received a really lack-luster response. I'm just not sure how a team of (what, 3-5?) developers could go about a transition like the one they went through without doing it safer. Not to mention the surplus of bugs that were a surprise when they launched and they forced users to use their bug ridden website for a week before most of them were fixed. Mind boggling. And yet, they don't see a single thing wrong.
Agreed. Moreover, I'm aghast at how they thought a 48ish hour downtime to some users was an even remotely acceptable user experience. I mentioned this in their chat during the transition and received a really lack-luster response. I'm just not sure how a team of (what, 3-5?) developers could go about a transition like the one they went through without doing it safer. Not to mention the surplus of bugs that were a surprise when they launched and they forced users to use their bug ridden website for a week before most of them were fixed. Mind boggling. And yet, they don't see a single thing wrong.
Yeah, given my professional experience, had I ever told an executive at a company I worked for that a major change would require 48 hours of downtime, I'd have been lined up against the wall with the rest of the engineering team.
Your experience in the chat room is really disheartening, and leads me to believe that there may be a bit of Richard Hendricks style hubris at play here. If your users are taking the time to tell you something is wrong, that means you have users who care and want to use your site. If you meet their complaints and concerns with indifference and indignation you're an idiot, because the people giving you money are spending their time (money) to tell you how to make the experience better, which will make you more money.
Agreed. Moreover, I'm aghast at how they thought a 48ish hour downtime to some users was an even remotely acceptable user experience. I mentioned this in their chat during the transition and received a really lack-luster response. I'm just not sure how a team of (what, 3-5?) developers could go about a transition like the one they went through without doing it safer. Not to mention the surplus of bugs that were a surprise when they launched and they forced users to use their bug ridden website for a week before most of them were fixed. Mind boggling. And yet, they don't see a single thing wrong.
In my earthly, non-forum life, I am a software test lead/manager, and deal with the testing and rollout of software on a daily basis. I've been chatting with the Puca admin guys a little during rollout. They've been quite free to admit to me that the 48hr gradual rollout was not ideal, and not what they wanted.
I suspect they're simply a team of devs who haven't had to worry too much about pre-release testing before now. And if they haven't got any test leads on their roster (which I believe they don't), they did a pretty admirable job for a bunch of guys who are new to this. I think they'd be pretty happy to admit now that they didn't properly load test the site before sending it live, and hadn't catered for the idea that every man and his dog would crush it as soon as it came up. And I know experienced test managers who wouldn't prioritize this (and no, that doesn't make them good experienced test managers).
I certainly don't envy the position they found themselves in - it would've been pretty much coding, pizza and as much caffiene as you can muster for as long as it takes, sod the idea of sleep. Feedback to users is important, but it's pretty hard to do it well while you're in the grinder like that. I'm hoping that once they iron out the last couple of particularly offensive bugs (especially the cards reporting their value as 0 in people's want lists and screwing up the package totals accordingly) that we'll start to see more open and honest communication from the lads like we have in the past.
Just be glad it didn't end up like some of the more classical screwed-up software rollouts.
Queensland Govt Health Payroll System Inquiry Proceedings (PDF)
TL;DR Queensland Govt wants new payroll system, IBM codes it, it's full of bugs, third-party guy does the acceptance testing and fails it, IBM fudges the test report and the Govt deploys it. Three months of payroll chaos ensues.
You know what, it's not even that there was a metric ****ton of bugs or that the site itself had a terrible UI. Those aren't the deal breakers for me because surprise surprise, there was the old site which was functional and was simplistic to use.
They have stubbornly refused to go back to the old site while working out the future site fixes and doing proper testing for the next hypothetical rollout. This is what I have a beef with because it says a lot about their decision making. They would rather subject us to a site that is riddled with bugs and is hard to navigate than going back to what works 100%. I don't know if it is laziness or hubris but if I had been in charge of the roll out, I would've 100% double backed after 48 hours even if migrating back took time.
Every time the pucatrade rep on discord says "the only way out is through" as a response to "why not go back?", I want to reach through the monitor and slap the ***** out of whoever thought that is an acceptable answer.
You know what, it's not even that there was a metric ****ton of bugs or that the site itself had a terrible UI. Those aren't the deal breakers for me because surprise surprise, there was the old site which was functional and was simplistic to use.
