5 ms·

It's a shame really, but as an outsider, it wasn't clear that Persona was even given a chance to fail.

by wamatt 11y ago

StavrosK 11y ago
That's my impression as well. A friend and I created a pastebin as a fun weekend project this weekend (https://www.pastery.net/ https://www.pastery.net/) and we used Persona for auth, since it's so damn convenient. The only problem is that it loads a bunch of files on every request, and my questions about lazy-loading went unanswered. I really wish Mozilla would change their mind and start developing this again. It feels like they expected all major websites to support it on the first year, and, when that didn't happen, they decided to drop it.
- click170 11y ago
  With Thunderbird at least, the repo is open and available. Does the same stand for Persona? Perhaps instead of asking Firefox to do the legwork we should be contributing some of our own time to the projects we want to see succeed.
  - StavrosK 11y ago
    How is that relevant? Are you saying that Persona failed because they didn't get enough pull requests from other users?
    - citruspi 11y ago
      I think they're saying that even if Mozilla may not decide to actively develop Persona anymore, Persona users and fans would be able to contribute to and further develop the project if the repo is "open and available."
      - StavrosK 11y ago
        That's not happening, unfortunately. Sometimes you need someone with weight to throw around, or with technical knowledge, or both.
        click170 11y ago
        I agree sometimes it takes someone with weight to throw around, but are you saying that person needs to be a Mozilla employee? Because I wouldn't agree with that.
        StavrosK 11y ago
        No, it needs to be Mozilla itself.
  - citruspi 11y ago
    It does - https://github.com/mozilla/persona https://github.com/mozilla/persona
- DHowett 11y ago
  Incidentally, I launched a pastebin (https://ghostbin.com/ https://ghostbin.com/) some time ago, used Persona for authentication, and absolutely loved it. Pastery looks awesome. It appears that you're using Pygments server-side?
  - rsync 11y ago
    Serious question for you and your parent: Why would a pastebin site need auth/login/userids ? All I can think of is editing pastes, but isn't it easier to just create a new (anonymous) paste instead of editing ?
    - lotyrin 11y ago
      Deletion. Modification of pastes and maintaining the same id (URL). Though, these can be solved with tokens per post, it's tough to remember/maintain multiple tokens, doesn't work out in practice.
    - DHowett 11y ago
      They're primarily for content ownership management. I'd prefer not to have accounts, and Ghostbin tries to avoid keeping identifiable information around, but they do come in fairly handy for maintaining control of the things you've pasted and the right to revoke them. Session cookies aren't quite up to the task, since they can't stick around forever and are difficult to bring with you between workspaces. (edit to elaborate on "avoid keeping identifiable information": account names are hashed, the account never stores user identification, and a paste doesn't track which account created it. For most intents and many purposes, the creator of a paste is anonymous. We could get into drawing correlations between all the pastes owned by an account, but I'm not particularly interested in sitting around and doing that.)
    - r-w 11y ago
      Preventing inundation with pastes. This is less of a problem for a small project like this compared to one of the major sites, whose server architecture is probably more robust.
  - StavrosK 11y ago
nickbauman 11y ago
Absolutely right. Dan Callahan's statement on why: "with Mozilla Labs somewhat sudden dissolution, we were unexpectedly asked to demonstrate traction and commercial adoption that simply wasn't there." Commercial adoption? What does that even mean in the context of MoCo? What does MoCo sell? And 24 months? Makes no sense, even for MoCo. Google is consistently a small player in the Cloud Computing Platform space. Their response: "we're not going away." MoCo has scores of millions in the bank. I'm still baffled.
- derefr 11y ago
  I think that for Persona in particular, "commercial adoption" would probably have meant "random startups deploying Persona as their customers' identity provider." The idea being that a government/nonprofit/enterprise can be convinced to adopt any old ridiculous thing if you say the right things about it, but "the market" (i.e. a critical mass of SMBs) will only adopt your product if it's giving them a higher ROI than the alternative.
  - nickbauman 11y ago
    Yeah I don't buy that either. Was there an ROI case made about the iPhone? I don't think so. There's no similar ROI case anyone could make for Firefox, either.
    - derefr 11y ago
      The original iPhone wasn't a platform (didn't have "apps"), remember; it had value without third party support. Persona, meanwhile, has no value apart from its adoption. Obviously, it doesn't need adoption to "survive", but it does need adoption to succeed in its goals: rescuing users from stupid hack-job login systems. I would compare it more to Chrome: Google made Chrome with the goal of getting the web to work better for some people, so they could more easily consume Google's various services. You could say that Chrome's bottom line was "increased Google service adoption": if Chrome renders faster and provides more HTML5 features, then Google gets more and happier customers for their products. Mozilla doesn't have "services" that Persona is a loss-leader for, but there's still that same ROI calculation of whether Persona is achieving its goals, and measuring adoption rate is a good proxy for that. The important thing, that Mozilla failed to do, was to actually understand what an adoption curve for something like Persona should look like, and what things affect it. They needed to heavily advertise it (to developers, not to users!) over about two years, and then start looking at the adoption curve.

