CASPAR.green – That's just, like, my opinion, man.

Further Thoughts on Ghost

December 16, 2018 by Caspar Leave a Comment

As I’ve continued to delve into the Ghost CMS, I’ve found there are a few things I miss in addition to those I pointed out in my original review.

A media manager. The support forums say this is “in progress”, but currently there is no user mechanism to re-use images or browse images that have already been uploaded to content. If you use an image in a post, then in a later post want to use the same image, you have to upload a fresh copy. I could, maybe as a developer, live with this on a self-hosted installation where I could have direct access to my content on the server where I could see what is uploaded and where, on the GhostPro service, you don’t have access to your installation’s file system. So media management is currently either a royal pain (if you’re a developer with server access) or non-existent (if you’re an end user).
Basic form processing. In addition to the lack of comment and discussion noted previously, you also have to rely on 3rd party solutions to implement any kind of forms (like a contact form). To be fair, this is not an issue unique to Ghost. WordPress, for example, requires installation of a forms plugin to handle this. But, unfortunately Ghost has …
No capacity for plug-ins. Which means that the only way to implement things that Ghost doesn’t do is through a 3rd party service (or you could develop your own service and run it on a separate server.) You’ll also need to take the lack of extensibility into account in assessing the cost of your site, since you’ll likely have to pay for service subscriptions for the missing functionality or develop your own and host it elsewhere.

While I still like Ghost in general, and I still absolutely love the editor, these 3 things are blockers for me to actually port my own blog to Ghost.

Review: Ghost CMS

December 15, 2018 by Caspar Leave a Comment

Ghost debuted in April 2013. It’s focus is on “professional publishing”. They refer to websites as publications. It’s actively developed by a non-profit foundation and it’s open source. All good.

It’s built in Javascript on top of Node.js using the Handlebars templating engine for its presentation layer. Content is stored using MobileDoc format.

When I installed a local copy of the current version (2.8.0) to take it for a test drive, I ran into an issue where it wouldn’t start after the install. It took a few minutes to figure out that the issue was an older version of Node, which once updated to the recommended version (currently Node 8.x), cleared the way to a smooth start.

Once up and running, the experience of getting around in the CMS was about as intuitive as one can ask for. The editor is, hands down, the easiest I’ve ever seen. Easier than the old WordPress editor, and one can see where some of the Gutenberg styling has been borrowed from the Ghost editor, except that this one really works.

The controls for both pages and site settings are laid out where you can find them easily, and are pretty much self-explanatory. Installing the sample data gives you a tour of the basic concepts in 7 short introductory posts. Once you get through them (it only takes about 15 minutes to read them all) you’re up and running. I pretty much fell in love with the experience within the first few minutes.

From a development perspective, the code is beautifully organized, and the documentation is thorough. Within a couple hours, I’m able to find my way around the templating system, the data structures, create custom routing (for setting up custom sitemaps, etc). Within another hour or so, following the well-documented source code of the included sample theme, I’m able to lay out the basics of a new custom theme. Docs are available for using ghost as a headless CMS with REST API endpoints.

For a site that is about publication (what Ghost was designed for) this is, hands down, the best thing I’ve seen in a long time.

That said, Ghost does have its limits.

If you’re running a blog and want to have open comments, you need to integrate it with a comments service. Disqus is the one recommended as being the most common choice. Discourse is another that appears to be a popular choice, especially if you also need a forum or discussion companion piece.

E-commerce is not supported. At least not within Ghost itself. So this is probably not the CMS you’re looking for if you’re trying to set up an online store.

It’s API allows it to be extended with any number of other online services – anything you can run through a web hook, embed in an iframe, or inject in a script. But all these things are just that: integrations with other platforms, none of which are handled natively in the Ghost CMS itself.

On one hand, this is understandable. Ghost focuses on publishing. What it does, it does exceptionally well. But it’s not trying to be all things to all people. That means that you can’t do all your online business in one place. You’ll use Ghost for your publishing, and you’ll use Discourse (or whatever) for your social stuff, and maybe Shopify for your sales.