They have stubbornly refused to go back to the old site while working out the future site fixes and doing proper testing for the next hypothetical rollout. This is what I have a beef with because it says a lot about their decision making. They would rather subject us to a site that is riddled with bugs and is hard to navigate than going back to what works 100%. I don't know if it is laziness or hubris but if I had been in charge of the roll out, I would've 100% double backed after 48 hours even if migrating back took time.
Every time the pucatrade rep on discord says "the only way out is through" as a response to "why not go back?", I want to reach through the monitor and slap the ***** out of whoever thought that is an acceptable answer.
I'm not answering for them, because they'll have their own reasons and so forth, but the usual answer is "we'd lose the last 48hrs worth of data".
If they reverted, the old site would be lacking the last 48hrs worth of trades, and they'd have to find some way of either reverting them or importing them (which is as big of a pig to do as it sounds). Most companies will choose to keep going, simply because they lose time reverting (and that means lost deadlines, which means money). Obviously, Puca isn't going to have some high-flying corporate deadline underpinning its decisions to revert or not, but I fancy data integrity would've been high on their list of reasons.
You know what, it's not even that there was a metric ****ton of bugs or that the site itself had a terrible UI. Those aren't the deal breakers for me because surprise surprise, there was the old site which was functional and was simplistic to use.
They have stubbornly refused to go back to the old site while working out the future site fixes and doing proper testing for the next hypothetical rollout. This is what I have a beef with because it says a lot about their decision making. They would rather subject us to a site that is riddled with bugs and is hard to navigate than going back to what works 100%. I don't know if it is laziness or hubris but if I had been in charge of the roll out, I would've 100% double backed after 48 hours even if migrating back took time.
Every time the pucatrade rep on discord says "the only way out is through" as a response to "why not go back?", I want to reach through the monitor and slap the ***** out of whoever thought that is an acceptable answer.
I'm not answering for them, because they'll have their own reasons and so forth, but the usual answer is "we'd lose the last 48hrs worth of data".
If they reverted, the old site would be lacking the last 48hrs worth of trades, and they'd have to find some way of either reverting them or importing them (which is as big of a pig to do as it sounds). Most companies will choose to keep going, simply because they lose time reverting (and that means lost deadlines, which means money). Obviously, Puca isn't going to have some high-flying corporate deadline underpinning its decisions to revert or not, but I fancy data integrity would've been high on their list of reasons.
To second what Cadaverous is saying, in the engineering space it's often far easier to roll forward than to have a rollback plan. Often companies that spend the time on rollback plans have major data constraints that can't be violated, such as financial or medical data. In situations like that you need to have a rollback plan because a roll forward that creates significant downtime due to unforeseen bugs can cost enormous amounts of money or even lives.
I don't fault them for "the only way is through" mentality given their size and resources, but I do fault them for rolling forward with such a massive change without more due diligence (load testing, usability testing, etc).
I don't see how rolling back the UI could cause data inconsistencies between platforms, they're still saving to the same databases. Or if they're not, they should have rolled out the back end platform and done some sort of AB migration. All in all, they're talented people doing good work, but I feel as though they're lacking some sort of 'Senior' Engineer leadership to assist with more advanced things like the migration and rollout of an entirely new site. *shrug*
I don't fault them for "the only way is through" mentality given their size and resources, but I do fault them for rolling forward with such a massive change without more due diligence (load testing, usability testing, etc).
But they have to do something in the now to fix the problem. Trudging forward is not going well at all and the responses are only going to get worse as time goes on, not to mention people will just cancel subscription or stop using the site completely which affects their reputation and their bottom line.
Bite the bullet, swallow their pride, whatever. Sacrificing resource and time on a rollback is far better than getting a *****ty reputation, devalued currency, and potentially even going out of business.
I don't see how rolling back the UI could cause data inconsistencies between platforms, they're still saving to the same databases. Or if they're not, they should have rolled out the back end platform and done some sort of AB migration. All in all, they're talented people doing good work, but I feel as though they're lacking some sort of 'Senior' Engineer leadership to assist with more advanced things like the migration and rollout of an entirely new site. *shrug*
Given the amount of downtime, I'm going to go out on a limb and posit that they had a number of backwards incompatible changes. I wouldn't be surprised if they made a number of schema changes that were incompatible with the model that was previously used in the view, and then updated their view/stack (I don't know what technology they're using off the top of my head. PHP? Ruby?) to make use of the new schema.
