5 ms·

Show HN: AvatarPicker – User-friendly, in-page gravatar alternative

42 points by bgdam 12y ago

aakilfernandes 12y ago
Well done. How come you decided to go with an iframe instead of a generic popup? Also what do you use for image hosting? s3?
- bgdam 12y ago
  Thanks. I'm not sure what exactly you mean by 'generic popup'. I'm assuming you are referring to overlaid div. I went with the iframe mainly to avoid having to provide CORS access to all the implementing sites. This way, it's always a same origin ajax/form post. Since this is just an MVP (I repurposed the site from an older project, the only thing new is the actual widget), the images are currently on a Linode VPN. When/If it gets popular, i'll be re-engineering this.
  - aakilfernandes 12y ago
    I think you misunderstand CORS. You have to enable CORS on your server, not the sites that use your widget. http://stackoverflow.com/questions/298745/how-do-i-send-a-cross-domain-post-request-via-javascript/7605119#7605119 http://stackoverflow.com/questions/298745/how-do-i-send-a-cr...
    - bgdam 12y ago
      Yeah, I know that. What I meant was that I did not want to set wildcards for the access-control-allow-origin header. If you use hostname reflection like in the linked stackoverflow example, issues arise when the widget is cached browser side by site A and then invoked by site B.
  - callumjones 12y ago
    So they're not on a CDN as mentioned on the product page?
    - bgdam 12y ago
      As of now, no they are not. If/when the widget takes off, I will start saving avatars to a CDN. I guess that makes it a bit dishonest to mention the CDN on the landing page, but since it will definitely be part of the standard package before I start charging, I let it slide.
m4tthumphrey 12y ago
This is cool. I've thought about this sort of the service before. It surprises me that Gravatar has never bothered creating a widget.
- bgdam 12y ago
  Thanks. This idea only occurred to me when another dev friend of mine was whining about all the support calls he got from non-tech people who did not know what Gravatar was, and how he had to spend so much time walking people through the Gravatar process. Also the fact that setting up a gravatar the way you like requires you to get a Wordpress.com account is insane. Anyway, glad you liked it.
funkyy 12y ago
I think you could adjust your pricing. Usual for new products a year long subscription is priced at 10 times the monthly plan. For few dollars savings I am not going to risk signing up with new service if I am not sure if it will run or not for a year. 2 months worth of subscription is good enough incentive though. btw - "up to" not "upto". "upto" is not correct afaik Looks nice, I will hopefully see it around the web soon!
- bgdam 12y ago
  Glad you liked it. The pricing is only tentative. I thought it would be better to set expectations higher rather than set expectations lower and then irk people by raising prices (if required). I do agree that making it two months free would be a better incentive than one month free. Thanks for the language correction too. English is not my first language and at times I mess it up. Will fix ASAP.
  - keerthiko 12y ago
    Since you're not on an app store that restricts your pricing flexibility, I think it's totally okay to raise your price later when you've gained stability. You won't really be irking anyone. I think making a more convincing sale value to early adopters is worth it, and later you can raise the price to justify the increased stability you will be offering to the risk-averse later-adopters. At worst you'll cause on-the-edge customers to suffer "loss by missing out" on the "deal" when it was cheaper, but that's not a bad thing. It's how airline ticket sales pages use cookies to tell if you're looking at the same flight 2 days in a row and raise the price by a couple tens of dollars to make you fear the price will keep going up if you wait too long to buy :)
    - bgdam 12y ago
      Thanks for your advice. I think I will be giving the pricing page some serious thought in the next few days.
  - j_s 12y ago
    Avoid upsetting customers by allowing them to keep their current price and only charging new customers the higher price.
toothbrush 12y ago
Does this mean that if sites i like (and i) use your service, my machine hits your server very frequently, thus giving you relatively thorough insight into my browsing habits? This question comes from someone who tries to be strict with the number of RequestPolicy rules allowed. Forgive me if this is a stupid question, but humour the paranoid ;).
- bgdam 12y ago
  Well, if you are going to upload your own image, then only a salted hash of the username you use on that site will be available AvatarPicker. However, if you use Twitter, Facebook or Gravatar, then yeah AvatarPicker will have your Twitter/Facebook username or hashed gravatar email. Without this, it would be impossible retrieve your avatar from those services. I guess the best I can do is codify in the privacy policy, that we will not track any kind of correlation between your service usernames and the websites that they are used on
donatj 12y ago
Gravatar is free though, its hard to compete with free.
- bgdam 12y ago
  Hard, but not impossible. I believe for a lot of companies, the reduced customer support pertaining to changing user avatars (esp. for a non-tech based user base) and the better UX, will more than make up for what they pay.
  - voidz 12y ago
    Gravatar is too ubiquitous for me now that Wordpress owns it. Plus the Wordpress account requirement is really annoying. So I hope your project will become an alternative on many sites, OP!
programminggeek 12y ago
What pain does this solve? Is that pain worth the money?
fiatjaf 12y ago
For the indieweb enthusiasts, http://webvatar.com/ http://webvatar.com/ is Gravatar for websites, and you don't need to set up anything, just put the link there and it will fetch and cache the image from the referrer website. There are not much people marking their websites with microformats, but this can be a minor problem.
skeoh 12y ago
The pricing table shocked me until I realised the plans were ordered most to least expensive, whereas I had subconsciously expected it to be ordered the other way. Looking at http://ui-patterns.com/patterns/PricingTable http://ui-patterns.com/patterns/PricingTable it seems there's no real standard approach to ordering pricing tiers. Does anybody have any insight into the psychology of pricing tables from a UX perspective?
developer1 12y ago
I don't mean to overly offend, but something like this is not worth more than $5/month. Maybe $20/month for 1 million+ users. Your $19/month ($11/month if paid annually) is more than a lot of individuals and small businesses spend on VPS hosting. Your only target with this pricing would be medium-to-large companies, and yet your published pricing ends at an extremely meagre 100,000 users which would not begin to cover them. By the way, your method of popping up the div overlay removes the scrollbar if present (at least in your demo), which makes centred layouts jump around. Raises a yellow flag regarding implementation quality.
- developer1 12y ago
  Whoa. You don't even have actual facebook or twitter integration requiring authentication. You just pull the photo by username, which means no identity validation. Facebook/twitter should show the username along with the photo to indicate a verified connection. Plus after picking someone else's account, I can't delete the photo due to an error. Oh, and then refreshing the page gives an nginx error. This service is far from ideal, sorry. And at this price? Another get-rich-quick personal project deployed by someone who hasn't managed to write a scaleable or truly useful service. Now that probably offends :/
  - bgdam 12y ago
    Considering that this is an MVP, I am quite happy with the way the site managed the HN crowd. And yes, it doesn't have actual validation for Facebook/Twitter. This was done mainly for ease of use and preventing user confusion seeing as it is aimed at non-tech end users. If actual paying customers require this, then I'll implement it. As for the errors, it would be helpful if you could copy paste whatever is listed on the console. While I have been hearing that the pricing is a bit on the higher side, I don't think I will ever be lowering it to levels you have suggested. I guess the companies who believe it's not worth paying the extra few dollars to have a dedicated service can always implement things in-house. Finally, this is an MVP, and of-course it is not going to scale. I don't even know if there is demand to actually develop this into a real product. As such no project is 100% bug free on the day it is launched. Projects with such requirements are still stuck in a room full of programmers despairing at the latest bug found. Going so far as to call it a get rich quick scheme just because it does not meet your idealized notions of what a service is worth, or how it should be implemented is unwarranted. The entire tone of your comment appears to be more into trashing me/the project. As such, it would have been more helpful if you had simply stuck to the constructive criticism.

