TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Speed as a Habit

136 点作者 kanamekun将近 10 年前

24 条评论

calinet6将近 10 年前
I take offense to <i>idea</i> of speed as a priority, though not necessarily to the whole article.<p>&gt; All else being equal, the fastest company in any market will win.<p>All else is never equal when speed is the driving factor. I have <i>never</i> seen a company execute well while pressuring for speed.<p>It&#x27;s significantly more important to think about systems. What is the cooperative source of your speed? What are the barriers to making good product quickly? What allows us to make faster decisions?<p>Invariably, without fail, unequivocally, the answers to these questions—which are sources of speed, quality, and market fit—lie in the systems that compose a company. Not systems as in slow methodical heavy processes, but systems as in a clear understanding of how things are connected and how to continually improve how they work.<p>Start with systems, then you can go fast. It&#x27;s ludicrously dangerous to put the cart before the horse. Granted, much of this article dives into detail on how to make things run smoother in order to go fast—that&#x27;s great, but if you make it the primary focus, imagine how fast you could go.<p>Speed... a noble goal, but as W. Edwards Deming said, &quot;By what method?&quot;
评论 #9926900 未加载
评论 #9925585 未加载
评论 #9925939 未加载
评论 #9926339 未加载
评论 #9927451 未加载
bsder将近 10 年前
&gt; Speed is a defining characteristic — if not the defining characteristic — of the leader in virtually every industry you look at.<p>Not even close.<p>Apple was <i>WAY</i> late to the MP3 player party. AltaVista was way ahead of Google, but DEC marketing (&quot;DEC has it now, but you can&#x27;t have it.&quot;) couldn&#x27;t sell water to a man dying of thirst in the Sahara. Android had far better and faster competitors, but it brushed them aside.<p>In addition it is <i>well</i> documented that the first company in a category takes all the arrows while the second company makes all the profit.<p>Things rarely need to be done that fast.
评论 #9925418 未加载
评论 #9926914 未加载
评论 #9927236 未加载
评论 #9927587 未加载
评论 #9927788 未加载
Animats将近 10 年前
He quotes Patton. It&#x27;s worth remembering that Patton was not in overall charge of the war in Europe. Eisenhower was. Eisenhower, who was a logistics officer by training and background, had a basic plan for winning in Europe - build up a mountain of resources so large that the Allies could roll over any possible opposition. Which is what happened.<p>Eisenhower put Patton into play when a part of the operation needed speed and violent execution. Eisenhower took Patton out of play when it didn&#x27;t.
评论 #9928728 未加载
评论 #9927588 未加载
dbg31415将近 10 年前
Well, so I can see this article being twisted all over the place.<p>It&#x27;s hard enough to get proper time for planning technical projects without this kind of crap.<p>Guess what, planning time does save time and lowers your cost of ownership. ROI-wise, planning time is the best gift you can give your project.<p>But... then someone is like, &quot;Speed, I hear speed is good... move faster. Dependencies... fuck &#x27;em... I don&#x27;t understand &#x27;em anyway so fuck &#x27;em. Let&#x27;s just go as fast as we can and we&#x27;ll worry about everything we should have worried about up front over the next few years after all our devs quit because they&#x27;re tired of working on shitty code that was rushed through.&quot;<p>This article so rubs me the wrong way. &quot;Technical debt? What&#x27;s that?!&quot;<p>Anything worth doing, is worth doing right. To the best of your ability. Doing things right takes time. Take your time, do things right.
评论 #9925164 未加载
评论 #9925009 未加载
评论 #9925717 未加载
评论 #9925069 未加载
评论 #9925203 未加载
评论 #9926385 未加载
rcarrigan87将近 10 年前
Surprising number of people on this thread are conflating first mover advantage with moving fast.<p>First mover just means you are first to a new market. Hypothetically, there could have been 10 years of product development behind your first mover advantage.<p>Moving fast, as I see it, is not becoming that huge, slow organization (think GE), that has so much bureaucracy baked-in nothing ever gets done.<p>Most decisions really don&#x27;t require the level of debate they&#x27;re given. The author is pointing out that a lot of the decision making process is influenced by shit that doesn&#x27;t have anything to do with the actual decision.
