幕后丨Castro 2 开发者采访英文原稿

nerd_du nerd_du 2016-08-26

前不久我们采访了 Castro 2 的开发团队 Supertop,以下是采访的英文原稿,你可以点此去网站阅读翻译后的完整文章


Supertop is an indie iOS and Mac software company run by Pádraig Ó Cinnéide and Oisín Prendiville. We met in Dublin City University in 2005 and have been working together for over 5 years. Pádraig resides in Vancouver, Canada and Oisín is currently in Sarajevo, Bosnia and Herzegovina.

We both work full time at Supertop. In general we both do everything, but Oisin tends to focus more on design/motion than Pádraig who prefers to work on servers and performance.


We use Slack as a messaging app. This lets us keep all most of our communication in one place that we can switch off when we’re not working.

Oisin is in the habit of working late, so usually there’s a good overlap in the morning and afternoon with Padraig, so we’re able to catch up. We see each other about twice per year, and we usually get a lot more done then. It’s not ideal to be so far apart but Padraig loves Vancouver and Oisin loves travelling, so we make it work!

π:为什么 Castro 2 开发了 22 个月?是遇到了什么困难吗?

That is a long story! I would say that it breaks down into a few different reasons.

  • My wife and I (Pádraig) had a baby last year and that was a big adjustment. I don’t have as much time for coding as I used to.
  • We fell victim to a phenomenon that called second system syndrome (https://en.wikipedia.org/wiki/Second-systemeffect). We built Castro 1 under very tight time and resource constraints, so when it was time to build 2.0 we gave ourselves much more space to try to do everything. This caused us to lose focus a bit and not prioritize as much as we should have. Early in 2016 we realized our mistake and started to manage our time better.
  • Maybe the biggest thing though is that we wanted to make something original. There are 100s of podcast apps out there that follow the same basic pattern but we had a new idea. We wanted to make it easier to subscribe to a lot of feeds and pick out the best episodes. It’s hard to refine a new idea, and to build an interface that teaches users how it works. It took a lot of iteration. The app has changed layout completely several times during development.

π:为什么会想到使用类似 GTD(或者说邮件)的管理方式?对于这个功能,你们有哪些考虑?

This is the big idea of the app. We feel that since podcasting is growing so fast, it’s no longer enough to just subscribe to 5-6 feeds and listen to every episode of shows you’re committed to. You miss great episodes if you do that. There are so many amazing stories, different perspectives on politics and diverse voices available now. So we wanted to make an app that lets you subscribe to 100+ feeds, quickly pick out the best episodes and easily skip the rest.

π:从开发周期来看,Castro 2 的设计风格是早于 iOS 10 的,但目前看来它和 iOS 10 非常契合,这是巧合吗?

That really is a coincidence! We can show screenshots from before iOS 10 was announced where we have this card approach. It was nice to see Apple do it, it felt like a vindication of the design direction. Once small reason that we like the card is that we can leave the status bar outside the app. It’s very minor but managing the light/dark status bar seamlessly through transitions was always annoying and almost impossible to get perfect.

π:Castro 2 的设计属于「看截图很平淡,用起来有惊喜」,因为应用中有不少精致的动效,但也有用户反映应用不太流畅,这是否和动效有关?

We completely agree — when we were planning our teaser site before launching, it was really frustrating because, exactly as you said, the still screenshots aren’t very glamorous but the actual interactions are nice.

Since we’re introducing a new idea to podcasts, the animations of episodes dropping into the tabs as you queue or archive them are vital to teaching users how things work. They could seem like unnecessary flourishes, but in our testing, users had no idea where an episode went without them.

We’ll improve performance issues over time, again it was a matter of getting the app to a good enough state that we could release it.

π:为什么 Castro 2 里没有智能加速功能?

It’s a good feature. I don’t personally use it because I prefer well edited podcasts where the pacing might be important or where the pauses give me a chance to digest the story better. However, it’s the single biggest feature that users have mentioned since we launched. As I said before, we are shy about promising features or schedules, but I can say that we understand the value of it completely and we feel a lot of pressure to add it.