42 points by bgdam 12y ago

Well done. How come you decided to go with an iframe instead of a generic popup? Also what do you use for image hosting? s3?

bgdam 12y ago
Thanks. I'm not sure what exactly you mean by 'generic popup'. I'm assuming you are referring to overlaid div. I went with the iframe mainly to avoid having to provide CORS access to all the implementing sites. This way, it's always a same origin ajax/form post. Since this is just an MVP (I repurposed the site from an older project, the only thing new is the actual widget), the images are currently on a Linode VPN. When/If it gets popular, i'll be re-engineering this.
- aakilfernandes 12y ago
  I think you misunderstand CORS. You have to enable CORS on your server, not the sites that use your widget. http://stackoverflow.com/questions/298745/how-do-i-send-a-cross-domain-post-request-via-javascript/7605119#7605119 http://stackoverflow.com/questions/298745/how-do-i-send-a-cr...
  - bgdam 12y ago
    Yeah, I know that. What I meant was that I did not want to set wildcards for the access-control-allow-origin header. If you use hostname reflection like in the linked stackoverflow example, issues arise when the widget is cached browser side by site A and then invoked by site B.
- callumjones 12y ago
  So they're not on a CDN as mentioned on the product page?
  - bgdam 12y ago
    As of now, no they are not. If/when the widget takes off, I will start saving avatars to a CDN. I guess that makes it a bit dishonest to mention the CDN on the landing page, but since it will definitely be part of the standard package before I start charging, I let it slide.

m4tthumphrey 12y ago

This is cool. I've thought about this sort of the service before. It surprises me that Gravatar has never bothered creating a widget.

bgdam 12y ago
Thanks. This idea only occurred to me when another dev friend of mine was whining about all the support calls he got from non-tech people who did not know what Gravatar was, and how he had to spend so much time walking people through the Gravatar process. Also the fact that setting up a gravatar the way you like requires you to get a Wordpress.com account is insane. Anyway, glad you liked it.

funkyy 12y ago

I think you could adjust your pricing. Usual for new products a year long subscription is priced at 10 times the monthly plan. For few dollars savings I am not going to risk signing up with new service if I am not sure if it will run or not for a year. 2 months worth of subscription is good enough incentive though. btw - "up to" not "upto". "upto" is not correct afaik Looks nice, I will hopefully see it around the web soon!

