3 ms·
It's my initial reaction too, but if I stop to think of Firefox as a huge, difficult-to-manage project where each feature costs dev resources and increases secu
by jszymborski 5y ago
It's my initial reaction too, but if I stop to think of Firefox as a huge, difficult-to-manage project where each feature costs dev resources and increases security risk, FTP support is probably the feature I wouldn't mind removed right next to the Pocket integration.
- handrous 5y agoOn the one hand I'm OK with removing features to keep the browser slim. OTOH Firefox is vastly fatter than it was in its early versions, which supported FTP (and Gopher!), so it doesn't seem like FTP or other minor protocols are an actual cause of complexity- and bloat-creep. [EDIT] I'd actually be a lot happier with this if FF were more aggressively adding new Internet or alternative-Web protocols, or trying to use their platform to give things like federated social networks more visibility and better prominence (via, say, optional but tastefully promoted integration with the browser). That might also, you know, differentiate them and give me a reason to recommend it to non-nerds again, despite "it's way, way faster than anything else, and lighter on resources, and blocks pop-ups and has tabs and doesn't have an ad banner!" no longer being enough (or even true, for some of those qualities) like it was in the early days.
- signal11 5y ago> OTOH Firefox is vastly fatter than it was in its early versions, which supported FTP (and Gopher!), so it doesn't seem like FTP or other minor protocols are an actual cause of complexity- and bloat-creep. I’m sure someone will correct me, but without looking at the source code, I suspect the following will have played a role in increasing code size since the earliest days of Phoenix: * support for newer media formats, from MP4 through AVIF * DRM maybe? * WebGL * SVG? * The complexity of modern CSS
- handrous 5y agoMuch, much bigger JS engine/sandbox, probably. Backwards compat with an ever-growing set of HTML, CSS, and JS. I'd guess a modern CSS engine is several times (dozens of times?) larger than what Phoenix shipped with. I expect their internal UI-drawing code/framework is a lot larger than in the Phoenix days. Webassembly. Canvas. And yeah, several of the other things you mentioned—video used to be the responsibility of plugins, for example. [EDIT] in fact, some things browsers do now were seen as terrible ideas when Phoenix was released, because they would kill performance, since they're so much more expensive than alternatives. Vector graphics were regarded as to-be-avoided for a long time, because they're so much more taxing on the client than raster. CSS doing animations would have been considered insane, fit only for "look how I can bring your machine to a crawl with a web page" demos. Maintaining an entire "shadow DOM" of Javascript objects would have been regarded as parody-levels of resource wasting for all but very niche use-cases. No wonder the modern web is so slow and resource-hungry....
- hutzlibu 5y agoOr Wasm, or WebRTC, or indexedDB ... plus anything else on that long list: https://developer.mozilla.org/en-US/docs/Web/API https://developer.mozilla.org/en-US/docs/Web/API
- spiritplumber 5y agoFirefox supporting Gemini would be awesome.
- sambe 5y agoI don't think my estimate of the cost of FTP maintenance would be as high as yours. Why do you think FTP would still be such a cost at this stage?
- asoneth 5y agoI'm curious what your estimate would be. Mine would be somewhere between 0.5FTE assuming the bare minimum of keep-the-lights-on QA testing for each release and ongoing bug triage up to maybe 3FTE if it requires periodic security and design updates to keep up with changing standards/platforms/protocols.
- coffeeling 5y agoThe Pocket integration should be about the last thing removed: Independent revenue streams are something Firefox desperately needs.