currently beta release is set to occur near the end of next week at current pace.
The current goals which need to be achieved for this transition are as follows: * Subscriptions as opposed to blocking-based content presentation model. To deal with low volume initially, every sub will be subscribed to by default, and an opt-out model will be followed. This will change when/if defaults become established.
* user blocking
* mod ui improvements to at least a passable state
* RSS
* general style improvements/bug fixes. no more boring gray 'comment' button, messages will look better, etc. sub names will appear in top left as normal text similar to how the 'about' page is now. jquery message polling (messages will change color within a few seconds of receiving one)
* change to /i/ sub strings as opposed to /r/. old /r/ format will redirect to /i/
* an actual damn logo. I like the idea of a greek letter or similar... perhaps just the 'beta' character for beta? lol. i'd love the name of the site to be 'agora' or similar, but I doubt i'll ever be able to snag a decent tld
* post hide button
* warrant canary
* traffic logs will no longer be kept past 1 day of age on the production server
* youtube videos embedding instead of just the image
and finally
* tor hidden service (should be easy enough to implement, the only thing i'll have to change is moving the DB to a dedicated server).
* (optional) full client-to-client encryption on user messages
there are a few different ways i can implement end-to-end encryption, but right now the plan goes something along the lines of:
\1. allow accounts to opt into encrypted messaging. when opted in, a public/private keypair will be generated for the user account. an icon/text/something will show on the user profile showing they have enabled encrypted messaging
\2. on opt-in, the user is prompted (or assigned, haven't decided yet) for a 'secret phrase/password' that will be used to generate their private key and decrypt messages
\3. messages sent to users will have an optional 'encrypt' button. when this button is pressed they will be prompted for their passphrase. from a ui perspective, this could be inconvenient. after one-decryption, i could store the result in a session or cookie, but that severely diminishes the security. so i don't think i have another option here but to prompt users for their passphrase every time a message is decrypted.
\4. handling of public keys, and encrypted messages, beyond the initial 'enter passphrase' will be seamless
if there is a better way of doing this i would love input
============
finally, attracting users. when beta is reached, i will post ieddit.com again to a few places including hackernews. while the reaction last time was great, there is no gaurentee a similar amount of interest will occur.
people using, and being aware of, the platform is by far the most important goal to achieve. that being said, I have no fucking clue how to do this after the initial beta announcements to HN/reddit/etc.
i'm not adverse to bots. i hate to artificially create false usage, but there may not be another option initially. reddit relied heavily on bots in it's early day as well.
any bots which post to the site must operate organically. for posts pulled directly from reddit, for example, a few comments/comment threads should be pulled too. bonus points if there is additional sophistication such as pulling comments from last iterations of the same URL posted to reddit, some verbage changes, etc.
if any of you guys think you can create such a bot, contact me once the prototype is written, and if it's decent i'll give it a 'bypass captcha' key. i'll probably write such a bot myself if it comes down to it, i already have written an extremely basic bot but am not running it currently.
the post requests are simple enough to where an api shouldn't really be required.
anyway, that's the status of things, the roadmap, and so on. i'll "sticky" this post as an announcement to ensure it's seen for a while... hopefully the shift from alpha to beta will breathe new life into the platform.
i'm not adverse to bots. i hate to artificially create false usage, but there may not be another option initially. reddit relied heavily on bots in it's early day as well.
I just accidentally pressed on "New Comment" and I was able to leave a blank comment LMAO Anyways, your goals look pretty good. I concur that you giving the site identity (/r/ -> /i/, logo, etc...) will provide a unique "feel" to the place instead of being a pure clone for Reddit. That said, the site currently suffer from 1 issue: - Not enough incentive for a switch from R or HN. Are there any unique subs here that R doesn't have? I think /i/imagine is a good one but we need more. Do we have a distinct competitive advantage that may bring more users here? Something has to be "attractive" enough. Anon posting is a start, but you need more than that to retain users.
