5 ms·

> I bet they also have a "community" support forum or Google Group full of helpless people talking to a virtual wall. This is Google support summed up perfectl

by mstade 9y ago

> I bet they also have a "community" support forum or Google Group full of helpless people talking to a virtual wall. This is Google support summed up perfectly in a single sentence. It really is appalling.

mgkimsal 9y ago
> It really is appalling Never quite understood what's "appalling" about it. They've never had strong support for any dev/cloud-oriented services. I know some folks who've had good experiences, but those were exceptions, not the rule. I'll say, I get the frustration with it, but it's why I've never considered using their cloud services for anything. I had a client who asked me to migrate all their stuff to google 'cloud compute' and their 'sql cloud' offering last year. It was a performance disaster; queries that were minor .1ms on localhost or in the same network were suddenly 18-20ms hitting google cloud. Spent a few days trying to dig in, and a few responses from various groups were "shouldn't be that slow" and "works for me" and that was about it. Gave up trying to get any real response from google (we may have got a 'looking in to it' response - I don't remember). Thing was, I knew going in it wasn't going to end well, I just wasn't sure exactly why (maybe that's just too cynical of a view, but I prefer 'realistic').
- samastur 9y ago
  Just as a restaurant can be appalling from when it was first opened and even if you don't visit it ever, other services can. Appalling doesn't depend on personal experience or lack of prior knowledge.
- elorant 9y ago
  It's appalling because you're a paying customer. I can understand free services not having support, but any service that's costing money should at the very least have a human being responding through email.
  - mgkimsal 9y ago
    But you KNOW going in - there's no support. You know this by a) hundreds and thousands of horror stories from people saying "I did not get any support from google". You can counter that with "here are people who did get support", but the existence of the thousands of support nightmare stories out there gives you enough evidence to know that this isn't really "supported" in the way you might expect. b) there are no phone numbers. https://firebase.google.com/support/ https://firebase.google.com/support/ - no phone numbers. That's, of course, not the sole criteria you should use to judge potential support, but... it's google. It's 2017. We have multiple years of knowing what Google's track record is re: support for paying customers. Outside of large adwords advertisers (where they make a lot of money), it's non-existent from a practical standpoint. The big problem here is that firebase was fine until google acquired them. I'm reminded of a time when I had a bad experience at a restaurant. Local place I went to with friends from work. We went... probably 4-5 times per year, and I loved the chicken sandwich they did. Last time I went, it was ... bad. Not what I'd expected at all. It was different. Server asks "how's everything?". I said I was disappointed that the chicken sandwich had changed - it was not as good as before. Now... I wasn't actually complaining as in "comp the meal", but I was not a satisfied diner. "Nothing's changed sir". "um... yeah - this is nowhere near as good as it used to be". "You're wrong, nothing's changed". We had a back and forth, and they went back and came out "the cooks say the recipe is exactly the same". This was getting a bit ridiculous. Manager came over, same exchange. He leaves, the comes back and says "they're right - the recipe is 100% the same. We have changed a few ingredients last month, but it's the same". Maybe that doesn't actually relate to the google service situation at all now, come to think of it, but I was reminded of that situation when I started typing it :) Oh yeah - nothing changed from OP's end, except how the new owner (google) decided to start monitoring and accounting for usage. Think of a "$9.99 all you can eat buffet" getting loyal diners, then changing owners, and the new owners giving some people a $95 bill at the end of the meal because they'd decided to also charge for any visits to the salad bar over 2. Most diners would never hit that limit, but a few - loyal diners for years - would have been (rightly) upset at the after-the-fact change.
    - mstade 9y ago
      > But you KNOW going in - there's no support. I know this but to say everyone does or should isn't exactly fair. As well, you assume I am holding the purse strings, that I'm in any way responsible or even have a say in procurement. This may be true in my own business (it is) but not necessarily elsewhere, like my clients (it more often than not isn't) and you'll just have to live with poor decisions made on your behalf. Also quite often, there's a sunk cost that makes switching that much harder. So no, it's not fair to say "Google has shitty support but that's ok because we all kind of know this." Paying customers can and should complain about this, loudly, shouting it from the roof tops. Ideally they vote with their wallets and just leave Google, but again that's just easier said than done in many cases. Some semblance of professionalism shouldn't be too much to ask from one of the world's biggest corporations, an ask coming from paying customers no less.
- mstade 9y ago
  I don't think it's fair you're being down voted, I get your point of view, so I gave you an upvote. But the thing is, this isn't some kid-in-a-basement free-but-ad-supported service where you-get-what-you-pay-for. Google's cloud offering isn't free (for any use beyond hello world anyway) and as a paying customer I'd expect: 1. Things to work 2. People to answer when things don't work 3. Things to start working again in a timely manner, with timely and concise updates along the way Points two and three carry a lot of nuance, but Google's support does not. It's a virtual wall that people cry to and get no response back from, except for the rare few occasions when the Google deities decide to grace forums or mailing lists with their presence. It's just not a serious offering. Now, you can pay extra for support, but even so it's going to be appalling. You probably will get to talk to someone at least, but in my experience you rarely get ahold of anyone who'll actually read what you've sent them, much less actually try to help. My take away is that support and customer service is just not a priority for Google, what so ever. It's not even a low priority thing, it's a no priority thing.