This is judging on how long it took and how things got out of hand for them. Backwards compatible changes rolling forward usually don't cause the same kind of deployment dumpster fires.
I don't see how rolling back the UI could cause data inconsistencies between platforms, they're still saving to the same databases. Or if they're not, they should have rolled out the back end platform and done some sort of AB migration. All in all, they're talented people doing good work, but I feel as though they're lacking some sort of 'Senior' Engineer leadership to assist with more advanced things like the migration and rollout of an entirely new site. *shrug*
Given the amount of downtime, I'm going to go out on a limb and posit that they had a number of backwards incompatible changes. I wouldn't be surprised if they made a number of schema changes that were incompatible with the model that was previously used in the view, and then updated their view/stack (I don't know what technology they're using off the top of my head. PHP? Ruby?) to make use of the new schema.
This is judging on how long it took and how things got out of hand for them. Backwards compatible changes rolling forward usually don't cause the same kind of deployment dumpster fires.
Exactly what I was going to say.
Given the scope of the changes, they'd more than likely have an old database for the old site, and a new database for the new one.
If someone traded a non-English, less-than-NM card inside the first 48hrs, there's no way that they'd have been able to roll that back into the old database.
while I understand the importance and frustration of the roll out, I am more concerned with how I seemingly cannot get cards sent to me. Only 2 in the last 3 weeks. With 30,000 points siting I am done sending out cards for now. I have over 1000 trades in sent and received. Real question is the site slowing down in the amount of trades?
Private Mod Note
():
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
I pestered them about the usability issues already on Twitter, and was invited to post something on their support site.
https://pucatrade.uservoice.com/forums/223627-ideas/suggestions/15823765-serious-usability-issues-with-future-site
I'm hoping they take me seriously and give me an opportunity to help out.
Want to be a better Magic player? Read the rulings forum and check out the comprehensive rules!
I should have been more specific.
I meant the search engine in "Send Cards" section. It was able to search with set abbreviations but now it can't, even full set names.
Package trader is great IF you actually keep an inventory on Puca, which I never did. And even then, it still shows you packages of ten cards even if people only have points for one. Their one good feature is still somewhat annoying :/
GReki, the History of Kamigawa Legendfall
UGEdric, Spymaster of Trest Drawmaster of Trest | GBGlissa the Traitor A Touch of Death | WBTeysa, Orzhov Scion Spinning in Graves
UWIsperia, Supreme Judge A Riddles of Sphinxes | RG Mena and Denn, Wildborn Beware Falling Rocks | GWSigarda, Host of Hurons The Enchantress
WRGRith the Awakener Superfriendly Tokens
The site doesn't have the information density or efficiency the old one had. But once I turned on the Lightweight mode, I was able to use it without any blocking issues.
The package trading is efficient and has gotten me more trades already. The condition support and language support has facilitated lots of trades. The MTGO support, while still experiencing some issues, is a major step forward in trading tech.
Yes, the UI needs improvement to clear some bugs and meet the standards for power traders. But the site is usable, more advanced, and a better platform than it was before the change.
The best measure for me is how many cards I can send and receive. Since the new site overhaul, I've been sent 4 Japanese fetchlands and have sent out two packages. I'm receiving more and sending more, so its an improvement in the most basic measure.
Check out Odds//Ends - My articles on Quirky Cards and Oddball Builds
Long-time PucaTrade member and sometime author. Send me cards!
Currently playing Knight of the Reliquary - Retreat to Coralhelm Combo
It would appear the site adding lower condition has netted me one card I've had on my list a while - someone has staked a claim to send me a SP Phyrexian Dreadnought (a kinda key card for the Varolz, the Scar-Striped EDH I have almost built). I've had it listed on my wants for quite a while, but as one of what appears to be 172 want'ers, it was... unlikely, to say the least. Well, the majority of want'ers haven't added tolerances for lower conditions yet, so I guess someone who had an SP one on the shelf they thought they couldn't move on the old site (my profile said I'd happily accept SP but when you're in a sea of want'ers you can't expect your profile to get read) finally saw an opportunity to move it.
