Rebuilding the Textpattern demo site

gaekwad · 2016-02-07 15:24:40

(Previous demo site discussion here [2014])

The time has arrived for me to build a better Textpattern demo site. The current site is lacking in some areas, and I want to rebuild it from scratch.

There are a bunch of things I want to do properly. In no particular order:

stable release and development release ~~and alpha/beta/release candidates available~~
European, North American, and Asia Pacific points of presence
different versions with as-released content (default article, comment, categories, etc), zero content (all non-essential items removed), and additional content (safe images, files, example articles, comments, etc)

The new demo site will be containerised, with appropriate security, and if the consensus is that the ability to use plugins would be helpful then that functionality will get added to the shopping list.

I’m not a developer, I’m a hacker — and I tend to learn by doing. This will take a while.

I’ve also been out of the Textpattern loop for some time (i.e., depression), and this is an appropriate project for me to warm up with.

Please help me out by listing your requirements (need to have) and feature requests (nice to have). Also, if you can +1 or -1 other people’s replies that will help me gauge what to do.

Thank you.

Last edited by gaekwad (2016-02-08 13:39:32)

bici · 2016-02-08 05:27:30

Sir Gaekwad. I applaud and welcome your project. looking forward to it.

If you do decide to add plugins please had the option for flat file editing of the templates. I think it would be useful.

gaekwad · 2016-02-08 09:21:52

bici wrote #297791:

If you do decide to add plugins please had the option for flat file editing of the templates. I think it would be useful.

Hey bici – I think this may be out of the scope, but I’ll see what I can do.

MarcoK · 2016-02-08 11:14:32

For me i’s a great Idea.
But in my opinion, orginal admin theme is very ugly. Better is use Hive theme. More…. web 2.0 :)

philwareham · 2016-02-08 11:57:38

MarcoK wrote #297799:

For me i’s a great Idea.
But in my opinion, orginal admin theme is very ugly. Better is use Hive theme. More…. web 2.0 :)

If using Textpattern 4.6 for this, then Hive is the only theme that ships with it anyway (although there are 2 variants of it bundled).

gaekwad · 2016-02-08 12:18:08

philwareham wrote #297800:

If using Textpattern 4.6 for this […]

It’ll likely end up being 2 or 3 versions:

current release
dev version from GitHub
beta/RC release (if there is one at that point in time)

So, yeah, it’ll be Hive-friendly from the outset.

jakob · 2016-02-08 13:17:44

Pete, what you have setup now is already great, and this could be even better.

gaekwad wrote #297782:

stable release, development release and alpha/beta/release candidates available

I think it would be great to have a development release. From the replies on the forum, it seems to me that there are plenty of interested forum users who would like to see what’s coming but haven’t installed the dev version, for one reason or another. Sometimes it’s a fear that a particular plugin will stop working on it. In that sense, it would be good if plugins are allowed, but you would need to make sure your demo security can’t be compromised…

Of course, being a work in progress, the dev version could also be broken from time to time. Maybe the dev’s aren’t so keen on that…

Regarding the beta/RC: Is there ever a betas/RC that is markedly different to the current dev version? How often are betas/RCs actually in existence? Maybe the dev version is sufficient.

For beta-testing purposes, it’s better (I imagine) that people test out a beta/RC in as many different real-world situations as possible, e.g. with their data, on their servers, with their plugin combo.

European, North American, and Asia Pacific points of presence

Sounds like a good idea, if it’s not too involved. Maybe that’s a nice to have though. What might be an idea is to have more languages already loaded from the get-go.

different versions with as-released content (default article, comment, categories, etc), zero content (all non-essential items removed), and additional content (safe images, files, example articles, comments, etc)

I think some realistic data and pages is always a good idea if people wan’t to test out a CMS. As much sense as the limited standard articles make sense for the txp distro version (I see so many sites where the standard article, links and comment still exist years later, so people don’t delete them), I think it’s a bit thin on the ground to really show off the potential of txp and the admin area. A good demo site would be much more convincing. But, of course, it is much more work…

An admin dashboard with some “welcome + what can I do now?” infos (don’t they call that “onboarding” nowadays?) might also be an idea, but that needs a plugin, and isn’t part of txp’s standard setup. Perhaps that’s a nice to have too.

rah_flat

As much as I am a fan of this plugin, I can’t imagine how this would work for a demo site. It would need a whole upload mechanism for the flat files. Maybe I’m underestimating how people use a demo site, but my guess is that developers who use flat file templates do that because they already have a local web-dev setup.

gaekwad · 2016-02-08 13:38:13

jakob wrote #297803:

Pete, what you have setup now is already great, and this could be even better.

