Go to main content

Textpattern CMS support forum

You are not logged in. Register | Login | Help

#11 2015-06-24 13:22:38

Bloke
Developer
From: Leeds, UK
Registered: 2006-01-29
Posts: 9,839
Website

Re: Textpattern vs WordPress Based on the Real Using Experience

hcgtv wrote #291953:

rm -r /home/www/wordpress

lol :-)

Nice comparison though, thanks Bert. Based on these findings, it seems a stock Textpattern has around four times the mean transfer rate out of the gate than WordPress, despite the latter having a smaller document size. I call that a result and we hope to maintain or improve that with an improved parser and optimisations at some point.


The smd plugin menagerie — for when you need one more gribble of power from Textpattern. Bleeding-edge code available on GitHub.

Txp Builders – finely-crafted code, design and Txp

Offline

#12 2015-06-24 14:29:29

gaekwad
Admin
From: People's Republic of Cornwall
Registered: 2005-11-19
Posts: 3,252

Re: Textpattern vs WordPress Based on the Real Using Experience

Bloke wrote #291955:

I call that a result and we hope to maintain or improve that with an improved parser and optimisations at some point.

I secretly (OK, not so secretly) wish that one day the Textpattern vs WordPress stop. They are not helpful, in the main, because they’re aimed at two very different demographics each with wildly different needs. A stock Textpattern vs a stock Wordpress? Nobody I know of has either of those in production.

WordPress was a fork of b2. I know this because I was a user of b2 when Matt Mullenweg forked it and announced WordPress v1. The emphasis is, and always has been, on ease of use and being accessible to as many people as possible to make websites as easy as possible. As a result of this, WordPress powers approximately 1 in 4 websites. I’ve seen stats to say that it accounts for 3 out of every 5 websites where the CMS is known. That’s amazing and terrifying at the same time.

I have heard anecdotally that the quality of code in Wordpress is not great. Some people say it’s bad. I know anecdotally, and also from some previous comparisons that Wordpress (typically) uses more memory than Textpattern out of the gate.

Automattic, the folks behind WordPress, pump a lot of time, resources and developers into making something that makes them a whole bunch of money. People spin up Wordpress, apply a theme, optionally add some plugins where core code doesn’t quite cut it, and they’re on the air. That’s it. Done. Save for the scintillating blog content and cat pics, that’s it. Wordpress runs on low-end hosting, Textpattern runs on low-end hosting. Most people use low-end hosting. Go figure.

When I upload a new Wordpress for my clients, about 18MB goes up the pipes to get up and running. Textpattern is about 2MB, and I know I can strip about 30% of that back if I need a super-minimal Textpattern instance on a very slow connection. Textpattern is, as near as makes no difference, about one ninth the size of Wordpress. There is no fair feature/speed comparison here, whatever way you look at it.

For me, it’s as much about the bytes in a file as the people involved. From my perspective, I know Robert, Stef, Phil and Jukka are on the board of development and design. They all know far more than I do about all kinds of things. They all have day jobs (I assume), lives outside of Textpattern and all the drama that accompanies it. They all have, I’m sure, a lot more on their Textpattern to-do lists that either a) I know about or b) they tell everyone about.

I have a tribal loyalty to Textpattern. It’s what I’ve used for almost a decade. I write about Textpattern. I teach people how to use it properly. I make website for people. I fix other websites where a dev has set it up, lost interest, and left the client in a yellowy 4.3-era hinterland. I come here when I can (which is not as often as I’d like) and see what’s happening, try to help a few people and delete a glob of spam if I get here before 0900UTC. That’s my current involvement.

I find the whole Textpattern project tricky, personally challenging and often frustrating. It makes me mad, sometimes. I’ve stopped wondering whether it’s just me that gets like this, or whether I give an open source software project too much space in my head. When I met up with Stef a month or so ago, we talked about Textpattern and I left the pub feeling a bit bewildered. There’s a hell of a lot achieved from the time and expertise of relatively few people. That’s also amazing and terrifying at the same time. The respect I have for the developers and designer is high, but I am always reminded that there are 4 people on that list. Four. How many tens of thousands of sites ultimately depend on those people? I don’t know. I make sure my Textpatterns ping textpattern.com to send a heartbeat, but for all I know I could be one of a thousand, ten thousand, hundred thousand, or a million websites. How big is Textpattern? Does it even matter? Shrug.

