The recreation of existing capabilities on the web is really starting to irk me. None of these working groups collaborate, they all go off and invent their own thing, every time!
> We propose a new Integrity Block format for signing an entire Web Bundle. This is different from bundling Signed HTTP Exchanges because we don’t intend to create a verifiable mirror of a subset of a site’s resources, but a holistically verifiable version of an entire application.
Damnit. So because your goal is lesser you are going to ignore the existing work?
Then there's the fact that Signed HTTP exchanges also closely parallels various web archival formats. But the two groups once again had no interest in collaborating, were all but our use case for storing a copy of the site is different (so we'll do our own).
Nothing is iterated on, nothing gains and grows, because these endless working groups sprawl outward towards their own fringe.
My recent target of frustration is the Observable API, which is yet another brand new system for async activity. Closely mirroring web streams & async iteration helpers, except it has its own toolkit & is totally different from those! Why?! It's not even like it's some bumbling outsider just scratching their itch; there's a whose who of people who damned well know what else is afoot & what's available on the web already, and here they go recreating what we already have but in very subtly different tweaked form.
Generally I am so pro Fugu & pro the web. But there's seriously a lack of counterforce to prevent things from flying out in a million directions. There's so little binding & pulling things together, so little effort to garden & home our specs, and instead to just keep doing your version afresh. Integrity Block format feels like another classic NIH, another classic desire of everyone to see their special tree while denying the forest.
> We propose a new Integrity Block format for signing an entire Web Bundle. This is different from bundling Signed HTTP Exchanges because we don’t intend to create a verifiable mirror of a subset of a site’s resources, but a holistically verifiable version of an entire application.
Damnit. So because your goal is lesser you are going to ignore the existing work?
Then there's the fact that Signed HTTP exchanges also closely parallels various web archival formats. But the two groups once again had no interest in collaborating, were all but our use case for storing a copy of the site is different (so we'll do our own).
Nothing is iterated on, nothing gains and grows, because these endless working groups sprawl outward towards their own fringe.
My recent target of frustration is the Observable API, which is yet another brand new system for async activity. Closely mirroring web streams & async iteration helpers, except it has its own toolkit & is totally different from those! Why?! It's not even like it's some bumbling outsider just scratching their itch; there's a whose who of people who damned well know what else is afoot & what's available on the web already, and here they go recreating what we already have but in very subtly different tweaked form.
Generally I am so pro Fugu & pro the web. But there's seriously a lack of counterforce to prevent things from flying out in a million directions. There's so little binding & pulling things together, so little effort to garden & home our specs, and instead to just keep doing your version afresh. Integrity Block format feels like another classic NIH, another classic desire of everyone to see their special tree while denying the forest.