评论 #9927507 未加载
InclinedPlane将近 10 年前
During the Vietnam war the US began to look into the unexpected underperformance of its pilots in air-to-air combat. They weren&#x27;t exactly losing but they weren&#x27;t doing nearly as well as they were expected to, and there was a litany of mistakes being made again and again that were hurting combat performance, all of which came to light in the &quot;Ault Report&quot;. A lot of research went into studying what was happening and why and ultimately a new theory of aerial combat fundamentals came out of that research.<p>The main difference that was causing problems for American pilots was that the planes they were fighting against were more maneuverable. It turns out that greater maneuverability naturally led to a faster and more responsive cadence in combat, leading to MiG pilots gaining the upper hand in encounters.<p>The two theories that came out of this study were Energy-Maneuverability Theory and the idea of the &quot;OODA loop&quot;. As it happens, OODA loop theory is generally applicable in many situations outside of aerial combat.<p>OODA is an initialism standing for: Observe, Orient, Decide, Act. It is the fundamental event processing loop for &quot;doing stuff&quot; in almost any situation. The faster one&#x27;s OODA loop the faster one can respond to changing situations, which translates into huge practical advantages. And the amazing thing is that this is somewhat independent of maneuverability.<p>OODA loop theory should be well understood by anyone in the tech industry. It&#x27;s a critical element for the competitive advantages of small startups over big, entrenched competitors. And it&#x27;s the reason why agile methodologies are so often advantageous.<p>&quot;Speed&quot; is a crude substitute for understanding the value of feedback-based iteration and the competitive advantage of tighter OODA loop cycles.<p>P.S. Raw speed is not a substitute for good design, thoughtful engineering, or keeping technical debt under control. It is quite possible to speed yourself right into a wall where you lose agility because of excessive technical debt. Also consider the relative merits of building on good foundations, increasing value exponentially over time versus constantly tearing down and rebuilding disposable structures. Don&#x27;t confuse busy work with craftsmanship, value doesn&#x27;t scale linearly with effort.
评论 #9927525 未加载
评论 #9929642 未加载
panic将近 10 年前
This exact attitude is the reason why modern software is so bad. Dates are set before anyone has any idea how long the software will actually take to design. Then, when it turns out the dates were too aggressive, the design is compromised so the software can be shipped faster.<p>That&#x27;s not to say the article isn&#x27;t right. Maybe speed really does win. But if that&#x27;s true, we might be stuck with terrible software forever.
yoklov将近 10 年前
&gt; &quot;You know you&#x27;re going fast enough if there&#x27;s a low-level discomfort, people feeling stretched. But if you&#x27;re going too fast, you&#x27;ll see it on their faces, and that&#x27;s important to spot too&quot;<p>Most people can&#x27;t tell the difference here, and failing to do that is deadly. You can lose a lot of good engineers by forcing too much crunch, and that can kill or significantly delay projects.<p>This happens not rarely in game development.
DrBazza将近 10 年前
&quot;A good plan violently executed now is better than a perfect plan next week.&quot; General Patton. Because software development is so much like dropping as many bombs onto a bunch of people as possible.<p>&quot;I’ve long believed that speed is the ultimate weapon in business. All else being equal, the fastest company in any market will win.&quot;. The irony that the authors two previous employers were Google and Apple. Both of which are massive companies that were (and are) famously never first to market with anything.<p>&quot;All things being equal&quot;: it&#x27;s the company with the deepest pockets and the best PR. Which is not equal.
logicallee将近 10 年前
&gt;All else being equal, the fastest company in any market will win.<p>All else being equal, the best-funded company will win. All else being equal, the company with the catchiest name will win. All else being equal, the company with the best product will win. All else being equal, the company with the best-advertised product will win. All else being equal, the company with the most stories being written about it will win. All else being equal, the company with the better cafeteria will win. etc.<p>all true statements. don&#x27;t mean much though...
