Textpattern's face to the public

wet · 2025-10-10 04:02:52

Just stumbled across a Facebook ad from a web host that caught my eye. The headline goes:
“Does your capital_P_dangit load like chewing gum? How fast does your capital_P_dangit load?”

Now, let me sit with this for a moment. Performance issues have apparently become so widespread and expected with capital_P_dangit that hosting companies are literally building their marketing strategy around fixing it.

This tells me two things:

Performance has become a pain point worth advertising against
We’ve been doing something right all along

Look, I’m not here to throw shade at other platforms. That’s not our style. But I will take a moment to appreciate what we’ve got:

Fast load times right out of the gate – no plugin juggling, no cache gymnastics, no “premium optimization packages” required
Lean, mean architecture – That went well ;)
Happy visitors – who aren’t staring at loading spinners wondering if their browser crashed

It’s kind of wild when you realize that what we consider baseline normal has become a premium upsell feature elsewhere.

Sometimes you need a reminder like this to appreciate what you’ve got. That reminder may make it onto our homepage just as well.

Bloke · 2025-10-10 07:45:12

I’ve just had that exact experience migrating a sprawling WordPress ecommerce site to Textpattern. The original site was painful to use. Like, anywhere between 4 and 8 seconds to FCP. Using the backend was like wading through quicksand. I’ve drummed fingertip holes in my desk waiting for it to do anything.

I cloned the HTML and CSS, Txp-ised stuff with a few forms and a single page template, and trimmed some of the 150Kb CSS rules down to 60Kb so it fit in our stylesheet database column.

The HTML is still a bloaty div soup mess that could be optimised, and I could probably cut the CSS down to a fraction of its size, but as a path of least resistance, it works as-is and can be gradually iterated down to a leaner footprint.

The crowning moment when I demoed it came when the client sat down at his computer, logged into the front end site and typed part of a product code into the search box. He hit enter. The article list came back instantly and he sort of blinked, did a double take and said, partly involuntarily, “Jeezus, that’s quick.”

There are currently over 6500 articles in the database.

So, yes, we do take performance for granted here. And we should shout about it.

bici · 2025-10-10 15:09:50

Bloke wrote #340903:

So, yes, we do take performance for granted here. And we should shout about it.

Textpattern: A lean CMS with no Performance Anxiety

etc · 2025-10-10 19:54:16

We have spent hours to save your seconds.

Maintaining this level of performance will be challenging with distributed cf storage in txp5.

giz · 2025-10-11 19:43:48

I’m including the demo access page in my homepage design as it is likely to be the first stop for interested visitors.

It’s been a long time since I viewed/used a default install. It hasn’t changed much…

pages/default.txp is long and complex for new users to understand, and could do with some <txp:output_form name="head" /> or <txp::head /> etc. forms to keep it visually manageable. It’s also an opportunity to showcase these critical tags to those starting out.

Similarly the default public site is underwhelming as it is neither a wireframe starting scaffold (too opinionated in its styling), nor a real design. It’s somewhere between the two… The css is not very easy to change (some css variables would go a long way in placing all the critical specifications at the top).

For the demo site we’re missing an opportunity to both show how simple Textpattern can be to use, and a chance to promote all the relevant commentary that has surfaced in this thread in the form of pre-populated content. Multiple sections, articles, in-use custom-fields, and themes to showcase developer staging et al. Images could be svg ‘marketing’ snippets from the homepage etc., or maybe a simple theme eg. yellow flowers so they continue the brand colouring (& ‘dots’ :).

giz · 2025-11-17 23:09:07

Bump my previous post⇡

.

I’ve taken the design prototype as far as I dare using plain html. There is still refinement to do, but that comes later…

Dark Mode is in (I’m still playing with the definition of a ‘dark yellow’), and :focus, :focus-within aren’t finalised (I’m struggling with getting some kind of consistency between Safari & Chrome treatment of the <details> tag). This level of detail is better handled when I refactor the css for production.