- logicallee 9y ago
  is there a typo here? 18ms seems like soemthing that could be designed into their whole infrastructure for some reason you don't know about (such as partitioning of networks, whatever) -- after all, it's a cloud solution, right? Maybe 18 ms tacked onto every request is fine for them. To be honest, if all of my Internet requests were 18 ms slower I wouldn't even notice. if I read you correctly and it wasn't a typo, could you mention why you cared so much? The .1 ms you quoted on localhost is extraordinarily quick and I wouldn't expect it to translate to a cloud...
  - mgkimsal 9y ago
    Because we had a shitty codebase that made 350 SQL calls on one page (the user dashboard), so the primary page that everyone used went from < 500ms to > 8 seconds, and the end client who had to test and accept this change didn't like it. It became unusable. To (re?)clarify - this was 18ms per query to the SQL cloud (from a google compute engine). Yes, were there things that could be done to reduce that? Of course. The other guy on the project put some caching in, reduced duplicate DB calls, etc, and got it down to around 3-4 seconds, but without severely diving in and learning far more about this inherited codebase, there wasn't much that could be done to make it much faster. The other steps involved heavier caching, which changed the actual functionality of the service that was contracted before. "real time status updates" had a functional definition of being within X seconds, IIRC. Caching everything so 'new' information wasn't necessarily available to an end user for, say, 3 minutes, wasn't an option, without getting a lot of buy-in from a lot of people. Not saying it wouldn't work technically, but the client relationship was already strained. The move to google was primarily to 'save money' (because someone had $10k service credit with google), but the effort to make it work well on google cloud and SQL combo (wasn't my call) was too much. You can say this isn't google's fault entirely - you'd probably have a strong case - but not everyone is moving pristine clean/strong/good code over, and problems with the original base were amplified by seemingly small issues.
    - logicallee 9y ago
      Thanks for that pretty detailed write-up. I hope you don't think I'm blaming your code base, but just out of pure curiosity, as an amateur question could you mention why the SQL calls couldn't be done in parallel -- did each one depend on the next or something? At the high level I get everything you just said, just curious about this point, if you remember.
- scriptkiddy 9y ago
  I made it a point a long time ago to never use any SaaS, PaaS, or cloud solutions from Google for anything important. First off, their offerings tend to be much more expensive than competitors. Second, Google tends to randomly drop features, services, and support for applications without warning whenever they want. Third, they don't have any actual support at all for anything. I use Linode for all of my hosting, and I can get tech support on the phone in less than 5 minutes even without the upgraded support plan. On top of this, Linode and DigitalOcean offer extensive community curated documentation for almost everything. The above reasons are also why I don't use any open source software developed or maintained solely by Google. There's no guarantee that Google won't just stop maintaining it after a year or less. For a somewhat recent project, my co-founder and I evaluated using Angular2 for the front-end of our application. During our evaluation, we learned about Angular4. Knowing about what happened between Angular1 and Angular2, we decided to steer clear of the angular ecosystem. This decision also made more sense from the performance and ecosystem side of the argument. Most groups that develop open source software do their best to ensure that there is a well documented and simple upgrade path. Look at Django, rails, React, Vue, postgresql, etc for example. As far as I have experienced, Google products offer no such simple upgrade path. To be honest, I'm not really sure what value Google is offering anymore. The only quality product from them I still find any use for is Gmail. Even Gmail has competitors that do Email better. I guess there's also youtube, but Google are kind of shitting that up as well. I think the tech space needs to walk away from worshiping Google and Facebook. In all reality, the technical quality of their products is lacking and their entire goal is to lock both consumers and developers inside their own walled gardens.
  - mgkimsal 9y ago
    I hear you. I've made an exception on angular2, partially because it's being baked in to some other tools I use, and I think those communities will help support it as well. And from what I can tell, the 'a2->a3->a4' is much more minor support numbering vs the "everything is redone" from a1->a2. But this is a somewhat easier choice to make because I still have the code locally. Google could stop supporting it tomorrow, and it would not be fun, but things wouldn't stop working, unlike the firebase situation above, where the option is "turn off your app". That said, I've generally been using knockout and vue (more vue recently than knockout) and both have been fine. "Third, they don't have any actual support at all for anything". There's a lot of people on this thread that feel like because they're paying for something they should get support. And... while I'm sympathetic to that viewpoint, and agree they should, we've had 10+ years of seeing google consistently under- or non- support many services that people seem to want to base their entire businesses on.

