小程序服务器域名5次

by Howard Lo

霍华德·罗

为什么您不应该在100美元的服务器上用5天的时间构建面向500,000个用户的应用程序 (Why you shouldn’t build an app aimed at 500,000 users in 5 days on a $100 server)

A few days ago, I read Erik Duindam’s “How I built an app with 500,000 users in 5 days on a $100 server” article. Basically, he spent an extra 2–3 hours (totaling 24) to make his app (GoSnap) scalable by using a CDN and some optimized databases. He gives another app (GoChat) a hard time for building an app that had technical issues at launch.

几天前，我读了Erik Duindam的“我如何在100美元的服务器上用5天的时间构建500,000个用户的应用程序”一文。基本上，他通过使用CDN和一些优化的数据库花费了额外的2-3个小时(总共24个小时)使其应用程序(GoSnap)可扩展。他使另一个应用程序(GoChat)难以构建在发布时遇到技术问题的应用程序。

Those reading his article should take a few things into consideration:

那些阅读他的文章的人应该考虑以下几点：

他是一位优秀的开发商。 (He is a good developer.)

Much better than me.

比我好多了。

I doubt I could push whatever he did out in 24 hours. He’s quick, and knows what he’s doing.

我怀疑我能在24小时内将他所做的一切都推开。他很快，并且知道自己在做什么。

To him, 2–3 hours isn’t a big deal. You can shave that off your sleeping schedule and feel crappy for a day. But what if you’re not a strong developer? 24 hours turns into a week, and 2–3 hours can be nearly a full day.

对他来说，2到3个小时没什么大不了的。您可以将其从睡眠时间安排中剔除，并一天变得cr脚。但是，如果您不是一个强大的开发人员，该怎么办？ 24小时变成一周，而2–3小时可能几乎是整天。

If you’re racing to push something out because you’re worried someone else might have the same idea (more common than you think), a full day can make or break you.

如果您因为担心其他人可能有相同的想法(比您想像的更普遍)而竞相推出某些东西，那么整整一天都会使您成败。

Gauge this for yourself, but I’d rather be first with a 90% solution than second with 100% solution.

自己衡量一下，但我宁愿以90％的解决方案排在第一位，而不是以100％的解决方案排在第二位。

％失败>％成功 (%Failure > %Success)

Apps you build are far more likely to fail than they are to succeed.

您构建的应用失败的可能性远大于成功的可能性。

Erik speaks from a winner’s point of view. Winners speak a lot louder than losers, whose stories about their failures probably aren’t as impactful (or exciting) as Erik’s is.

埃里克(Erik)从赢家的角度讲。胜利者的声音要比失败者的声音大得多，失败者的失败故事可能不及埃里克的影响力(或令人振奋)。

If you do a quick search for Pokemon Go related apps, there are dozens of failed apps for every successful one.

如果您快速搜索与Pokemon Go相关的应用程序，那么每一个成功的应用程序都会有数十个失败的应用程序。

Barring massive partnerships (Google and Nintendo, in the case of Pokemon Go) the app store is essentially one giant numbers game. Yes, you can spend 10% more time to make your app scalable. But assuming your chance of failure is so high, in most cases that extra 10% of time is going to waste. If you spend 10% more time to make all your Minimum Viable Products (MVPs) scalable, but fail 9x, that’s one fewer MVP you could have built.

除非建立大规模的合作伙伴关系(例如Google和Nintendo，在《 Pokemon Go》中为Pokemon Go)，否则该应用程序商店实质上就是一个大型数字游戏。是的，您可以花费10％以上的时间来使您的应用程序可扩展。但是，假设您失败的可能性很高，那么在大多数情况下，将会浪费10％的时间。如果您将所有最低限度产品(MVP)的可扩展时间花费了10％以上，但失败了9倍，那么您可以构建的MVP就会减少一倍。

And are you really going to stop at 10%? Do you have so much self control and a birds eye view of what you’re doing?

而且您真的要停在10％的位置吗？您是否拥有如此多的自我控制能力，并且对正在做的事情有鸟瞰图？

For me, that 10% can become 20% really quick because “well, I put in a CDN on my app, why not put in caching, it’ll only take another 10%.”