@Pete: I hope you don’t mind my taking on your demo login page; I’ve made it part of the main site with the same banner and footer…

On the subject of the banner & footer used throughout the site, does textpattern.com serve these same forms to docs.textpattern.com, plugins.textpattern.com and forum.textpattern.com via curl etc.?

My next step is to use real content and existing structures i.e. a clone of the textpattern.com database. Whilst I could work on a dev template on the live database, it’d be better to keep it separate so the forum can continue with commentary.

In time I’ll publish my fork to GitHub for public perusal, and will flesh-out my cloned database content as far as possible…

Can someone send me a copy of the database?

.

I need advice on how to use the file structure of the main repository textpattern-com-website.

I’ve forked a branch, placed it on my Mac: ~/Sites/local.txpcom/textpattern-com-website
I’ve set up a vanilla install of Textpattern on txpcom.all-sorts.biz, and added a theme named txpcom, exported it to the file system.
Nova is configured so that its local path is textpattern-com-website/src/templates, remote path is /blah/txpcom.all-sorts.biz/themes/txpcom

All good for the txp templates, but now ~/Sites/local.txpcom/textpattern-com-website/src/assets is omitted on the live server. Phil uses Grunt to populate assets and themes etc. into their respective production directories, and Grunt is more than I can chew.

I can restructure the repository so it can live in /themes/txpcom without the need for Grunt, but I suspect this will ruffle many feathers!

How should I proceed?

etc · 2025-11-18 22:08:15

giz wrote #341223:

Can someone send me a copy of the database?

I have no db access anyway, but sharing txp db looks potentially leaky, whatever degree of trust we had. It contains some sensitive data, like pwd hashes and cookie nonces, so few tables need to be purged first.

giz · 2025-11-19 01:47:47

etc wrote #341240:

…so few tables need to be purged first.

Thanks. The primary content tables are all I need. textpattern, categories, sections and maybe images.

gaekwad · 2025-11-19 11:04:41

giz wrote #341223:

@Pete: I hope you don’t mind my taking on your demo login page; I’ve made it part of the main site with the same banner and footer…

The demo site is deliberately on another domain / host so there’s a level of separation for security (along with other considerations). It’s part of my own Textpattern Consulting biz thing, which has been in the skunkworks for some time officially and will launch properly when it makes it up the priority queue far enough. There are other ‘big’ things above it at the moment, but they’re due to launch in spring next year, so fingers crossed after that. Maybe. Ish.

On the subject of the banner & footer used throughout the site, does textpattern.com serve these same forms to docs.textpattern.com, plugins.textpattern.com and forum.textpattern.com via curl etc.?

Each of the sites above is independent of the others, but there are multiple links behind the scenes for sharing of assets, security, optimisations etc.

I appreciate there’s a bunch of work that’s gone into this so far, but this redesign just turns me off and feels like a step back. We’ve got a coherent-ish brand established already, thanks to Phil’s work over the years, and I’m a bit baffled why this is being considered for retirement.

If anything, what we’re lacking is content. Tutorials. Docs. Examples. Themes. Comparitively, a website redesign with bouncing balls is a real head scratcher for me, and churning out some code generated by an LLM seems counter-intuitive to our project ethos of thoughtful, considered code that’s lean, fast, and efficient.

This isn’t gatekeeping, so please don’t take it that way. Perhaps it’s just me that feels this way, but there are plenty of tasks that can be dug into that will take us forward as we go into the Textpattern 4.9 chapter.

giz · 2025-11-19 18:57:21

gaekwad wrote #341253:

The demo site is deliberately on another domain / host so there’s a level of separation for security (along with other considerations).

Exactly; it’s another sub-site. Each sub site should be readily identifiable as part of the Textpattern project.

Each of the sites above is independent of the others, but there are multiple links behind the scenes for sharing of assets, security, optimisations etc.

Are common html elements like the header and footer repeated manually for each sub site, or are they derived from a single source?

