A Year to The Day | Hacker News Reader

A Year to The Day

blog.ghost.org

81 pointsNaushad52 comments

Is there anything about Ghost that, intrinsically, makes it better than other blogging platforms?

Seems to me it's just reinventing a many times reinvented wheel (except this one is hip and popular).

It's main selling point is that it's fast and uses Markdown for writing posts. It's perfect for a developer blog.

For a developer blog, what would be the advantage of Ghost over a static site generator?

I can't see any reason to use it for any type of blog. Setting up ghost on your server and getting used to the UI (yes, it's as simple as it can be but it's still new) sounds more to deal with than pushing markdown files to your Github account.

Web apps can be used from any device - your home machine, your phone, your tablet, your friend's machine, a computer in an internet cafe halfway around the world. Pushing markdown files up to your github account... well, okay, I dunno, maybe you can figure out a way to do it in all of those cases if you're the kind of person who thinks that's the simplest thing to do, but most people are not.

If you have A LOT of posts and pages it can take minutes to generate your blog, then push it to your server. Ghost is pretty much instant. Write in the browser, publish in the browser.

I don't have to spend 1 second figuring out how to convert layouts and templates and blah blah blah. I just want to write content.

I thought the same. What is the advantage of this platform over something equally simple like Jekyll or Octopress?

Not a great comparison. Ghost has an admin that's pretty nice. While it might be easy for all of us to jump into code to compose our posts, with Ghost you can jump in on your mobile device, make some edits and hit save.

If you're a programmer, I don't see the advantage over, say, SGit. Just launch the app, edit the file, commit & push. And hey, you can even edit offline.

> except this one is hip and popular

And sometimes that's all it takes.

I guess we'll never get this horrible metaphor out of software development, but I keep trying:

It's not _reinventing_ the wheel. The wheel is an old concept, so are blogs. This is a new/different implementation of that wheel. "Reinventing the wheel" is as pointless as it sounds. But what about making new, better wheels? Maybe not everyone needs the same wheel?

(Edit: Just to clarify, I'm not exactly a fan of Ghost.)

That's the point: while Ghost claims to be a new, better wheel, it actually seems (to some people) to be reinventing it.

It would only be "reinventing" the wheel if it was (or, at least claimed to be) an independent first-time creation of a blog engine without prior knowledge of them existing.

So, unless we're talking about something like that (unlikely) there's no such thing as "reinventing the wheel" in real life.

There's just "creating a different wheel". BBS, Enkei and all the wheel companies are not "reinventing the wheel". The create a new version of wheel.

Even if it's boring and similar to existing things, it's not a re-invention. That would be like calling another CRUD app that is "re-inventing the CRUD".

If you're going to complain that it's not actually reinventing, then can I point out that Ghost is not a wheel, either?

It's an idiom with a known meaning. It doesn't have to be literally true.

Yes, but the problem is that it's not a good idiom either, for the reasons explained:

It's based on a broken metaphor, and it doesn't reflect how things work on software.

They could have at least said "they made just another blog engine".

I thought the implication in "reinventing the wheel" in software was that the current de-facto implementation obviously got it perfectly right, so trying to improve on that would just be absurd. Of course, i'm sure people were saying that when someone decided what the world needed was a better Movable Type.

And to go all the way down the pedantic rabbithole, there's never been software which is elegant or utilitarian enough to be worthy of considering a "wheel" in that regard.

It gained a lot of its initial traction by unveiling a very slick, built-in dashboard. This was a design concept, and hasn't materialized (yet).

In its current state, it's an extremely basic blogging platform with a built-in Markdown editor. For me, that was enough to become an early adopter, because I think it has more long-term potential for interesting new features than, say, Jekyll. That said, there are still a lot of rough edges.

Regarding the dashboard, there is lower-hanging fruit with greater payoff for Ghost, including the API (which is apparently a prerequisite to the dashboard anyway). They seem to be prioritizing the right things.