Smushman将近 10 年前
&gt; All else being equal, the fastest company in any market will win.<p>Yea, no doubt. Don&#x27;t even need an MBA to see that. If you have two equal competitors, the first one to market has a useful head start.<p>I think this article is what I would call aggressive trolling; scoring points by &#x27;thinking different&#x27;. But essentially this is not more than standard knowledge, repackaged as a hype laden pseudo-intellectual conversation meant to score points.
评论 #9926917 未加载
AnimalMuppet将近 10 年前
See <a href="http:&#x2F;&#x2F;integratedleader.com&#x2F;articles&#x2F;40-70rule.pdf" rel="nofollow">http:&#x2F;&#x2F;integratedleader.com&#x2F;articles&#x2F;40-70rule.pdf</a> for another take on this. (At least, I think it&#x27;s the same thing - we want certainty, and that causes us to take far too long to make decisions.)
sparkzilla将近 10 年前
&quot;Move fast and break things&quot; is a way for immature and arrogant people to mask the guilt they should feel for offloading their poor decisions and planning to the end user.
评论 #9926861 未加载
somberi将近 10 年前
&quot;There is more to life than increasing its speed.&quot; - Mahatma Gandhi<p>Also a commonly cited zen story:<p>Eager Student asked: “How long will it take me to be enlightened?”<p>“Ten years” said the master.<p>“What if I work twice as hard?”<p>“Twenty years” said the master.
评论 #9928185 未加载
0xdeadbeefbabe将近 10 年前
&gt; There are decisions that deserve days of debate and analysis, but the vast majority aren’t worth more than 10 minutes.<p>Quick! I need a vast majority detector.<p>Setting a deadline for deciding something sounds good to me, but the above is some kind of puzzle.
评论 #9925676 未加载
评论 #9925175 未加载
omouse将近 10 年前
To the nay-sayers on speed as priority:<p>I&#x27;m currently on a two-week deployment cycle. It&#x27;s not only annoying to wait but it&#x27;s also <i>dangerous</i>. There have been several incidents where the code that broke something was committed a whole week or two weeks before! Worse still is that we&#x27;re deploying two systems at once and usually we&#x27;re patching the systems at the same time. So you have a perfect storm of bad things that can happen with multiple causes. Faster means you see the problems sooner.
评论 #9926137 未加载
mizzao将近 10 年前
&gt; There’s not always a stark tradeoff between something done fast and done well.<p>Yes, there is: <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Project_management_triangle" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Project_management_triangle</a>, and to have both requires more money (resources).<p>This article also doesn&#x27;t seem to have been written by an engineer. I&#x27;d love to get a more technical perspective on how to move fast.
hamxiaoz将近 10 年前
On a side note, the hero photo looks realistic (has the &#x27;pop&#x27; feeling), wondering how to produce this kind of photo?
评论 #9926924 未加载
skybrian将近 10 年前
The answer as usual is &quot;it depends&quot;. At the very least it depends on the consequences of missing a deadline and how painful &quot;undo&quot; is if you screw it up. A software update is much easier than a hardware recall. Formal processes are necessary when undo becomes painful enough.
pbreit将近 10 年前
I was prepared to disagree but, you know what? He&#x27;s right. Very few decisions are irreversible or benefit from extended debate or analysis. Some keys are that you have to be prepared to undo and not crap on the decision-maker.
maxxxxx将近 10 年前
I hate it when people narrow things down to one factor. Yes, speed can be helpful, but you still need to balance it with other factors like quality. Speed for its own sake can kill you quickly.<p>In short, you need to know what you are doing and why.
itistoday2将近 10 年前
I remember how many <i>_years_</i> it took Apple to figure out how to do right-click properly while their competitors were speeding along with that silly second button.<p>Speed is important. Patience is important. Stepping back and smelling the flowers is important. It&#x27;s not about fast or slow. It&#x27;s about right, and you can&#x27;t be right all the time if you&#x27;re stuck in the habit of zooming along all the time.
评论 #9927517 未加载
michaelochurch将近 10 年前
Am I being asked to take seriously the company that produced this atrocity? <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=bMJIBxtDUHc" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=bMJIBxtDUHc</a>
评论 #9931948 未加载
rglover将近 10 年前
Wheeee!<p><a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=m4UPvorv2qY" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=m4UPvorv2qY</a>