I appreciate there’s a bunch of work that’s gone into this so far, but this redesign just turns me off and feels like a step back. We’ve got a coherent-ish brand established already, thanks to Phil’s work over the years, and I’m a bit baffled why this is being considered for retirement.

If anything, what we’re lacking is content. Tutorials. Docs. Examples. Themes. Comparitively, a website redesign with bouncing balls is a real head scratcher for me, and churning out some code generated by an LLM seems counter-intuitive to our project ethos of thoughtful, considered code that’s lean, fast, and efficient.

There is a lot to unpack here. The reason I started this thread is we all know Textpattern is an excellent platform, yet languishes in the uptake of new users. The question remains: why is it largely unknown to most? textpattern.com is where it starts; content & design have dated, and both could do with a refresh.

On the subject of lean, fast, and efficient, the current codebase is not a shining exemplar :p

The hero animation may not be your thing, but many appear to like it. I’m not that fussed if it is dropped, even if it took many, many design iterations to pull off (the opposite of churn).

Thanks for your comments Pete; positive or negative, all discussion is useful as design abhors a vacuum.

Others, please chip in. Eliciting commentary on this thread is like drawing blood from a stone…

gaekwad · 2025-11-19 19:31:15

giz wrote #341259:

Exactly; it’s another sub-site. Each sub site should be readily identifiable as part of the Textpattern project.

It’s another site operated by me (or more specifically, my business), not the project. It’s operationally independent in that sense. It won’t be under the textpattern.com banner since it’s essentially public access and there can’t be any cross contamination should there be a breach. I don’t want any random drive-by miscreants doing anything as an ‘official’ Textpattern action because they’ve popped a lid on a .com property. This is the reason the demo runs on textpattern.co and it’s air gapped from any other project server or domain.

Are common html elements like the header and footer repeated manually for each sub site, or are they derived from a single source?

There are a variety of forms, pages, shared assets, and scripts – much of which is internal to the site and not on a public repo. The design decisions there were made by the development team, so I don’t have full insight as to why each decision was made but I trust their combined expertise.

On the subject of lean, fast, and efficient, the current codebase is not a shining exemplar :p

I disagree. It might not be to your liking, but the site itself is fast to load (see GTmetrix, Core Web Vitals, etc) and performant (effective caching). There are many person hours that go into this project from a CMS angle and the infrastructure, and making it all work together. There’s a reason why we have design patterns – because they give us a coherent look and feel, and changing one thing then means we have to change the other sites in the network.

We can agree to disagree. I wish you well. I’ll leave this thread.

wet · 2025-11-19 21:15:31

We’re invisible. Here’s proof.

Click this link and scroll down.

See where Textpattern ranks? That’s (part of) our problem.

We keep talking about redesigning our website. Making it prettier and more modern. I was nodding along too.

Then I realized something: When was the last time someone who’d never heard of Textpattern barely found our site?

Here’s what really happens: Developers duckduckgo “WordPress alternatives.” They ask on Reddit. They check sites like AlternativeTo. And we don’t show up.

We’re like a great band playing in an empty room while everyone crowds around a mediocre act on the main stage. It’s not about talent. It’s about being where people actually are.

What we can do:

I checked how AlternativeTo works. Their ranking uses three things:

Likes from users with accounts
“Good alternative” votes on other CMS pages
Real comments and reviews

Right now, we barely register. Other systems rank higher because their users actively vote and engage.

Let me spell out the obvious truth: A beautiful website means nothing if nobody visits it. We can have the best code and fastest performance. But if people don’t know we exist when they’re choosing a CMS, none of it matters.

What if we spent the next month just getting visible?

Here’s what you can do:

5 minutes: Make an AlternativeTo account and like Textpattern
10 minutes: Visit the capital_P_dangit and Drupal pages there. Vote for Textpattern as a good alternative
15 minutes: Write a real review about why you use Textpattern

This isn’t manipulation. It’s making sure we’re part of the conversation where people make CMS decisions.