There seem to be plans for expanding the range of Ghost-specific add-ons (Ghost Apps, as they’re called) that may eventually start handling some of these things. But they are not to that point yet.

The beauty of it being so simple is also one of it’s drawbacks. What’s there is highly customizable. But if Node, Handlebars templates and custom routing with yaml configurations scare you, you won’t be able to take advantage of it. (Or, you can hire someone who is comfortable to do that for you.) The platform is developer friendly; but it will require some developer chops to actually build it out the way you want it. “Developer friendly” does not mean (as it often does in WordPress) that you can just install a plugin. There aren’t many (if any) readily available such things. You have to know something about code. But for someone who knows what they’re doing, it’ll be a pleasure to work with.

That said, there are a couple dozen (as opposed to many thousand in WordPress, for example) pre-fab themes you can purchase from various vendors. Your mileage may vary.

The other limiting factor around Ghost is that it’s not going to run on a lot of popular shared server hosting plans. Most of these plans are set up for more traditional server-side scripting languages (mostly PHP). Ghost runs on Node. I asked my go-to hosting company, SiteGround, a couple years back whether they had any plans to allow node based platforms on their shared hosting. The answer was, no. (I could set up anything I wanted, though, if I bought a VPS plan.)

So hosting a Ghost site means either hosting it with the Ghost foundation. Pricing starts at $29/month. If you’re a serious blogger, that’s probably worth it. And hosting with the foundation helps pay for more development, which can only help make good thing better. And, since they take care of it, you don’t have to worry about updates, security and all that other server-side stuff. You can focus on your publication. Which is the whole point.

If you want to go it on your own, you can install it on a DigitalOcean droplet, or you can run a Docker container on your own VPS. Or you can install it on any VPS running Ubuntu Linux with a minimum of 1GB memory. The Digital Ocean 1GB droplet is currently $5/month.

In all, I have to say I liked it a lot. I could happily run this blog on Ghost. Someday, in the not-too-distant future, I just might.

See also the follow-up to this review with a few added cautions.

The Odd Universe of Floating Point Math

December 14, 2018 by Caspar Leave a Comment

Twice this week, I’ve run into bugs in programs caused by not accounting for doing math with floats.

Floats are a numerical data type used by computers when there is a decimal or fraction involved in the number.

1.5
$125.47
2/3
3.14159

For somewhat arcane reasons I won’t go into here, computers frequently end up just a smidge off when they try to do math with these.

As it happened both instances of the errors were happening when dealing with adding and subtracting money amounts. The program was looking for a certain amount to go to zero as a condition to end a loop. Comparing the condition against the result of the floating point calculation came out to some small, but not zero amount, $0.000000000099995, or something. Since it’s not zero, the loop wouldn’t terminate.

Here are a couple quick ways to deal with these small remnants of cents (or whatever fractions) when you come across them. (Hat tip to Bernhard Breytenbach.)

When calculating with money, convert everything to cents before you start, then do the calculations with integers. Convert back to dollars (or whatever) when you’re done.
After doing floating point calculations, you can cut off the long tails of the decimals by rounding (most languages have the capacity for rounding) the number times 10 to the power of however many decimal places you want to keep, then dividing the rounded number by that same power of 10.

So, for example, if you have n = 8.22222222222244444 and you want 2 decimal places:

round (n * 100) / 100

Gives you round (822.2222222244444) / 100 == 822 / 100 == 8.22

Round as late as possible though. If you still have more calculations to do, you’ll end up wasting a lot of cycles rounding.

Don’t Nag Your Customers

December 13, 2018 by Caspar Leave a Comment

Last night a WordPress security update was released.