From an end-user perspective:

1. It's easier to use than WordPress, Drupal etc. That's because it does much less than those platforms on purpose. The WordPress admin area has become a scary place for new users, particularly once they start to add plugins that litter the admin menu with extra entries.

2. The admin area is fast. Moving around feels slick – it is a desktop-like experience; hitting Command+S will save the draft you're working on, for example, rather than bringing up your browser's “Save Page As...” prompt. The Markdown preview is as-you-type, and they use the screen estate very well. It's a comfortable environment to write long posts in, and it hasn't been tacked on as an afterthought, like WordPress' full-screen editing mode.

3. The organisation behind Ghost, the "Ghost Foundation" is a non-profit. It doesn't have investors or shareholders; it is answerable only to its users and funded by those who pay for official Ghost hosting.

4. Pricing for official Ghost hosting is very reasonable ($5/month for 1 blog, $14/month for 5 blogs).

5. Although it's still in development, the proposed Dashboard looks far more useful than competitors. WordPress' Dashboard often looks like a wasteland once a few plugins have been installed – a mish-mash of "Upgrade to the PRO version!" prompts and RSS updates from the WordPress feed.

And from a developer perspective:

1. GitHub-based workflow.

2. MIT licence instead of GPL.

3. Less code debt (for now!).

4. Handlebars-based theme templates that enforce separation of presentation and logic. (It will be possible to add custom handlebars tags via the Ghost Apps API in the future for more advanced stuff, but it's already pretty solid.)

5. The official hosted version allows you to upload your own themes (unlike wordpress.com, for example).

As a developer and end-user, it's already a more enjoyable experience than competing platforms for simple blogs, and I think by the end of the year – once the app API has launched and developers are adding forum apps and more – it will be a real competitor for more complex sites too.

Hey, that was actually a pretty clear and detailed answer. Thank you for not treating it as a rhetorical question.

I've been looking to switch away from wordpress for awhile, but I don't know of an easy migration path that transfers all my data. Is this possible?

Not at the moment, but there will be an official importer app later this year – it's one of the major items on the roadmap: https://github.com/TryGhost/Ghost/wiki/Roadmap#importer-part...

disclosure: I'm a former founder of a successful WordPress hosting platform and spent many years in the WP space.

I'll take the opportunity to counter, if I may...

1a. It's easier to use than WordPress, Drupal etc. That's because it does much less than those platforms on purpose.

A key stumbling issue with Ghost is writing in Markdown - most non-tech people use WordPress's WYSIWYG editor because they don't understand markup as a concept. That's a BIG barrier to entry that Ghost needs to address or remain niche.

The other big barrier to entry with Ghost is that it's super difficult to install compared to WP - requires Node.JS which doesn't run on most shared accounts (where most people host blogs). The average non-technical person shouldn't be trying to manage a VPS or server.

Yes, there are hosted Ghost services but those cost money, hosted WordPress starts a $free on wordpress.com.

1b. The WordPress admin area has become a scary place for new users, particularly once they start to add plugins that litter the admin menu with extra entries.

Most plugins don't abuse the addition of menu items - most of the ones that create a new root-level menu item do so because they have sophisticated/deep enough configuration that requires it. I'm not sure that added menu items makes WordPress particularly "difficult to use" especially if the alternative is that functionality isn't available in the first place.

WordPress's /wp-admin/ can be super fast - that's down to where you are running it. Drafts are saved automatically and with browsable history so the need for Command-S is removed.

Materially, that's no different to the WordPress Foundation.

4. Pricing for official Ghost hosting is very reasonable ($5/month for 1 blog, $14/month for 5 blogs).

Pricing for official WordPress hosting is even more reasonable - free. But also has a clear path to scalability via various Managed WordPress Providers if needs be.

The "Upgrade to Pro" part is a bit of a straw-man argument. I'm sure there must be some WP plugins that do that but not many, and any widget on the Dashboard can be easily moved or removed. Users can customize their dashbord with what they want, and a reasonable counter-perspective would be that a small RSS feed of WordPress news and articles isn't a negative.