bgdam 12y ago
Glad you liked it. The pricing is only tentative. I thought it would be better to set expectations higher rather than set expectations lower and then irk people by raising prices (if required). I do agree that making it two months free would be a better incentive than one month free. Thanks for the language correction too. English is not my first language and at times I mess it up. Will fix ASAP.
- keerthiko 12y ago
  Since you're not on an app store that restricts your pricing flexibility, I think it's totally okay to raise your price later when you've gained stability. You won't really be irking anyone. I think making a more convincing sale value to early adopters is worth it, and later you can raise the price to justify the increased stability you will be offering to the risk-averse later-adopters. At worst you'll cause on-the-edge customers to suffer "loss by missing out" on the "deal" when it was cheaper, but that's not a bad thing. It's how airline ticket sales pages use cookies to tell if you're looking at the same flight 2 days in a row and raise the price by a couple tens of dollars to make you fear the price will keep going up if you wait too long to buy :)
  - bgdam 12y ago
    Thanks for your advice. I think I will be giving the pricing page some serious thought in the next few days.
- j_s 12y ago
  Avoid upsetting customers by allowing them to keep their current price and only charging new customers the higher price.

toothbrush 12y ago

Does this mean that if sites i like (and i) use your service, my machine hits your server very frequently, thus giving you relatively thorough insight into my browsing habits? This question comes from someone who tries to be strict with the number of RequestPolicy rules allowed. Forgive me if this is a stupid question, but humour the paranoid ;).

bgdam 12y ago
Well, if you are going to upload your own image, then only a salted hash of the username you use on that site will be available AvatarPicker. However, if you use Twitter, Facebook or Gravatar, then yeah AvatarPicker will have your Twitter/Facebook username or hashed gravatar email. Without this, it would be impossible retrieve your avatar from those services. I guess the best I can do is codify in the privacy policy, that we will not track any kind of correlation between your service usernames and the websites that they are used on

donatj 12y ago

Gravatar is free though, its hard to compete with free.

bgdam 12y ago
Hard, but not impossible. I believe for a lot of companies, the reduced customer support pertaining to changing user avatars (esp. for a non-tech based user base) and the better UX, will more than make up for what they pay.
- voidz 12y ago
  Gravatar is too ubiquitous for me now that Wordpress owns it. Plus the Wordpress account requirement is really annoying. So I hope your project will become an alternative on many sites, OP!

programminggeek 12y ago

What pain does this solve? Is that pain worth the money?

fiatjaf 12y ago

For the indieweb enthusiasts, http://webvatar.com/ http://webvatar.com/ is Gravatar for websites, and you don't need to set up anything, just put the link there and it will fetch and cache the image from the referrer website. There are not much people marking their websites with microformats, but this can be a minor problem.

skeoh 12y ago

The pricing table shocked me until I realised the plans were ordered most to least expensive, whereas I had subconsciously expected it to be ordered the other way. Looking at http://ui-patterns.com/patterns/PricingTable http://ui-patterns.com/patterns/PricingTable it seems there's no real standard approach to ordering pricing tiers. Does anybody have any insight into the psychology of pricing tables from a UX perspective?

developer1 12y ago

I don't mean to overly offend, but something like this is not worth more than $5/month. Maybe $20/month for 1 million+ users. Your $19/month ($11/month if paid annually) is more than a lot of individuals and small businesses spend on VPS hosting. Your only target with this pricing would be medium-to-large companies, and yet your published pricing ends at an extremely meagre 100,000 users which would not begin to cover them. By the way, your method of popping up the div overlay removes the scrollbar if present (at least in your demo), which makes centred layouts jump around. Raises a yellow flag regarding implementation quality.

developer1 12y ago
Whoa. You don't even have actual facebook or twitter integration requiring authentication. You just pull the photo by username, which means no identity validation. Facebook/twitter should show the username along with the photo to indicate a verified connection. Plus after picking someone else's account, I can't delete the photo due to an error. Oh, and then refreshing the page gives an nginx error. This service is far from ideal, sorry. And at this price? Another get-rich-quick personal project deployed by someone who hasn't managed to write a scaleable or truly useful service. Now that probably offends :/
- bgdam 12y ago
  Considering that this is an MVP, I am quite happy with the way the site managed the HN crowd. And yes, it doesn't have actual validation for Facebook/Twitter. This was done mainly for ease of use and preventing user confusion seeing as it is aimed at non-tech end users. If actual paying customers require this, then I'll implement it. As for the errors, it would be helpful if you could copy paste whatever is listed on the console. While I have been hearing that the pricing is a bit on the higher side, I don't think I will ever be lowering it to levels you have suggested. I guess the companies who believe it's not worth paying the extra few dollars to have a dedicated service can always implement things in-house. Finally, this is an MVP, and of-course it is not going to scale. I don't even know if there is demand to actually develop this into a real product. As such no project is 100% bug free on the day it is launched. Projects with such requirements are still stuck in a room full of programmers despairing at the latest bug found. Going so far as to call it a get rich quick scheme just because it does not meet your idealized notions of what a service is worth, or how it should be implemented is unwarranted. The entire tone of your comment appears to be more into trashing me/the project. As such, it would have been more helpful if you had simply stuck to the constructive criticism.