Versions of WordPress prior to 5.0 also got security releases. So you don’t have to update to 5.0.1; you can (and should) update to the next point in whatever older version you happen to be running. (So, for example, if for whatever reason, you’re still running 4.6.x, you should update to 4.6.13.)

When I did update to 5.0.1, I got a nag on the update:

Yes. I know the Classic Editor prevents use of the Block Editor. That’s why I installed it.

Come on.

This is nagging a customer who’s already made up his mind. It’s not going to convince me to change. It’s just driving toward resentment.

Don’t do this to your customers.

De-Google-izing

December 12, 2018 by Caspar Leave a Comment

As part of the end-of-year housecleaning I’m working on I’m transitioning as much as possible, away from Google.

The inspiration for this is similar to my leaving Facebook earlier this year: they just collect too much data. Not that I have anything to hide. I search for all kinds of odd stuff. But since they’ve taken “do no evil” out of their company policy my perception is that they’ve been on a slippery slope into the world of stalking.

So, for example, I’ve already switched my default search engine to Duck-Duck-Go, and I’ve minimized my use of Drive and Gmail.

Now it’s time for a few changes to this site (and others that I administer) to move away from Google Analytics, so that I’m not complicit in Google tracking my visitors. Also, I’m going to be transitioning from hosting my videos (the few that I make) from YouTube to Vimeo.

(I expect that Vimeo also does some tracking, but since I’m paying for the service, at least I won’t feel so much like I’m the product rather than the customer.)

I’m also going to be working on bringing all my sites into GDPR compliance over the next couple months.

Some regulation is important. The internet has always been a bit of “wild west”. Nowadays, though, it’s a wild west with several Big Brothers in the mix. One must do what one can to make it hard for them.

Top 4 Books, 2018

December 11, 2018 by Caspar Leave a Comment

The end-of-year clean out continues.

I’ve made a pile of books to take over to the library for donation to their weekly book sale.

I’ve read a lot of books this year. Now it’s time to make room on the shelf for next year’s reading. I’m only keeping the few that were so compelling that there’s a chance I’ll look back at them again. Those 4 are:

Sapiens. Yuval Noah Harari.
Being Mortal. Atul Gawande.
Thinking, Fast and Slow. Daniel Kahneman.
The Black Swan. Nassim Nicholas Taleb.

These 4, then, get my award for books of the year for 2018.

It’s Time for Arras to End

December 10, 2018 by Caspar 2 Comments

With the coming of WP 5.0 and Gutenberg, and my new focus on moving in directions other than WordPress, it’s time to finally put the Arras WordPress theme to bed.

I realize that I did this once before, and then thought that I might resume. This time, however, the decision is certainly final.

To those who have been waiting for further updates, my apologies. You are welcome to fork the source code and attempt to continue on your own. That said, my recommendation to Arras users is to find an up-to-date, reliable theme that will serve you better into the future. There are plenty of them out there.

The 5 Stages of WordPress Grief

December 9, 2018 by Caspar Leave a Comment

The other day I posted that the direction WordPress has taken has saddened me and made me occasionally angry.

In a subsequent exchange on Twitter, I somehow attracted the attention of someone who was definitely angry. Very angry.

As I reflected on this further, I realized that what many in the WordPress space have been going through can be described by the classic Kubler-Ross model of grief. Also known as as the five stages of grief.

(One note about the so-called stages, before I go further. Since Kubler-Ross first published the theory, it’s been widely recognized that transition to one stage to the next isn’t always linear. More often than not, grieving persons move from one to any of the others, in different order, sometimes repeating certain stages.)

1. Denial

For a long while, most of 2016, this was where I was. The Gutenberg project had been announced. Work had begun. But it was not really on my radar, and when it was, it was far enough in the future that I could choose to ignore it.

I revisited this stage several times through 2017 and ’18 as well. I saw people in this stage right up to the very last moment, as recently as last week: “It’s not too late to convince those in charge (whoever they are) to change their minds about it.”

