Anyone know what happened to Intel over the past 5+ years? They used to be unstoppable, relentless, the best. Have they had trouble recruiting the best engineers? How have TSMC and Apple been been able to pull ahead in terms of being at the lead of CPU development?
I thought maybe Intel was losing ground to TSMC for cultural reasons – being the best while being based in the US is getting harder and harder. But I'm not sure.
I'm the author of that Medium article. The "second one" you referred to isn't the second bug I reported – it's just an aside on Wired's crazy headline font. They've since replaced it with a normal readable font...
Faster still would be to skip the request for the CSS altogether by inlining it in the head. Therefore you'd just request the font immediately and directly. This makes the preconnect moot, and should be faster.
Faster still is to use already installed fonts, as others have noted. But don't use "system" fonts if that means UI fonts like San Francisco or Segue. Exploit the rich font options on the desktop: Georgia, Sitka Text (Windows 8.1+), Calibri, Hoef, etc. Bootstrap and Medium started a ridiculous fad of using system UI fonts for body text. There are actual body text fonts available people.
Modern browsers are smart enough to only download the required file type, sniffing is unnecessary the vast majority of the time and you save an entire 3rd party connection.
The tool you linked does not solve the same problem, as the problems isn’t file types but URIs. Google is able to provide the tiniest possibly font files by hinting them for the specific browser/platform requesting them. We never consistently or predictably know the resultant URI of a Google Fonts-hosted font file. Here is a Tweet to the same effect from two days ago. Believe me, I’ve researched this :(
The fonts are static, but the CSS (from fonts.googleapis.com), and the font files it references, vary depending on the User-Agent. For instance, with a Chrome UA it will request WOFF files; with a blank UA it may request TTF files.
The idea is that you take all the links to the different font types and put them directly in your local CSS. Your browser knows which formats it supports, if you have TTF and WOFF2 fonts linked, Chrome will only download the WOFF2 version. No need to send an extra 3rd party request just to get a piece of CSS that only links a single file type.
Well yes, of course you can do that. But at that point, why would you reference third-party font URLs, instead of self-hosting them? (The article pretty much begins by stating that this is the fastest option.)
If performance is your utmost priority, forego web fonts entirely. If you value performance but want the benefit of a web font, self-host your CSS and fonts. If the convenience of Google Fonts is appealing, the suggestions here can improve first-paint performance. Different projects will have different trade-offs.
You could use a couple microseconds on the server to inline CSS to match the HTTP U-A; assuming whatever you needed to know is still in there by the grace of Google whenever they slush up the U-A strings.
Stability of the physical media aside, keeping anything you intend to ever read again on an obsolete storage format makes no sense when you consider that the number of working readers/players in the world decreases with time. The only reason I have my Apple II-era data is because I keep kicking it onto newer storage formats.
The potential red flag in this account was Jared interrupting the interviewer who was asking a question that mentioned microservices to say that he had no experience in microservices.
Why did he interrupt the interviewer? And why interject that he had no experience in microservices? So what? You can still tackle a question that mentions microservices...
It's not like one's programming skills are useless for questions that merely mention an architecture you haven't officially worked with before. It's very strange to me that he interrupted that way, for that reason, and it makes me wonder if he acted similarly in other interviews. If his attitude is that he shouldn't have to answer questions about architectures and technologies not specified in his resume, that wouldn't go well.
Also, there's a lot of hype and looseness around the term "microservices" these days. You might have worked with what some people call microservices without knowing it. All the more reason not to cut off the question.
The reason governments waste time on recycling, or fail to authorize landfills, or orchestrate the export of trash to other countries is because environmentalists exist and are very active. They're especially powerful in rich countries – in fact, environmentalism seems to be a function of affluence. Affluence looks like a necessary but not sufficient condition, loosely speaking.
Interesting read. It always surprises me that companies go many years running major apps and infrastructure with interpreter-based languages like Python and Ruby to begin with. It's an incredible waste of energy, compute, and for web apps sometimes, users' time.
Developers need to be a lot more disciplined about performance and efficiency. I'm glad Khan went to Go, but man all those years wasted.
And what exactly is wrong with their business model? That's not clear to me from the article. It says they take long-term leases and sell short-term leases. That seems straightforward and potentially lucrative if they're adding value, making it easier to find suitable short-term office space, etc. Is it just that they're not making money doing this? Why not?