π:用户呼声很高的同步和 iPad 通用版在计划之中了吗?

We’ve had a lot of requests for this, and it’d be exciting to work on. After 22 months of development, we’re a little shy about promising any kind of schedule though.

π:在 iOS 10 大量使用 3D Touch 操作的现在,Castro 2 没有 3D Touch 特性是基于什么原因?

We like 3D Touch and we want to add it, it was just a matter of priorities.

π:我们对 Tentacles 服务器很感兴趣,请问你们为什么想要自建这样一个服务器?未来还打算用它做什么?

Tentacles is our central feed aggregation server. We built it to keep all podcasts up to date so that the app wouldn’t have to use battery/network on your phone while scanning for episodes itself. It gives us a few nice features; one great thing is that when there are any issues with a feed, we can fix them instantly, without an app update.

In future, it could allow us to support additional discovery features, sync, much better search and sharing.

π:Overcast 有基于 Twitter 的推荐,Castro 2 也有了自己服务器,并且支持了分享链接,看起来以后可能会有社交元素?

We wanted to place more emphasis on episodes, and in particular sharing episodes with friends or on social media. It’s much easier to say what’s good about a single episode than an overall podcast, and its much lower commitment to say “ok, I’ll queue one episode and see what it’s like”.

Castro provides https://castro.fm/episode/z5lkHk links which you can scroll up in Safari on iPhone and then tap “Open In Castro” to immediately open the episode. If your recipient doesn’t have Castro yet, they can use the web player.

π:在 Overcast 宣布免费时,你们有说「并不担心 Castro 2 的竞争力」。但按目前版本来看,对于非重度播客用户来说,Overcast 仍是更有性价比的选择,对此你们怎么看?

There’s certainly a lot of competition in the podcast market but if you look at the charts, there are a few apps that manage to make enough money each day despite Overcast being free. If users are deciding entirely based on price, then they will always choose Overcast. If you want to listen to a lot of podcasts, you’re probably going to use your podcast app more than any other app on your phone. It makes sense to invest a bit of money in making that experience better. Castro 2’s triage features are unique, and will let you expand your subscriptions list.

π:听说之前的 TestFlight 版本采用了订阅制,为什么最后上架决定改为一次付费的买断形式?

We tried a patronage model for a while in Castro 1, where the app was completely free and users could choose to pay $1 per month to support the app without getting any additional features. A few passionate users were happy to pay but not enough to make it a feasible revenue model for Castro 2.

π:随着 App Store 对订阅制的优惠政策,相信会有越来越多 App 转为订阅模式,但不少用户并不买单,一方面是一些开发商策略做得不好,另一方面是认为某些 App 不应该订阅。你们怎么看?

We’re very interested to see how the market evolves. Apple’s new policy opens the door for potentially seeing this used in a lot more apps but customers will ultimately decide what they’re willing to pay for and it’s up to developers to make the case for why their app is worth it. It will be interesting to see if Apple take any steps to help make subscription pricing more palatable to customers.

If embraced, it would certainly be a much more sustainable model for software development that would help developers continue to improve their apps and serve the needs of users for longer, but there are hurdles to getting there for sure.

π:你们做的 Castro、Unread 其实都是挺小众的市场,这次 Castro 2 开发周期这么长,推出之后万一不叫座,对你们来说会存在不小的风险,是否考虑过未来缩短开发周期?

Spending 22 months on Castro 2 may not have been a pragmatic business decision but it’s emblematic of our goal of being able to work on the things that we want to work on. We were trying to get it right rather than just get it done and so concerns about cost probably took a back seat for quite a while. We’re not trying to build a huge business, or even necassarily make as much money as we could if we did contract work or got jobs in bigger companies, but yes we’re acutely aware of the risks and will do everything we can to mitigate them.