2. Anger

This is where I occasionally find myself, and where I’m encountering quite a few folks. Something we loved is lost. Anger is natural. Some people handle it better than others. In anger, some (like the fellow I encountered on Twitter) are lashing out at anyone and everyone they can find to blame. Some feel hurt and want to hurt back. That’s unfortunate.

As much as I disagree with and am occasionally angry about the way things have happened, I want to go on record that I’m not holding any personal grudges. I’ve said before, there are a lot of people who believe in the new direction of the WP project. A lot of people have put in a lot of hard work and effort to make what they believe is a good thing. People do what they think is right. I happen to disagree. I wish them well in their endeavors.

But I also stand by what I’ve publicly said before: WordPress, with this most recent release can no longer claim to be enterprise level software, and I will be much more cautious in my own use of it, and in my recommending it for others to use.

Anger need not always be destructive. As Kubler-Ross pointed out, it’s an essential thing to work through, a part of separation and self-definition, that when done well helps one to be stronger and more resilient in the long view.

3. Bargaining

Over the past few weeks, I’ve been here myself, and I’ve seen plenty of this back and forth over the interwebs. Given the realization that the loss is impending, or has already happened, we make deals to try to avoid what is inevitable. All the calls for “It’s not ready.” All the complaints about, “The timing is terrible.” All the offers of “If you will just add this feature (or remove that feature), I will accept it.” All bargaining.

The addition of the Classic Editor plugin and the Core team’s commitment to maintain it for 3 years was a grand bargain. Sometimes bargaining is useful. It bought some of us 3 years before the loss we are feeling becomes permanent. It is best ro remember, however, it only forestalls the inevitable. The day will come when the Classic Editor will be abandoned, and those who have remained in this stage will have to move… again.

4. Depression

I confess, I’ve spent quite a bit of time here as well. And still do. On the road for 4 hours one day recently, I found myself singing along with the Dixie Chicks: Hello, Mr. Heartache, I’ve been expecting you!

I’ve invested a lot in WordPress – not as a contributor, but as a promotor, as someone who learned the software through and through, wrote themes and plugins for it, and enthusiastically recommended it to clients I hoped it would serve long and well. For about 10 years, I was all-in. And, though no software is perfect, and no community of enthusiasts is perfect, I felt a part of something worthwhile and was invested in it.

Losing that is bound to make one feel blue. Even nostalgic. In these moments I can recognize that I learned a tremendous amount from WordPress. I benefitted from it educationally, emotionally, and financially. I made a lot of good friends, and I hope that those who are staying with WordPress even when I’m no longer as engaged in that community will remain friends. Acknowledging that is a good thing. Going on record as being grateful for the opportunities it gave me is appropriate here.

It’s also helpful to recognize here that nothing lasts forever. Sadness. Which leads to…

5. Acceptance

Acceptance can take 2 forms. One is a resolve that one might as well prepare for it. These are the ones that are signing up for the “How to make Gutenberg blocks” courses that are now popping up everywhere.

There’s another kind of acceptance, the kind I find myself moving into. This has happened. It is what it is. What can I take from this experience and move on?

“Learn Javascript deeply” seems to be one of those things. It’s a fact of the web, and even if WP implemented it in what I regard as a fatally flawed way, if I’m going to be in the web business, I need to know it. For a long time WordPress allowed me to skirt the necessity of it.

“Beware the silos” might be another. Even if the silo you happen to be in is a very large (approaching 35% of the web) one. Keeping your head buried in one hole too long blinds you to important things that are happening in the rest of the world.

Another reality that acceptance makes me aware of is that as big as it is, WordPress is still only 32% of the web (as it claims). Once you get beyond that silo you find that this number may not be as accurate as you once thought, and that even if it is, that still leaves 68% of the web that is doing something else, and most of it is getting along just fine, thank-you very much.