对我来说，这10％真的可以很快变成20％，因为“好吧，我在应用程序中放入了CDN，为什么不进行缓存，只需要再花10％。”

Go minimum or go all out. Drawing the line anywhere else isn’t optimal.

尽量减少或全力以赴。在其他任何地方画线都不是最佳选择。

失败还算不错。 (Failing isn’t so bad.)

Erik himself updated his article a few days after:

几天后，埃里克本人更新了他的文章：

The Google Play page says [GoChat] “back 100%” with “over 2 million users”.

Google Play页面说[GoChat]“支持100％”，拥有“超过200万用户”。

GoChat launched an MVP, didn’t know they would get so much traction, and their crappy MVP code buckled under load. GoChat failed.

GoChat推出了MVP，但不知道他们会获得如此大的吸引力，他们cr脚的MVP代码在负载下也屈服了。 GoChat失败。

Or did it?

还是呢？

Because even though GoChat went down, it’s still getting more users than Erik’s GoSnap, which is built to scale. They are different products, sure, but could the extra 10% of time Erik put into GoSnap been better used to hit the app market 2–3 hours sooner? Could Erik be sitting on more users with the 2–3 hours head start and a technical crash after?

因为即使GoChat出现故障，它仍然比Erik的可扩展规模的GoSnap吸引更多的用户。当然，它们是不同的产品，但是是否可以将Erik在GoSnap中投入的额外10％的时间更早地将其用于应用市场2–3个小时？ Erik会在2到3个小时的启动时间之后又出现技术崩溃，从而吸引更多的用户吗？

Hard to say what the results could’ve been, but it’s interesting to think about.

很难说出结果是什么，但是考虑一下很有趣。

Even though GoChat’s code failed them, they still bounced back. They optimized in the second step, and still captured over 2 million users. The failure of non-scalable MVPs doesn’t seem to have hurt them all that much in the end.

即使GoChat的代码使它们失败，它们仍然会反弹。他们在第二步进行了优化，至今仍吸引了超过200万用户。最终，不可扩展的MVP的失败似乎并没有对他们造成太大的伤害。

Could GoChat have gotten more users by this point if they didn’t crash? Possibly. But they sure as hell didn’t miss the gravy train as Erik suggests, so what do you have to worry about?

如果他们没有崩溃，那么GoChat能否在这一点上吸引更多用户？可能吧但是他们确信地狱不会像埃里克(Erik)所说的那样错过肉汁火车，那么您有什么需要担心的呢？

It’s better to go ahead and launch than to miss the train entirely because you were too busy scaling.

最好不要完全错过火车，因为您太忙于扩展了。

The thing about “scale” is that at the upper end there are so many users to go around, you can afford to run into problems initially and still capture a good market share post-fix.

关于“规模”的事情是，高端用户太多了，您可以负担得起最初的问题，但在修复后仍然可以保持良好的市场份额。

At the lower end of “scale” there is nothing. You have nothing on your side but speed of execution. You have to get there fast. Just ask the dozens of other apps that aren’t being downloaded.

在“规模”的低端，什么都没有。除了执行速度，您什么都没有。您必须快速到达那里。只需询问其他数十个尚未下载的应用即可。

Self promotion section!I made rabbut.com, a tool that lets you collect emails here on Medium (and other places). Oh look, here is one now:

自我推广科！ 我制作了rabbut.com ，该工具可让您在Medium(和其他地方)的此处收集电子邮件。 哦，看，现在是一个：

Looking for my older stories? I’ve got some. Here.Looking for older stories is a PITA on Medium. Click here for a shortcut.powered.by.rabbut.com

寻找我的老故事？ 我有一些 这里。 寻找较旧的故事是中级的PITA。 单击此处获取快捷方式。 powered.by.rabbut.com

Also, I’m giving away my free eBook on starting up a startup. Especially good for people who don’t know how to startup a startup:

另外，我还将赠送免费的电子书，以帮助您启动创业公司。 对于不知道如何启动创业公司的人尤其有用：

First 10 people to subscribe get my free eBook.How to startup your startup as a nobody.powered.by.rabbut.com

前10位订阅者可获得我的免费电子书。 如何以无人启动您的启动。 powered.by.rabbut.com

Man, these rabbut things are like everywhere now. I wonder where you could get one…

伙计，这些拉布特的事现在到处都是。 我不知道你能在哪里买到……