I wasn't aware of the package pricing problem (alliteration for the win!) so that's something to keep an eye on if it becomes a hassle, but that's probably something PucaTrade can fix very easily, such as graying out or minimizing results if the receiver doesn't have enough points for the whole package, or just adding a filter to only show people who can afford a whole package.
I see people wanting for 3000 worth of PP, then when I open the package, I realize they only have like 320 points worth of cards.
I'm having the exact same issue now. None of the packages at the top (I organize by pp) show up as the advertised pp amount when i click them, it's much less. They need to address this as it's making the package send (the only real benefit right now) worthless.
BUDGET DECK BUILDER SINCE 93'
Modern ------------------------- WU Azorius Titan Midrange
Commander ------------------- WUG Derevi, Empyrial Tactician
***Are you looking for cards priced below TCG NM "LOW" and with FREE SHIPPING? - Click HERE! for my sales thread - 485 Feedback 100% positive!***
Mine was doing that a couple days ago, but then inexplicably fixed itself 6-12 hours after I first noticed. I was never certain if prospective traders were seeing my wants as 0 either. I'd recommend just putting yourself on "vacation" mode for a few hours, and check back in at the site later tonight or tomorrow morning.
"Personally I love high-riak, low-reqars gambles. Life's best with a decent amount of riak. And f*** reqars."
They have bots set up. You trade your cards to the bot and then I presume get your points even if its a couple days before they pickup.
Ohhh, okay, that makes 100% sense then. That's awesome of them to be willing to act as intermediaries between mtgo traders to allow for a free flow of mtgo cards that can be converted into pucapoints and then into physical cards. Sounds like a win win for a lot of people (or for physical cards into points and then into mtgo cards).
I'm testing the waters for this right now. i had an open order for tix for about a week, but no bites since I don't pay bonus. I just uploaded a bunch of commons and uncommons from standard and modern so we'll see how that goes.
Agreed. Moreover, I'm aghast at how they thought a 48ish hour downtime to some users was an even remotely acceptable user experience. I mentioned this in their chat during the transition and received a really lack-luster response. I'm just not sure how a team of (what, 3-5?) developers could go about a transition like the one they went through without doing it safer. Not to mention the surplus of bugs that were a surprise when they launched and they forced users to use their bug ridden website for a week before most of them were fixed. Mind boggling. And yet, they don't see a single thing wrong.
Yeah, given my professional experience, had I ever told an executive at a company I worked for that a major change would require 48 hours of downtime, I'd have been lined up against the wall with the rest of the engineering team.
Your experience in the chat room is really disheartening, and leads me to believe that there may be a bit of Richard Hendricks style hubris at play here. If your users are taking the time to tell you something is wrong, that means you have users who care and want to use your site. If you meet their complaints and concerns with indifference and indignation you're an idiot, because the people giving you money are spending their time (money) to tell you how to make the experience better, which will make you more money.
At the risk of spamming, I'm going to include my support link for this again which is currently sitting at 18 votes in hope that it will actually instigate a response from them: https://pucatrade.uservoice.com/forums/223627-ideas/suggestions/15823765-serious-usability-issues-with-future-site
If you feel like Future Site is a bit of a dumpster fire with regard to usability, please upvote.
Want to be a better Magic player? Read the rulings forum and check out the comprehensive rules!
In my earthly, non-forum life, I am a software test lead/manager, and deal with the testing and rollout of software on a daily basis. I've been chatting with the Puca admin guys a little during rollout. They've been quite free to admit to me that the 48hr gradual rollout was not ideal, and not what they wanted.
I suspect they're simply a team of devs who haven't had to worry too much about pre-release testing before now. And if they haven't got any test leads on their roster (which I believe they don't), they did a pretty admirable job for a bunch of guys who are new to this. I think they'd be pretty happy to admit now that they didn't properly load test the site before sending it live, and hadn't catered for the idea that every man and his dog would crush it as soon as it came up. And I know experienced test managers who wouldn't prioritize this (and no, that doesn't make them good experienced test managers).
I certainly don't envy the position they found themselves in - it would've been pretty much coding, pizza and as much caffiene as you can muster for as long as it takes, sod the idea of sleep. Feedback to users is important, but it's pretty hard to do it well while you're in the grinder like that. I'm hoping that once they iron out the last couple of particularly offensive bugs (especially the cards reporting their value as 0 in people's want lists and screwing up the package totals accordingly) that we'll start to see more open and honest communication from the lads like we have in the past.