---

At the end of the day, debating WordPress vs Ghost is like debating Python vs Rails or Debian vs Centos. Differently philosophies - Ghost is certainly more focused on attractive features for developers, WordPress is focused on mainstream adoption. Makes sense given their relative positions in the market.

I think HN gives WordPress a hard time, applying niche values like code elegance and utilization of the latest technologies and ignoring in plain sight the fact that both of those are actually WordPress's advantage. Anyone can create a plugin, and while many of them are utter shite, there is an amazing breadth of quality ones out there. And PHP, while derided mostly on HN, actually is one of the most prevalent languages in terms of install base and suitability to low-resource shared-environment mainstream use.

A key strategic issue Ghost will need to deal with is that the use-cases for "blogging" have polarized over the years. People run a blog either as part of a wider site - which is where WordPress wins given it evolved into a site CMS. Or blogs are written to be part of a content network - which is where Medium and Subtle are playing where you get a "Ghost" like experience but the distribution network to go with it.

I agree with you – WordPress is often criticised within the tech community. That's because its philosophy has always been to optimise for end-user happiness first and developer happiness second. If you think something in WordPress is crazy – such as supporting PHP 5.2 – the reason is normally that it may be worse for developers, but it is much better for end-users. As soon as I understood and accepted that, I had much more fun with WordPress.

But I think the criticism in the tech community is also a cry for a strong alternative. And we have so few good ones today. That's why I'm happy to champion and celebrate anyone taking on an entrenched platform. If nothing else, Ghost and competing platforms will encourage WordPress to become something even better.

Some responses to your comments:

Right, so what happens when your needs grow and you need functionality that extends outside of the envelope of Ghost?

You'll install Ghost apps, just as you install WordPress plugins like Advanced Custom Fields if you want to bend WordPress into a more feature-complete CMS. It's too early to say whether Ghost will succeed at creating a better add-on system, so perhaps this particular debate is premature. The challenge will be in creating an API that allows developers to bolt-on features without also enabling UI cruft and the sort of performance headaches and conflicts you see with WP plugins, which don't enforce sane practises like namespacing (thanks, PHP 5.2!).

Ghost may be "easier to use" but one thing that is difficult with it over WordPress is it's a lot harder to install - requires Node.JS which doesn't run on most shared accounts (where most people host blogs). The average non-technical person shouldn't be trying to manage a VPS or server.

I agree with you – at the moment the install process is too hard for non-technical users and a burden for technical ones. But the official Ghost hosting platform is inexpensive, easy to get started with, and happy to accept custom themes, so there is little reason for most users to look elsewhere. Those who want to host a blog themselves already have several alternatives that don't require sysadmin skills – WebFaction has a Ghost installer, for example, and they charge shared-hosting prices ($9.95/month). I suspect we'll see other suppliers who offer Ghost-only hosting to bloggers with lots of traffic, just as WP Engine and others have done with WordPress.

The problem is that these menu items can be created in at least three different areas – under Settings, under Tools, and at the top level. This makes things harder for users because they have to remember where the menu item of the configuration page for that plugin they installed four months ago is. I use WordPress every day and the menu system is a giant game of hunt-and-peck.

The menu setup also makes things harder for developers because they have to make the right choice of where their menu item will live. A great many get it wrong. (They will frequently place items at the top level that should have been under Settings or Tools.)

Additionally, the menus API invites games of Competitive Menu Trumping, where developers fight to get their menu item above those of other developers. Automattic are guilty of this – install their Jetpack plugin and you'll find a new "Jetpack" menu item right at the top of the menu bar, just below Dashboard: http://d.pr/i/AZyt This is either a desperate call for attention or an admission that the menu system is broken – that if it was placed under the Settings header instead, the Jetpack settings page might never be found.