Thanks, Jakob. If only you could see the scripts holding it together behind the scenes. Yikes…

I think it would be great to have a development release. From the replies on the forum, it seems to me that there are plenty of interested forum users who would like to see what’s coming but haven’t installed the dev version, for one reason or another. Sometimes it’s a fear that a particular plugin will stop working on it. In that sense, it would be good if plugins are allowed, but you would need to make sure your demo security can’t be compromised…

Security is the main concern at the moment, and the motivator for the refresh. I’m building the replacement/successor on top of LXD and, assuming I can get it to what it needs to do, I’ll be locking down each of the containers so they can’t cause damage to anything (container host, other sites, etc), and looking into FastCGI for PHP, among other things. Lots of new things to learn.

The current site is in the US, and is the sole occupant of a Digital Ocean droplet (VPS). No obvious problems or issues at the moment, I’m just keen to lock stuff down where possible.

Of course, being a work in progress, the dev version could also be broken from time to time. Maybe the dev’s aren’t so keen on that…

An excellent point, and one I had considered. The trade off, in my humble opinion, is that more eyeballs on the dev version is a good thing overall – whether it’s overall project enthusiasm, curiosity, or helping to avoid the feedback echo chamber.

Regarding the beta/RC: Is there ever a betas/RC that is markedly different to the current dev version? How often are betas/RCs actually in existence? Maybe the dev version is sufficient.

For beta-testing purposes, it’s better (I imagine) that people test out a beta/RC in as many different real-world situations as possible, e.g. with their data, on their servers, with their plugin combo.

Maybe you’re right. A choice of two would be less work, and you’re absolutely correct about different scenarios. I’ll provisionally remove the beta release idea.

An admin dashboard with some “welcome + what can I do now?” infos (don’t they call that “onboarding” nowadays?) might also be an idea, but that needs a plugin, and isn’t part of txp’s standard setup. Perhaps that’s a nice to have too.

It needs a plugin right now, sure – but something akin to the WordPress dashboard would be a very helpful addition to core.

As much as I am a fan of this plugin [rah_flat], I can’t imagine how this would work for a demo site. It would need a whole upload mechanism for the flat files. Maybe I’m underestimating how people use a demo site, but my guess is that developers who use flat file templates do that because they already have a local web-dev setup.

I agree. The scope of a demo site is pretty narrow, I’d say. Anyone looking for a cheap web dev setup for home of the office could do a lot worse than VirtualBox or a Raspberry Pi with LAMP.

Bloke · 2016-02-08 15:08:12

This is a great move, thanks Pete. And it gave me an excuse at long last to play with the lxd demo. Very nifty, and perfect for virtualising demo sites as you say by setting base snapshots and cloning from them on a schedule or on-demand.

gaekwad wrote #297804:

something akin to the WordPress dashboard would be a very helpful addition to core.

We could do something more formal if it’d be useful. We were using the default article as a sort of front-end dashboard (of sorts), but it’s not as nice as a dedicated back-end dashboard, of which jakob’s installations are the gold standard, imo (believe me, if you’ve ever tried one of his admin-side experiences, it’s something else).

Recall that out of the box we have the Start / Home tab available, which I do use for dashboard purposes. But, uhhh, yeah, it needs a few lines of code to kick it off, or something like smd_tabber running. I’m willing to investigate ways to automate this population of dashboard content if it’d be a neat and broadly useful addition to the admin side experience. One immediate (although currently un-thought-through) method might be via a Form or Page naming convention, or by allowing Forms to be created under a ‘Dashboard’ type or something like that.

philwareham · 2016-02-08 15:41:37

Cheers Pete, hope things are well.

A dashboard could be considered for Textpattern 4.7, it’s something I’ve toyed with locally a bit. I’m keen that we get 4.6 out ASAP though so I’d rather devs (me included) focused on the current 4.6 release blockers and then we can look at dashboard feasibility in earnest after that release.

Regarding using the GitHub master branch, betas, etc. A 4.6 beta (tagged on GitHub) will be released once those blockers are resolved, with a tagged release and 4.6.x branch thereafter. 4.7 should be a quicker dev cycle then 4.6 (like, hopefully later this year instead of the 3+ years it’s taken 4.6!) and betas for that will also be tagged as and when available. I’d say clone off the tags instead of the unstable master branch.

Looking forward to seeing what you do with the demo install – it’s definitely essential to attract new users that this is available.

gaekwad · 2016-02-08 15:45:29

philwareham wrote #297807:

A dashboard could be considered for Textpattern 4.7, it’s something I’ve toyed with locally a bit. I’m keen that we get 4.6 out ASAP though so I’d rather devs (me included) focused on the current 4.6 release blockers and then we can look at dashboard feasibility in earnest after that release.

Totally. It was an idle comment, nothing else. Feature creep is already rearing its ugly head, and it should be stomped on.

I made some good progress yesterday (hooray for weekend projects when the weather is bad), and if time permits I’ll be on it again tonight.

Bloke · 2016-02-08 16:08:09

philwareham wrote #297807:

A dashboard could be considered for Textpattern 4.7

Ummm, yeah. It was an implied >4.6 in my post. Should have specified that, sorry.

Textpattern CMS

Textpattern CMS support forum

#1 2016-02-07 15:24:40