Sure, there are plenty of reasons of the whys and whens - on the other hand, we could easily ask "why didn't everyone choose Emacs when textmate's development came to a halt" or "Why didn't people choose Python as their web language of choice when they didn't like PHP" or "Why not Bazaar instead of SVN" or why Ryan Dahl chose JavaScript as a base language for Node.js and not any of Python, Ruby, Perl, Scheme, Lisp or Lua - but JavaScript - and so on. (As others already pointed out below so I'm summing this up in one sentence here..) Thankfully we have plenty of options to choose from nowadays.
And yet I'm still convinced that the choice of Git over Bazaar has something to do with the smoothness of Github and with "It's from Linus!", the choice of Ruby/RoR with the image 37signals so nicely projects, that JavaScript's recent rise and massive change in perception comes thanks to Douglas Crockford and Erlang would have probably stayed in its niche if it wasn't for CouchDB.
So I was surprised and amazed by the change in perception of Vim over the last year and therefore I blogged about it.
And yes, along your examples Perl faced a similar situation around 2000 (so thanks for the well chosen examples) and decided to do Perl 6 to re-ignite the Perl spark. (Let's set aside for a moment wether or not it worked and what happened after that and boy am I sure the second I hit the submit button people will _exactly_ totally get into this subject.. ;)
I also didn't ask wether or not one really wants the success of the masses or if it might be a good thing to stay in a well-defined niche with a community of your choice, creating your own culture - as for example shows the Linux distribution Slackware very happily year by year and to a great satisfaction of its users.
I also didn't mention how much it might have to do with the age of developers, wether some changes plainly might be a generation thing of "first generation web developers" and "second generation web developers" or how much Apple's regained success does play into all that.
But as we can see within the comments below, old Perl cliches aren't really dead and get repeated all over wether or not the subject was Perl's marketing and not Perl's qualities (or the perceived lack thereof...)
Or maybe we all get kicked our asses by Lisp next year - thanks to Peter Seibel's Practical Lisp Programming book or "Land of Lisp" and of course Paul Graham and Emacs wins all over. ;)
Predictably Irrational talks about what you are describing here. You see, despite many of these technologies having a smaller install base, what they have done is create a new anchor in technology, which allows them to monopolize the conversation. The iPhone and Rails are the most significant anchors within the past 5 years. You can't make Perl popular by acting like rock stars and making videos on youtube, you can only making Perl popular by creating a new technological anchor for others to hook themselves to.
I chose emacs when Textmate stopped getting updated. The joke page registered at textmate2.com redirects to emacs, and a most of the people harassing Allan on his blog a couple years ago were saying they'd switched to emacs.
I think it might just be that emacs doesn't lend itself to screencasts or golf quite as much as vim.
I could lower the lid on my macbook so the webcam is pointing at my fingers to show the chording. Hmm, yeah, probably not so helpful.
Emacs has also had a really strong community for over 30 years (perhaps driven somewhat by the higher education roots). As such, I think it had a more firm underpinning than vim did, so more of the tutorials and such already existed.
I think there still was one good point in the parent comment that you did not address. If screencast were not the reason of RoR popularity but it's consequence, then maybe we should not lose too much time on making screencasts and instead concentrate on some other things? I have to attest that screencasting does require enormous amounts of time, especially if you try to do that on Linux where the tools break every 5 minutes and you need to make manual savepoints to keep with moving ahead with your work. Even though it was also very captivating and quite entertaining.
I may be a rare case, but I honestly started using git and vim in particular long before the heaps of hype that built up around the tools, before I even started reading HN actually.
I found the tools I went with to be more intuitive and easier to learn than the others. When tools are legitimately pleasurable to use, developers talk about them, and 'hype' builds up.
The hype then attracts the next generation of users and perpetuates the trend. Most users are pulled in by the hype, rather than the rational decision making of the early adopters.
And yet I'm still convinced that the choice of Git over Bazaar has something to do with the smoothness of Github and with "It's from Linus!", the choice of Ruby/RoR with the image 37signals so nicely projects, that JavaScript's recent rise and massive change in perception comes thanks to Douglas Crockford and Erlang would have probably stayed in its niche if it wasn't for CouchDB.
So I was surprised and amazed by the change in perception of Vim over the last year and therefore I blogged about it.
And yes, along your examples Perl faced a similar situation around 2000 (so thanks for the well chosen examples) and decided to do Perl 6 to re-ignite the Perl spark. (Let's set aside for a moment wether or not it worked and what happened after that and boy am I sure the second I hit the submit button people will _exactly_ totally get into this subject.. ;)
I also didn't ask wether or not one really wants the success of the masses or if it might be a good thing to stay in a well-defined niche with a community of your choice, creating your own culture - as for example shows the Linux distribution Slackware very happily year by year and to a great satisfaction of its users.
I also didn't mention how much it might have to do with the age of developers, wether some changes plainly might be a generation thing of "first generation web developers" and "second generation web developers" or how much Apple's regained success does play into all that.
But as we can see within the comments below, old Perl cliches aren't really dead and get repeated all over wether or not the subject was Perl's marketing and not Perl's qualities (or the perceived lack thereof...)
Or maybe we all get kicked our asses by Lisp next year - thanks to Peter Seibel's Practical Lisp Programming book or "Land of Lisp" and of course Paul Graham and Emacs wins all over. ;)