Just be glad it didn't end up like some of the more classical screwed-up software rollouts.
Queensland Govt Health Payroll System Inquiry Proceedings (PDF)
TL;DR Queensland Govt wants new payroll system, IBM codes it, it's full of bugs, third-party guy does the acceptance testing and fails it, IBM fudges the test report and the Govt deploys it. Three months of payroll chaos ensues.
My Stupidly Large Number of Current Decks
PucaTrade with me!
The Multiplayer Power Rankings
Cube: the Gittening (My Multiplayer Cube) - MTGS Cube List | @ CubeTutor
The N00b Cube (Peasant cube for new players) - MTGS Cube List | @ CubeTutor
They have stubbornly refused to go back to the old site while working out the future site fixes and doing proper testing for the next hypothetical rollout. This is what I have a beef with because it says a lot about their decision making. They would rather subject us to a site that is riddled with bugs and is hard to navigate than going back to what works 100%. I don't know if it is laziness or hubris but if I had been in charge of the roll out, I would've 100% double backed after 48 hours even if migrating back took time.
Every time the pucatrade rep on discord says "the only way out is through" as a response to "why not go back?", I want to reach through the monitor and slap the ***** out of whoever thought that is an acceptable answer.
I'm not answering for them, because they'll have their own reasons and so forth, but the usual answer is "we'd lose the last 48hrs worth of data".
If they reverted, the old site would be lacking the last 48hrs worth of trades, and they'd have to find some way of either reverting them or importing them (which is as big of a pig to do as it sounds). Most companies will choose to keep going, simply because they lose time reverting (and that means lost deadlines, which means money). Obviously, Puca isn't going to have some high-flying corporate deadline underpinning its decisions to revert or not, but I fancy data integrity would've been high on their list of reasons.
My Stupidly Large Number of Current Decks
PucaTrade with me!
The Multiplayer Power Rankings
Cube: the Gittening (My Multiplayer Cube) - MTGS Cube List | @ CubeTutor
The N00b Cube (Peasant cube for new players) - MTGS Cube List | @ CubeTutor
To second what Cadaverous is saying, in the engineering space it's often far easier to roll forward than to have a rollback plan. Often companies that spend the time on rollback plans have major data constraints that can't be violated, such as financial or medical data. In situations like that you need to have a rollback plan because a roll forward that creates significant downtime due to unforeseen bugs can cost enormous amounts of money or even lives.
I don't fault them for "the only way is through" mentality given their size and resources, but I do fault them for rolling forward with such a massive change without more due diligence (load testing, usability testing, etc).
Want to be a better Magic player? Read the rulings forum and check out the comprehensive rules!
But they have to do something in the now to fix the problem. Trudging forward is not going well at all and the responses are only going to get worse as time goes on, not to mention people will just cancel subscription or stop using the site completely which affects their reputation and their bottom line.
Bite the bullet, swallow their pride, whatever. Sacrificing resource and time on a rollback is far better than getting a *****ty reputation, devalued currency, and potentially even going out of business.
Given the amount of downtime, I'm going to go out on a limb and posit that they had a number of backwards incompatible changes. I wouldn't be surprised if they made a number of schema changes that were incompatible with the model that was previously used in the view, and then updated their view/stack (I don't know what technology they're using off the top of my head. PHP? Ruby?) to make use of the new schema.
This is judging on how long it took and how things got out of hand for them. Backwards compatible changes rolling forward usually don't cause the same kind of deployment dumpster fires.
Want to be a better Magic player? Read the rulings forum and check out the comprehensive rules!
Exactly what I was going to say.
Given the scope of the changes, they'd more than likely have an old database for the old site, and a new database for the new one.
If someone traded a non-English, less-than-NM card inside the first 48hrs, there's no way that they'd have been able to roll that back into the old database.
My Stupidly Large Number of Current Decks
PucaTrade with me!
The Multiplayer Power Rankings
Cube: the Gittening (My Multiplayer Cube) - MTGS Cube List | @ CubeTutor
The N00b Cube (Peasant cube for new players) - MTGS Cube List | @ CubeTutor