by mstade 9y ago

mgkimsal 9y ago

> It really is appalling Never quite understood what's "appalling" about it. They've never had strong support for any dev/cloud-oriented services. I know some folks who've had good experiences, but those were exceptions, not the rule. I'll say, I get the frustration with it, but it's why I've never considered using their cloud services for anything. I had a client who asked me to migrate all their stuff to google 'cloud compute' and their 'sql cloud' offering last year. It was a performance disaster; queries that were minor .1ms on localhost or in the same network were suddenly 18-20ms hitting google cloud. Spent a few days trying to dig in, and a few responses from various groups were "shouldn't be that slow" and "works for me" and that was about it. Gave up trying to get any real response from google (we may have got a 'looking in to it' response - I don't remember). Thing was, I knew going in it wasn't going to end well, I just wasn't sure exactly why (maybe that's just too cynical of a view, but I prefer 'realistic').

samastur 9y ago
Just as a restaurant can be appalling from when it was first opened and even if you don't visit it ever, other services can. Appalling doesn't depend on personal experience or lack of prior knowledge.
elorant 9y ago
It's appalling because you're a paying customer. I can understand free services not having support, but any service that's costing money should at the very least have a human being responding through email.
- mgkimsal 9y ago
  But you KNOW going in - there's no support. You know this by a) hundreds and thousands of horror stories from people saying "I did not get any support from google". You can counter that with "here are people who did get support", but the existence of the thousands of support nightmare stories out there gives you enough evidence to know that this isn't really "supported" in the way you might expect. b) there are no phone numbers. https://firebase.google.com/support/ https://firebase.google.com/support/ - no phone numbers. That's, of course, not the sole criteria you should use to judge potential support, but... it's google. It's 2017. We have multiple years of knowing what Google's track record is re: support for paying customers. Outside of large adwords advertisers (where they make a lot of money), it's non-existent from a practical standpoint. The big problem here is that firebase was fine until google acquired them. I'm reminded of a time when I had a bad experience at a restaurant. Local place I went to with friends from work. We went... probably 4-5 times per year, and I loved the chicken sandwich they did. Last time I went, it was ... bad. Not what I'd expected at all. It was different. Server asks "how's everything?". I said I was disappointed that the chicken sandwich had changed - it was not as good as before. Now... I wasn't actually complaining as in "comp the meal", but I was not a satisfied diner. "Nothing's changed sir". "um... yeah - this is nowhere near as good as it used to be". "You're wrong, nothing's changed". We had a back and forth, and they went back and came out "the cooks say the recipe is exactly the same". This was getting a bit ridiculous. Manager came over, same exchange. He leaves, the comes back and says "they're right - the recipe is 100% the same. We have changed a few ingredients last month, but it's the same". Maybe that doesn't actually relate to the google service situation at all now, come to think of it, but I was reminded of that situation when I started typing it :) Oh yeah - nothing changed from OP's end, except how the new owner (google) decided to start monitoring and accounting for usage. Think of a "$9.99 all you can eat buffet" getting loyal diners, then changing owners, and the new owners giving some people a $95 bill at the end of the meal because they'd decided to also charge for any visits to the salad bar over 2. Most diners would never hit that limit, but a few - loyal diners for years - would have been (rightly) upset at the after-the-fact change.
  - mstade 9y ago
    > But you KNOW going in - there's no support. I know this but to say everyone does or should isn't exactly fair. As well, you assume I am holding the purse strings, that I'm in any way responsible or even have a say in procurement. This may be true in my own business (it is) but not necessarily elsewhere, like my clients (it more often than not isn't) and you'll just have to live with poor decisions made on your behalf. Also quite often, there's a sunk cost that makes switching that much harder. So no, it's not fair to say "Google has shitty support but that's ok because we all kind of know this." Paying customers can and should complain about this, loudly, shouting it from the roof tops. Ideally they vote with their wallets and just leave Google, but again that's just easier said than done in many cases. Some semblance of professionalism shouldn't be too much to ask from one of the world's biggest corporations, an ask coming from paying customers no less.
mstade 9y ago
I don't think it's fair you're being down voted, I get your point of view, so I gave you an upvote. But the thing is, this isn't some kid-in-a-basement free-but-ad-supported service where you-get-what-you-pay-for. Google's cloud offering isn't free (for any use beyond hello world anyway) and as a paying customer I'd expect: 1. Things to work 2. People to answer when things don't work 3. Things to start working again in a timely manner, with timely and concise updates along the way Points two and three carry a lot of nuance, but Google's support does not. It's a virtual wall that people cry to and get no response back from, except for the rare few occasions when the Google deities decide to grace forums or mailing lists with their presence. It's just not a serious offering. Now, you can pay extra for support, but even so it's going to be appalling. You probably will get to talk to someone at least, but in my experience you rarely get ahold of anyone who'll actually read what you've sent them, much less actually try to help. My take away is that support and customer service is just not a priority for Google, what so ever. It's not even a low priority thing, it's a no priority thing.
