A lot of them have timed out (auto destruct test features). So I need to turn a lot of them back on. I may stop playing with the auto destruct on new features. I'm not sure. It sucks to have to go in and turn back on obviously good things. But it is a hedge against making something that turns out to be a miss and then it just hangs around.
Maybe it would be cool if I automated it making a poll when the time runs out so I don't have to explicitly remember to make one for every feature I coded two weeks ago.
I like the auto-destruct feature idea - but it might be worth having a red and yellow exclamation in a box icon that you can hover over to read what the feature is, when it will expire, and a link to a poll about the feature that could become permanent with enough positive feedback. Also, a running list of these features would be nice to see in order to better know/notice what's new, what's going on, and how we can help with feedback.
It triples the UI considerations (actually more). Some of these features aren't even UI related. The timing conditional for when something is new or not would clutter the html templates considerably and those are already at their maximum mess level.
Timing a feature turning off is shockingly easy without much clutter. Just wrap the new code in a time check or have a flag set with a single if statement. And that goes in the backend code which is a lot easier to segment and keep organized even if there are a few more if statements sprinkled in.
So in short and to be nice about it, it's a solidly bad idea. If we want to iterate new features fast it doesn't make sense to make it 5 times harder to do each one (small features), and age organization critical code faster. All projects hit a wall of complexity at some point. Inviting that sort of thing would just encourage the project hitting that wall faster. I'm currently trying to refactor the html templates to push the organizational age down, and having tons of duplicat content behind conditionals whose relevence is shifting with time would impede an already complex problem.
Another way to look at it is obviously design is one of my bottlenecks. You are asking me to do more design work for something that will be temporary. It takes more than 2x design effort to make something that looks good in two cases (alerted as new and not, and possibly removed if the feature dies).
Maybe you can show me a CSS example of something that looks good. Making something pop out and keep the whole page tasteful is hard. It takes time. If some small improvement can be completed in 2 minutes it doesn't make sense meditating for 10 hours on an ideal design to alert it (tastefully) and then figure out a tasteful way to add it to code without cluttering the code. Compared to a if(Date.now()<17650401401) { do extra backend stuff}, it's a completely different burden.
I did make an honest attempt at your idea at one point. But it's a no. It's not going to be a thing. But I guess if after my next feature announcement you make a theme for how it should look I can push it out.
Log in
</end Sarcasm>
I just saw that that page still says "Submit a link post".
I like the new features I see.
A lot of them have timed out (auto destruct test features). So I need to turn a lot of them back on. I may stop playing with the auto destruct on new features. I'm not sure. It sucks to have to go in and turn back on obviously good things. But it is a hedge against making something that turns out to be a miss and then it just hangs around.
Maybe it would be cool if I automated it making a poll when the time runs out so I don't have to explicitly remember to make one for every feature I coded two weeks ago.
I like the auto-destruct feature idea - but it might be worth having a red and yellow exclamation in a box icon that you can hover over to read what the feature is, when it will expire, and a link to a poll about the feature that could become permanent with enough positive feedback. Also, a running list of these features would be nice to see in order to better know/notice what's new, what's going on, and how we can help with feedback.
That poll page's title is still wrong.
It triples the UI considerations (actually more). Some of these features aren't even UI related. The timing conditional for when something is new or not would clutter the html templates considerably and those are already at their maximum mess level.
Timing a feature turning off is shockingly easy without much clutter. Just wrap the new code in a time check or have a flag set with a single if statement. And that goes in the backend code which is a lot easier to segment and keep organized even if there are a few more if statements sprinkled in.
So in short and to be nice about it, it's a solidly bad idea. If we want to iterate new features fast it doesn't make sense to make it 5 times harder to do each one (small features), and age organization critical code faster. All projects hit a wall of complexity at some point. Inviting that sort of thing would just encourage the project hitting that wall faster. I'm currently trying to refactor the html templates to push the organizational age down, and having tons of duplicat content behind conditionals whose relevence is shifting with time would impede an already complex problem.
Another way to look at it is obviously design is one of my bottlenecks. You are asking me to do more design work for something that will be temporary. It takes more than 2x design effort to make something that looks good in two cases (alerted as new and not, and possibly removed if the feature dies).
Maybe you can show me a CSS example of something that looks good. Making something pop out and keep the whole page tasteful is hard. It takes time. If some small improvement can be completed in 2 minutes it doesn't make sense meditating for 10 hours on an ideal design to alert it (tastefully) and then figure out a tasteful way to add it to code without cluttering the code. Compared to a if(Date.now()<17650401401) { do extra backend stuff}, it's a completely different burden.
I did make an honest attempt at your idea at one point. But it's a no. It's not going to be a thing. But I guess if after my next feature announcement you make a theme for how it should look I can push it out.
Poll: how much wood could a woodchuck chuck if a woodchuck could chuck wood?