foobar20198

Article makes a false premise: that speed of writing the first version is important.

Much more time is spent maintaining software than writing the first version. Some poor sob is going to have to go and fix all those "silly" mistakes that were made and the bugs they caused (but of course, as we all know, they would have been avoided if the programmer had just been a better programmer). Save yourself (and us) the bother. Please use languages that stop you from shooting yourself in the foot.

(Not that rust will avoid bugs, but it seems to catch a lot of stupid "mistakes" before compile time. We're all humans and make mistakes, please accept that fact before inflicting software on us.)

My equivalent is to use incognito/private browsing (depending on browser of choice). However, once again, browsers are opinionated, and don't offer to save passwords in private/incognito mode (with no overrides). Which means I just avoid the whole experience when possible.

Similarly, things like Focus let you access a throwaway experience even more easily. Still no password saving though.

Switching between profiles in Firefox is the most painful experience ever. Running multiple Firefox profiles simultaneously requires using a terminal command.

Switching between profiles in Chrome involves clicking on an easy to find button, and lets you easily run multiple profiles in parallel. No terminal involved, accessible to anyone.

To be fair, running multiple chrome profiles in parallel kills my (somewhat old) machine (mostly by virtue of having too many tabs in each profile), so Firefox does win there.

To be completely fair: many projects at many companies turn out to be failures. Rust could've been bad and Firefox OS could have ended up ruling the world, and suddenly your comment would look very different. There's just a fundamental risk in projects, and viability often only becomes later in hindsight. (Most other companies failures just happen to be a lot less visible since they're not developing in the open.)

My comment wasn't specifically targeted at your blog post, more the general impression I've been getting of Mozilla. And sure, Mozilla is more open than Apple and MS, and probably a bit more open than Google.

But fundamentally, Firefox decisions are made by Mozilla employees, to further the goals of the Mozilla company (which makes money from those decisions - sure, to be hopefully reinvested). Firefox isn't like those other large open source projects where the codebase and thus decisions are owned by many independent (and mostly commercial) entities. Outside contributors have very little say in the product, so I wouldn't say its any better for people to contribute to Firefox vs e.g. Chrome. I.e. Mozilla wants their code - but in the many cases that have become publicised it doesn't seem like Mozilla wants their opinions.

(This is probably just the reality of end-user facing software projects with major UX components. But it's still sad to see contributors being invited in when they'll be ignored in the decision-making.)

// Edit: and it looks like fabrice_d made pretty much the same point while I was writing my comment : )

It seems a bit awkward to me when Mozilla keep asking for contributors. They're profiting from Firefox (via the huge search engine deal), it's in really poor taste when such organisations then beg for contributors. They have enough money to employ people to keep their product going... Kinda the same with e.g. Facebook's open source projects, although I guess there are worse places to contribute (and FB's projects might actually be useful for other companies work - hence contributing isn't a complete waste of time).

My only exception would be scratching my own itches, but here they're asking for help with company goals.

So does that mean Mozilla don't use 2FA?

Firefox does try to make it easy to build. However they're quite opinionated and want you to use a "bootstrap" script which downloads some stuff into ~/.mozbuild regardless of you potentially already having it.

It is however still very close to a clone & bootstrap & build process, at least on Linux and Mac - i.e. one of the better projects. (For comparison, I recently tried to build Facebook's Infer, and that was even worse than building OpenOffice back in the day.)