I think the reason you don't get an answer to this is because this is article number 100352752. WeWork isn't a tech company because there's no unique technology they use to rent out office space- which is how they make money. Their business model is "Sign long term contracts cheap, rent out to short term tenants expensive". Which is fine, but it doesn't justify their valuation (which is predicated on them 'disrupting' the office rental industry) and isn't fine - because so far they spend more money renting the office space than they do leasing it out to their tenants. So their real business model is "Rent $1 worth of office space out for 50 cents". They had no compelling path to profitability.
Airbnb's business is the app that connects property owners to renters. Their app is the technology, their marginal costs are 0- you can add another property to the website trivially. Their value is in the network effects in their app just like facebook. Wework actually sign rental contracts and take on the risk of not being able to fill the properties.
They are considered a tech company (by the people who consider them a tech company) because they say they're a tech company. And they say that because tech companies get higher valuations.
> And what exactly is wrong with their business model? [...] It says they take long-term leases and sell short-term leases.
Right. So in the good times, you can make a fair bit of money doing this, but in bad times you'll lose a lot of money, as all the short term leases drop off.
Traditionally companies in this space work around this by buying their own buildings, by having fat margins in the good times, by not making very much money, and by occasionally going bankrupt. :) It's a perfectly valid business model, but it's not super attractive to investors; it's capital intensive and not very profitable. IWC (formerly known as Regus) has about 5 times the locations WeWork does, and is worth about $3B. $3B is a lot of money, so there's nothing wrong with the business model, but it's also a lot less than the $47B people were talking about for WeWork until very recently.
WeWork is trying to find a way to do something different or cooler, in the hopes of finding a reason they should be values completely differently than IWC, but so far that's translated into renting buildings, thin margins, no profits, and no story for why what they're doing is fundamentally better or more valuable than what IWC is doing. Yes, they have a live DJ in one of their London buildings and IWC does not, but IWC is making money and WeWork is not, and I mean...you can hire a DJ if that turns out to be the critical feature missing from other office space.
> That seems straightforward and potentially lucrative if they're adding value
Right. Especially now that Adam Neumann has stepped down, there's no reason WeWork can't find their niche as another IWC. The problem is, as another IWC, there's no real reason they'll be worth even $10B (again, IWC is much larger and only worth $3B), and Softbank backed them at much higher valuations. Just because WeWork has a profitable niche doesn't mean their investors will be okay. For Softbank the difference between a $5B valuation and a $0B valuation is minimal, when they were expecting something more like $60B.
> It says they take long-term leases and sell short-term leases. That seems straightforward and potentially lucrative if they're adding value, making it easier to find suitable short-term office space, etc.
When times get tough the startups go back to their parent's garages or basements and don't renew their short term leases. WeWork is stuck with their long-term leases.
Conversely while times are good startups are willing to pay premium for short term leases, either to expand quickly if times stay good or to go back to some garage if times get tough. The business model is essentialy similar to an insurance, they make money by taking on other people's risks.
Of course WeWork doesn't operate financially like this. The business model is sound in principle, but WeWork's execution seems ill conceived (except for the personal enrichment of the CEO).
I like it. What would also be cool is another Go compiler and linker, developed by a different team not at Google. It would be good for the ecosystem and performance if there was more than one compiler and runtime.
I think there would be a market for a proprietary compiler and maybe an IDE to go with it — if the performance was better than the open source one. I think this is achievable because as good as Go's performance is now, there's still a lot of headroom. Google isn't exploiting modern CPUs very well, and the linker is not doing extensive LTO.
The biggest constraint is the blazing fast compile times. A compiler and toolchain that was able to take some time for optimization might deliver markedly better runtime performance.
Curious enough; we're working on just that, a commercial compiler and ide for Go. A go compiler that can compile for .NET, native Windows, native Linux and native OSX. Haven't done much performance analysis yet, the IDE and and compiler are a first priority now, and are almost done.
> I think there would be a market for a proprietary compiler and maybe an IDE to go with it
There may be a market but not large enough to pay few top notch compiler/low level system software hacker. I only know of a commercial Go IDE and constant refrain there is how will a poor third world developer afford it. Though I feel real issue is developers are raised on diet of free software feels entitled to it. Paying for good software seems alien to them.
Software licenses don't really matter if you implement the Go language spec with your own code.
It's only when you want to share code with other implementations that software licensing enters the picture. Gccgo manages to share most of the standard library with the golang.org implementation, so they appear to have the licensing figured out.
I thought maybe Intel was losing ground to TSMC for cultural reasons – being the best while being based in the US is getting harder and harder. But I'm not sure.