It's a shame really, but as an outsider, it wasn't clear that Persona was even given a chance to fail.

by wamatt 11y ago

It's a shame really, but as an outsider, it wasn't clear that Persona was even given a chance to fail.

That's my impression as well. A friend and I created a pastebin as a fun weekend project this weekend (https://www.pastery.net/ https://www.pastery.net/) and we used Persona for auth, since it's so damn convenient. The only problem is that it loads a bunch of files on every request, and my questions about lazy-loading went unanswered. I really wish Mozilla would change their mind and start developing this again. It feels like they expected all major websites to support it on the first year, and, when that didn't happen, they decided to drop it.

click170 11y ago
With Thunderbird at least, the repo is open and available. Does the same stand for Persona? Perhaps instead of asking Firefox to do the legwork we should be contributing some of our own time to the projects we want to see succeed.
- StavrosK 11y ago
  How is that relevant? Are you saying that Persona failed because they didn't get enough pull requests from other users?
  - citruspi 11y ago
    I think they're saying that even if Mozilla may not decide to actively develop Persona anymore, Persona users and fans would be able to contribute to and further develop the project if the repo is "open and available."
    - StavrosK 11y ago
      That's not happening, unfortunately. Sometimes you need someone with weight to throw around, or with technical knowledge, or both.
      - click170 11y ago
        I agree sometimes it takes someone with weight to throw around, but are you saying that person needs to be a Mozilla employee? Because I wouldn't agree with that.
        StavrosK 11y ago
        No, it needs to be Mozilla itself.
- citruspi 11y ago
  It does - https://github.com/mozilla/persona https://github.com/mozilla/persona
DHowett 11y ago
Incidentally, I launched a pastebin (https://ghostbin.com/ https://ghostbin.com/) some time ago, used Persona for authentication, and absolutely loved it. Pastery looks awesome. It appears that you're using Pygments server-side?
- rsync 11y ago
  Serious question for you and your parent: Why would a pastebin site need auth/login/userids ? All I can think of is editing pastes, but isn't it easier to just create a new (anonymous) paste instead of editing ?
  - lotyrin 11y ago
    Deletion. Modification of pastes and maintaining the same id (URL). Though, these can be solved with tokens per post, it's tough to remember/maintain multiple tokens, doesn't work out in practice.
  - DHowett 11y ago
    They're primarily for content ownership management. I'd prefer not to have accounts, and Ghostbin tries to avoid keeping identifiable information around, but they do come in fairly handy for maintaining control of the things you've pasted and the right to revoke them. Session cookies aren't quite up to the task, since they can't stick around forever and are difficult to bring with you between workspaces. (edit to elaborate on "avoid keeping identifiable information": account names are hashed, the account never stores user identification, and a paste doesn't track which account created it. For most intents and many purposes, the creator of a paste is anonymous. We could get into drawing correlations between all the pastes owned by an account, but I'm not particularly interested in sitting around and doing that.)
  - r-w 11y ago
    Preventing inundation with pastes. This is less of a problem for a small project like this compared to one of the major sites, whose server architecture is probably more robust.
- StavrosK 11y ago

nickbauman 11y ago

Absolutely right. Dan Callahan's statement on why: "with Mozilla Labs somewhat sudden dissolution, we were unexpectedly asked to demonstrate traction and commercial adoption that simply wasn't there." Commercial adoption? What does that even mean in the context of MoCo? What does MoCo sell? And 24 months? Makes no sense, even for MoCo. Google is consistently a small player in the Cloud Computing Platform space. Their response: "we're not going away." MoCo has scores of millions in the bank. I'm still baffled.

derefr 11y ago
I think that for Persona in particular, "commercial adoption" would probably have meant "random startups deploying Persona as their customers' identity provider." The idea being that a government/nonprofit/enterprise can be convinced to adopt any old ridiculous thing if you say the right things about it, but "the market" (i.e. a critical mass of SMBs) will only adopt your product if it's giving them a higher ROI than the alternative.
- nickbauman 11y ago
  Yeah I don't buy that either. Was there an ROI case made about the iPhone? I don't think so. There's no similar ROI case anyone could make for Firefox, either.
  - derefr 11y ago
    The original iPhone wasn't a platform (didn't have "apps"), remember; it had value without third party support. Persona, meanwhile, has no value apart from its adoption. Obviously, it doesn't need adoption to "survive", but it does need adoption to succeed in its goals: rescuing users from stupid hack-job login systems. I would compare it more to Chrome: Google made Chrome with the goal of getting the web to work better for some people, so they could more easily consume Google's various services. You could say that Chrome's bottom line was "increased Google service adoption": if Chrome renders faster and provides more HTML5 features, then Google gets more and happier customers for their products. Mozilla doesn't have "services" that Persona is a loss-leader for, but there's still that same ROI calculation of whether Persona is achieving its goals, and measuring adoption rate is a good proxy for that. The important thing, that Mozilla failed to do, was to actually understand what an adoption curve for something like Persona should look like, and what things affect it. They needed to heavily advertise it (to developers, not to users!) over about two years, and then start looking at the adoption curve.