logicallee 9y ago
is there a typo here? 18ms seems like soemthing that could be designed into their whole infrastructure for some reason you don't know about (such as partitioning of networks, whatever) -- after all, it's a cloud solution, right? Maybe 18 ms tacked onto every request is fine for them. To be honest, if all of my Internet requests were 18 ms slower I wouldn't even notice. if I read you correctly and it wasn't a typo, could you mention why you cared so much? The .1 ms you quoted on localhost is extraordinarily quick and I wouldn't expect it to translate to a cloud...
- mgkimsal 9y ago
  Because we had a shitty codebase that made 350 SQL calls on one page (the user dashboard), so the primary page that everyone used went from < 500ms to > 8 seconds, and the end client who had to test and accept this change didn't like it. It became unusable. To (re?)clarify - this was 18ms per query to the SQL cloud (from a google compute engine). Yes, were there things that could be done to reduce that? Of course. The other guy on the project put some caching in, reduced duplicate DB calls, etc, and got it down to around 3-4 seconds, but without severely diving in and learning far more about this inherited codebase, there wasn't much that could be done to make it much faster. The other steps involved heavier caching, which changed the actual functionality of the service that was contracted before. "real time status updates" had a functional definition of being within X seconds, IIRC. Caching everything so 'new' information wasn't necessarily available to an end user for, say, 3 minutes, wasn't an option, without getting a lot of buy-in from a lot of people. Not saying it wouldn't work technically, but the client relationship was already strained. The move to google was primarily to 'save money' (because someone had $10k service credit with google), but the effort to make it work well on google cloud and SQL combo (wasn't my call) was too much. You can say this isn't google's fault entirely - you'd probably have a strong case - but not everyone is moving pristine clean/strong/good code over, and problems with the original base were amplified by seemingly small issues.
  - logicallee 9y ago
    Thanks for that pretty detailed write-up. I hope you don't think I'm blaming your code base, but just out of pure curiosity, as an amateur question could you mention why the SQL calls couldn't be done in parallel -- did each one depend on the next or something? At the high level I get everything you just said, just curious about this point, if you remember.
scriptkiddy 9y ago
I made it a point a long time ago to never use any SaaS, PaaS, or cloud solutions from Google for anything important. First off, their offerings tend to be much more expensive than competitors. Second, Google tends to randomly drop features, services, and support for applications without warning whenever they want. Third, they don't have any actual support at all for anything. I use Linode for all of my hosting, and I can get tech support on the phone in less than 5 minutes even without the upgraded support plan. On top of this, Linode and DigitalOcean offer extensive community curated documentation for almost everything. The above reasons are also why I don't use any open source software developed or maintained solely by Google. There's no guarantee that Google won't just stop maintaining it after a year or less. For a somewhat recent project, my co-founder and I evaluated using Angular2 for the front-end of our application. During our evaluation, we learned about Angular4. Knowing about what happened between Angular1 and Angular2, we decided to steer clear of the angular ecosystem. This decision also made more sense from the performance and ecosystem side of the argument. Most groups that develop open source software do their best to ensure that there is a well documented and simple upgrade path. Look at Django, rails, React, Vue, postgresql, etc for example. As far as I have experienced, Google products offer no such simple upgrade path. To be honest, I'm not really sure what value Google is offering anymore. The only quality product from them I still find any use for is Gmail. Even Gmail has competitors that do Email better. I guess there's also youtube, but Google are kind of shitting that up as well. I think the tech space needs to walk away from worshiping Google and Facebook. In all reality, the technical quality of their products is lacking and their entire goal is to lock both consumers and developers inside their own walled gardens.
- mgkimsal 9y ago
  I hear you. I've made an exception on angular2, partially because it's being baked in to some other tools I use, and I think those communities will help support it as well. And from what I can tell, the 'a2->a3->a4' is much more minor support numbering vs the "everything is redone" from a1->a2. But this is a somewhat easier choice to make because I still have the code locally. Google could stop supporting it tomorrow, and it would not be fun, but things wouldn't stop working, unlike the firebase situation above, where the option is "turn off your app". That said, I've generally been using knockout and vue (more vue recently than knockout) and both have been fine. "Third, they don't have any actual support at all for anything". There's a lot of people on this thread that feel like because they're paying for something they should get support. And... while I'm sympathetic to that viewpoint, and agree they should, we've had 10+ years of seeing google consistently under- or non- support many services that people seem to want to base their entire businesses on.