Plugin author "WingerSpeed" has even created a plugin named "Slim Jetpack" that functions almost exactly like Jetpack, but removes the menu item and banners from the top of the page. It's been downloaded almost 50,000 times – nowhere near the 10 million or so downloads that Jetpack itself has, but still a sign that menu bar bloat, dashboard adverts, and enforced sign-ins are complications and annoyances that many users would prefer not to have.

WordPress's /wp-admin/ can be super fast - that's down to where you are running it.

I don't think it's just down to hardware – WordPress is not yet a full MVC-style JavaScript app. Side-by-side, it does not feel as fast as Ghost on the front or back end. I just logged into an account I managed with one of the big WP hosts, and it's a huge improvement over how it feels with shared hosting, but if you compare it side-by-side with a Ghost blog it feels slower. Perhaps that's a little unscientific and anecdotal, but raw perception matters. To be fairer, Ghost's perceived speed is largely because it got a fresh start and doesn't carry 10 years of baggage, but also because it does less and makes fewer requests (11 vs 18 for the basic admin page). But the upshot is that it behaves more like a modern front-end JS app than WordPress currently does (even though that's changing) – the perceived performance mirrors the underlying architectural choices.

Drafts are saved automatically and with browsable history so the need for Command-S is removed.

Auto-save is great. I agree that this is better for users. My point was that, for writers and users with muscle-memory Command+S twitches like me, the fact that the shortcut triggers a save action instead of a browser menu shows that the app's been developed to feel more like a desktop one. It's a nice touch. (I believe the WP post editor disables Command+S when the box has focus, so it's not like they haven't thought of it – it's just a different approach.)

Materially, that's no different to the WordPress Foundation.

Absolutely. Except that the Ghost Foundation is just the Ghost Foundation. There is no commercial arm like with Automattic – no for-profit company feeding from the work of open source contributors (and giving a lot back, to be fair). Most won't care or appreciate the difference, but it's an important distinction for some.

Pricing for official WordPress hosting is even more reasonable - free.

Sure, it's free if you want to use a modified version of WordPress. It's free if you don't need to add plugins or themes. It's free if you don't need to edit CSS. It's free if you don't want to use your own domain. Once you add the various upgrades or one of their (very decently priced) bundles, and once you pay for separate email hosting (which WP.com doesn't include), you'll soon be looking at shared hosting pricing or the headache of migration to a self-hosted setup. It's great that there's a free entry point, but perhaps sometimes that can turn into more of a burden than a benefit. Ghost Pro has a free full-featured trial, for what it's worth.

But also has a clear path to scalability via various Managed WordPress Providers if needs be.

Ghost Pro offers a clear upgrade path that scales based on traffic, just as managed providers like WP Engine do: http://d.pr/i/SWHO If you start with them you can grow with them too.

It's not a straw-man argument – it's a genuine problem. I started collecting screen shots of the Dashboards and various intrusive banners from Dashboards I've logged into over several years of web consulting and it paints an ugly picture. (I'll try to publish them in the future if I can strip identifying features.)

I just logged into an account I managed with one of the big WP hosts, and it's a huge improvement over how it feels with shared hosting

I'm curious, which one were you using?

The site I logged into was with WP Engine (not a site I operate; just one I help maintain). It's fast, but there's still a perceivable lag when clicking around the admin area.

I'm not sure if you've played with Ghost much, but I recommend signing up for a free demo account at http://ghost.org if you're at all curious. Create a few posts, click between them on the contents tab, dive into the (currently very bare!) settings page, and try editing and saving a post – everything feels almost instantaneous. I don't get that feeling with WP, even on WP Engine or WP.com.

I was wondering the same thing a while back.

I signed up for a bunch of platforms just to cover all the bases and see the differences: Ghost, Medium, Silvrback, Svbtle, and Postagon.

I won't say which is my favorite, because I think everybody has different needs. But I found a decent Medium article which compares most of the big platforms. Might be of help to anyone else looking for a new blog home: https://medium.com/blogging-and-seo-tips/d62682d9b0aa

I gave it a shot for a month. So far there was nothing that could convince me that goes beyond a fancy look & feel. Feature-wise it's ages behind WordPress (not suprisingly, WordPress is a huge eco system).