The website matters. But nobody will see it if they don’t know we exist first.

The best marketing we have is us. Right now, we’re the world’s best-kept secret.

Bloke · 2025-11-19 21:20:36

^^ This.

I agree, Robert. 100%.

I’ve liked, voted, suggested it as an alternative, and written a comment/review on alternativeto.

jakob · 2025-11-19 23:01:19

I think all those aspects are important. They need to work in unison. We can extol the virtues of Textpattern on other sites, forums, social media, etc. and get Textpattern added back to hosts’ automatic installers, but once people are guided to the Textpattern homepage, the website has to be engaging and convincing. For that it must be visually appealing, clearly communicate Textpattern’s benefits, and entice people not just to download it, but also to install it and start using it. The homepage is a key part of that, and going beyond that also what Pete mentioned about examples, tutorials and themes that help people do what they want … and in the process get hooked.

I agree with Gary: as attached as we might be to the homepage, to outsiders – those we’re aiming to convince – I’d wager it does look dated. The main four-point-x theme, likewise, serves its purpose but is not the desirable site most people might be looking to install as a starter for their prospective site. One can discuss the virtues of Gary’s design, but I think it’s a step in a good direction, and I think it’s more than about merely being pretty.

Would it not be possible to at least take a subset of a typical Textpattern site, so that Gary has some actual existing content to play with? e.g. make a temporary duplicate of an existing representative site, replace the users in txp_table with a standard_user, delete any contentious content / sections / drafts but leave enough there so that it realistically represents the remaining content. I can’t speak for Phil, but as I mentioned earlier, there may be sensible handover points between what Gary’s and Phil’s css handle, or it may be possible to carry over Phil’s fundamental groundwork into a new build structure that works for everyone. Sure, there’s plenty of work in all that, and also many hands :-)

etc · 2025-11-20 09:36:38

jakob wrote #341267:

I think all those aspects are important.

So do I, but, about ten years ago, txp regularly appeared in top10 cms choices. This alone did not help it to gain much popularity. Txp is a niche cms, for ‘lightweight and fast’ diy minded people. Every time I convinced someone else to try it, that was not really successful. Non-tech people looked for buttons to click (textile what?), youngsters were not even able to install it (where is this docker thingy?). It was like introducing TeX to Word users.

Would it not be possible to at least take a subset of a typical Textpattern site, so that Gary has some actual existing content to play with?

Isn’t that what txp designer account and dev line are for? Should we tighten it, to prevent designers from accidental live assets modifications?

Textpattern CMS

Textpattern CMS support forum

#76 2025-10-10 04:02:52

Re: Textpattern's face to the public

#77 2025-10-10 07:45:12

Re: Textpattern's face to the public

#78 2025-10-10 15:09:50

Re: Textpattern's face to the public

Bloke wrote #340903:

#79 2025-10-10 19:54:16

Re: Textpattern's face to the public

#80 2025-10-11 19:43:48

Re: Textpattern's face to the public

#81 2025-11-17 23:09:07

Re: Textpattern's face to the public

#82 2025-11-18 22:08:15

Re: Textpattern's face to the public

giz wrote #341223:

#83 2025-11-19 01:47:47

Re: Textpattern's face to the public

etc wrote #341240:

#84 2025-11-19 11:04:41

Re: Textpattern's face to the public

giz wrote #341223:

#85 2025-11-19 18:57:21

Re: Textpattern's face to the public

gaekwad wrote #341253:

#86 2025-11-19 19:31:15

Re: Textpattern's face to the public

giz wrote #341259:

#87 2025-11-19 21:15:31

Re: Textpattern's face to the public

#88 2025-11-19 21:20:36

Re: Textpattern's face to the public

#89 2025-11-19 23:01:19

Re: Textpattern's face to the public

#90 2025-11-20 09:36:38

Re: Textpattern's face to the public

jakob wrote #341267:

Board footer