I just accidentally pressed on "New Comment" and I was able to leave a blank comment LMAO
blank comments are valid comments : ) you can do the same thing on reddit
Not enough incentive for a switch from R or HN. Are there any unique subs here that R doesn't have
it's difficult to iterate on the platform.
the niche i originally wanted to fill was providing a platform for communities whose content was not relevant to hackernews, or not advertiser friendly on reddit.
that being said, saidit has fulfilled some of this niche. there are some communities in which the platform i'm trying to provide will have it's advantages, for example: anything warez focused, anything relating to tor/anonymity darknet-related shit, political speech in which the authors may see blowback from in their host countries, sourcing discussion surrounding kava/kratom/etc...
i also am attempting to run/setup things in such a why that user privacy is PARAMOUNT. i want things to be setup to where even if an adversary gains physical access to the server, they will compromise a minimal amount of data.
there is a LOT that goes into maximizing user privacy, pgp messaging being one example.
I think you've gotta go for something like the cock.li policy: he promises to not go out of his way to fuck you over but ultimately, you cannot trust him and you're just going off his word.
currently beta release is set to occur near the end of next week at current pace.
The current goals which need to be achieved for this transition are as follows:
* Subscriptions as opposed to blocking-based content presentation model. To deal with low volume initially, every sub will be subscribed to by default, and an opt-out model will be followed. This will change when/if defaults become established.
* user blocking
* mod ui improvements to at least a passable state
* RSS
* general style improvements/bug fixes. no more boring gray 'comment' button, messages will look better, etc. sub names will appear in top left as normal text similar to how the 'about' page is now. jquery message polling (messages will change color within a few seconds of receiving one)
* change to /i/ sub strings as opposed to /r/. old /r/ format will redirect to /i/
* an actual damn logo. I like the idea of a greek letter or similar... perhaps just the 'beta' character for beta? lol. i'd love the name of the site to be 'agora' or similar, but I doubt i'll ever be able to snag a decent tld
* post hide button
* warrant canary
* traffic logs will no longer be kept past 1 day of age on the production server
* youtube videos embedding instead of just the image
and finally
* tor hidden service (should be easy enough to implement, the only thing i'll have to change is moving the DB to a dedicated server).
* (optional) full client-to-client encryption on user messages
there are a few different ways i can implement end-to-end encryption, but right now the plan goes something along the lines of:
\1. allow accounts to opt into encrypted messaging. when opted in, a public/private keypair will be generated for the user account. an icon/text/something will show on the user profile showing they have enabled encrypted messaging
\2. on opt-in, the user is prompted (or assigned, haven't decided yet) for a 'secret phrase/password' that will be used to generate their private key and decrypt messages
\3. messages sent to users will have an optional 'encrypt' button. when this button is pressed they will be prompted for their passphrase. from a ui perspective, this could be inconvenient. after one-decryption, i could store the result in a session or cookie, but that severely diminishes the security. so i don't think i have another option here but to prompt users for their passphrase every time a message is decrypted.
\4. handling of public keys, and encrypted messages, beyond the initial 'enter passphrase' will be seamless
if there is a better way of doing this i would love input
============
finally, attracting users. when beta is reached, i will post ieddit.com again to a few places including hackernews. while the reaction last time was great, there is no gaurentee a similar amount of interest will occur.
people using, and being aware of, the platform is by far the most important goal to achieve. that being said, I have no fucking clue how to do this after the initial beta announcements to HN/reddit/etc.
i'm not adverse to bots. i hate to artificially create false usage, but there may not be another option initially. reddit relied heavily on bots in it's early day as well.
any bots which post to the site must operate organically. for posts pulled directly from reddit, for example, a few comments/comment threads should be pulled too. bonus points if there is additional sophistication such as pulling comments from last iterations of the same URL posted to reddit, some verbage changes, etc.
if any of you guys think you can create such a bot, contact me once the prototype is written, and if it's decent i'll give it a 'bypass captcha' key. i'll probably write such a bot myself if it comes down to it, i already have written an extremely basic bot but am not running it currently.
the post requests are simple enough to where an api shouldn't really be required.
anyway, that's the status of things, the roadmap, and so on. i'll "sticky" this post as an announcement to ensure it's seen for a while... hopefully the shift from alpha to beta will breathe new life into the platform.