It's missing several fundamental features that make it unusable (in my opinion) for any serious user, such as user management.

Just as an aside. If you're looking for something built in Node that's more of a CMS, my Kickstarter for Webhook recently hit it's goal. I really like Ghost as a blog service, but you might like our system if you need to build a more customized CMS experience for multiple editors.

https://www.kickstarter.com/projects/1749618880/webhook

Nifty!

Any plans for a self-hosted version?

Yes. By end of the year. Check the updates, I explain the reasoning there.

Whoo. I'm glad to hear you have firm plans about it, that was the only thing preventing me from backing you.

Why Firebase + static pages instead of, say, varnish? (Honest question)

Because people who know HTML and CSS don't know how to set up or manage a database locally. Even worse, they'd have no idea how to port that data onto a live production server. Firebase makes the install pretty painless. Our only dependency is Node. Right now that means even windows users can get up in running in less than a minute.

As a designer I really like the ease of systems like Jekyll, but it was a dead stop as soon as I wanted those sites to be editable by other users. This way we get the best of both systems.

I have multiple client sites I'm currently designing with yeoman (and dreading converting to Wordpress). Just sent you $20 - can't wait to kick the tires.

> - Ghost Dashboard

I'm really excited for that one, that was the reason I installed it in the first place only to realise it's not available yet.

-> "Very soon you’ll be seeing a fresh, slick Ghost(Pro) dashboard"

Is "(Pro)" indicating that the dashboard component will be available only for paying/hosted customers and will not be open-sourced?

I doubt it, since the dashboard was a main selling feature of its Kickstarter campaign. A move like that would probably sacrifice any goodwill its garnered among early adopters.

I think it is, worryingly for early backers.

http://blog.ghost.org/introducing-ghostpro/

I did the same thing - They didn't have the "coming soon" banner on their website back then. Kinda scummy marketing if you ask me. FOSS is awesome, just don't overpromise and underdeliver or you'll burn goodwill.

curious about what happened from current tagline and differentiating point from wordpress: 'Just a blogging platform.', to the statement here: 'Ghost has the potential to be far more than just a blogging platform' ?

Please don't rewrite titles unless they're linkbait or misleading. ("Ghost blogging platform, future." wasn't even intelligible.) The post's title, when combined with the site name, is fine.

Got it.

Ghost promised to be a simple blogging platform. It's simple if you use it as installed. It's not so straightforward to tweak and upgrade. I'm not very excited about it anymore.

They may improve the upgrade process for users. It's also open source, I'm sure they'd love to hear about your upgrading concerns.

On the other hand, there's also the paid service, which you don't have to manage at all.

Does anyone have any recommendations for comment systems on Ghost? It looks like they themselves are using Disqus and Jeff Atwood is using Discourse (obviously) are there any other contenders?

Some people like Livefyre, which can add some more real-time abilities if you happen to have a large user base.

I'm using NodeBB -- https://nodebb.org/

I wanted something I could host myself, which rules out Disqus and LiveFyre. And while Discourse lets you show the comments on the blog page, people have to go to the forum page if they want to comment.

I saw this one tweeted out yesterday by one of the Ghost devs. Integrated commenting similar to what Medium does. http://ouija.io/

I was confronted with this problem myself and decided to code up a jquery plugin that would let me add comments using firebase, so I wouldn't have to rely on a third party commenting system.

It's really basic right now, but I'm planning to extend it's functionality in the future significantly.

You can check it out here: https://github.com/ximi/jquery-firecomments

I'm really glad that Ghost is progressing nicely — I've used it for a small project, and it was a really nice experience.

I didn't think another blogging platform was really needed, but their stats and revenue make it sound successful. Congrats, they succeeded where I didn't think it could be done, or would have just hosted WordPress and tweaked it if I was doing it myself.

Reply on news.ycombinator.com