Part of acceptance is realizing that there are twice as many opportunities outside of WordPress as there are inside it. And, to make things even more interesting, one can note that WordPress is low-end on the pay scale for people who know what they’re doing. Part of its appeal, and part of the reason it has grown so rapidly is that it serves the long tail market of sites that are being built on a shoestring.

There are plenty of notable exceptions, yes. I’ve long pointed to them as proof that WordPress was viable in the enterprise web space. But I’ve always said, and still say, that if your bank is running its web operations on WordPress it’s time to find another bank. Now, even more-so.

Sure, we lost something. But we are not cut off from the future. For anyone looking to transition from the typical $2000 – $10000, one-off site build range toward larger projects, now is the perfect opportunity to make the change for those who have the skills. And for those who don’t, now is the perfect opportunity to get them. Start with Free Code Camp.

As I noted before, these stages don’t necessarily progress in order. I find myself moving between them, sometimes several times in the same day.

This post is my self-reflection and observations, shared with the intention that they might be helpful to anyone else who stumbles across it.

And to all my WordPress friends, be well and keep in touch.

Choose an IDE and Stick with It

December 8, 2018 by Caspar Leave a Comment

I’ve used the PHP Storm IDE for the last 4 years (give or take). All the cool kids seem to be transitioning to Microsoft Visual Studio now. I have to say, it looks pretty slick, from what I’ve seen of it, but I’ve not tried it yet myself. I’m looking forward to doing that, but not anytime soon.

For now, I’m happy with PHP Storm. It’s part of the Jetbrains IDE set, and for 2 years I’ve had the license for the whole set. Up to now I haven’t used many of the other IDEs in the set, but that’s changing.

Last year when I started looking into Python, and again this week as I’ve been delving more seriously into Javascript, I’ve appreciated having the companion pieces in the set. I downloaded WebStorm, and within 10 minutes was able to find my way around, run my code and debug.

Having a familiar interface to program in when exploring less familiar language has been a great help. Certainly a lot easier than debugging in the browser.

If you’re going to be doing serious programming, you owe it to yourself to find a tool (or set of tools) that help you. That means that you take some time to get familiar with them and configure them to fit your workflow. And, yes, it’s going to take some time and effort. Once you’ve done that, a good IDE setup will become invisible; it will feel like it’s not even there, so you can focus on your work, not your tools.

It says something for how well the tools work, when as I started exploring a new language in depth outside my familiar development environment I found myself trying to figure out as much “How can I edit this code?” as “What do I need to change in this code?” Returning to a familiar coding environment (for me, WebStorm) removed that first hurdle so I could get on with the second.

Whatever your tools, choose a set that works for you, and stick with it until it no longer stays out of your way. (That’s the indicator that you might need a change. Some people I know are still using VIM and are completely happy with it!) Some shiny new things that come along may be worth adding to your kit. You know them when you start feeling a pinch in your workflow: “I wish I didn’t always have to stop what I’m doing because…” At that point you find the tool that addresses that need. But switching from one to the next too frequently is a form of hiding from the more important work.

A Few Notes on Updating to WordPress 5.0

December 7, 2018 by Caspar Leave a Comment

I updated this blog to WordPress 5 yesterday.

A couple notes:

1. When I updated, even though the Classic Editor was installed and active in 4.9, the update disabled the plugin! Grrrrrr. So, first thing I had to do was go to Plugins and re-enable it.

2. There are new settings under Settings > Writing: “Default editor for all users” and “Allow users to switch editors”. Set them to the “Classic” and “No”, as shown. If you’re using Advanced Custom Fields you’ll get the white screen of death if you don’t. (Update: Updating the Classic Editor plugin to version 1.1 should clear up the ACF compatibility issue.)

3. That said, even with the “Don’t allow users to switch” setting, the posts page will ignore it and offer users the option to edit posts in either editor, so (at least for now) it’s not a bullet-proof prevention to disallow what’s now called the “Block Editor”. (Update: again, updating to Classic Editor 1.1 seems to have cleared this problem as well.)