I wish there was a project manager and that they were more visible. I wish there was less of a reliance on Dean Allen for the textpattern.com domain. I wish there was enough of a project metabolism to gee-up more people to get involved in contributing, which would add more features and better usability, which would get more people involved, which would create the cycle to grow. I wish there were two or three more people actively committing enhancements to the 4.6-dev code so it becomes more of a realistic proposition for the future. I wish I knew how far along the 4.6 release roadmap we are.

I wish. I wish. I wish.

I wish I could actively contribute more useful stuff. I can’t code, and although I’m almost at the point where I’ve convinced myself I should learn, that moment hasn’t yet arrived. I make small changes to language packs, I’ve made attempts at increasing community involvement in various things – none of which appear to have had much traction – and I’m really struggling with my role in this project, if indeed there is one for me. All of this frustration has clearly just boiled over into a forum post that sounds like I’m two clicks away from a ragequit.

Someone tell me I’m overreacting and need a nap.

Offline

#13 2015-06-24 14:54:09

hcgtv
Plugin Author
From: Miami, Florida
Registered: 2005-11-29
Posts: 2,721
Website

Re: Textpattern vs WordPress Based on the Real Using Experience

gaekwad wrote #291967:

I wish. I wish. I wish.

I took an almost 4 year break from the day to day, because I too had the same feelings you eloquently express. If you go back 10 years and trace our involvement, there’s ebbs and flows, good times and WTF times.

My biggest issue has always been the lack of communication, the one or two sentence responses from those on the trigger when you asked a “Where are we heading?”, or “How about we implement this?” forum type question.

How many times have I asked how to get more involved – “Go commit code, we’ll think about”, is the response – time marches on, years go by, you’re right, crickets.

Every popular Open Source project has departments, people in charge of this and that. This allows more growth, developers alleviated from mundane tasks, given more time to code, etc.

I want to be in charge of communication, tell the world what a great piece of software we got here. Hunt down and squash posts like the one that started this thread, eradicate the misinformation.

Bert, Chief Bullshit Wrangler ;)

Offline

#14 2015-06-24 14:55:21

Bloke
Developer
From: Leeds, UK
Registered: 2006-01-29
Posts: 9,839
Website

Re: Textpattern vs WordPress Based on the Real Using Experience

gaekwad wrote #291967:

WordPress powers approximately 1 in 4 websites. I’ve seen stats to say that it accounts for 3 out of every 5 websites where the CMS is known.

Somebody’s counting but I don’t know how they come up with those figures: <meta name="generator"...>? Accessing /textpattern assuming someone hasn’t renamed it in a fit of security-through-obscurity? Visiting /rss and scraping the content? A combination of the above or some other signal? Who knows.

Seems a bit low to me given the number of sites I personally know about, but maybe (as I said to Adi who dug up the above link), if that’s the extent of its use then I (for one) need to do less procrastinating and more… something else, to further Textpattern’s cause.

For me, it’s as much about the bytes in a file as the people involved.

Despite what my plugins look like, I’ve always been an advocate of small is beautiful. That goes back to my Spectrum and Amiga programming days when you had to squeeze every ounce of power out of tape or an 880 KB floppy disk respectively, and do amazing things in 128-512 KB (yes, KB) of RAM.

For me, the more important metric is how we’re comparing to ourselves with each version. In 4.6 that’s not promising right now. I refuse to believe that is the best we can do. That level of performance is unacceptable to me, as both a user and a developer and I shall be doing everything I can to make the 4.6 branch and beyond a worthy series. My mantra is there’s always a better way and everything — everything — I do starts there.

But as Pete says, it’s a lonely road without the support of every community member in whatever capacity you can offer. Heck, it took Bert’s resurfacing after a four-year hiaitus to kick me into some kind of gear regarding themes. That’s a good thing: it’s about ten years overdue. The debate on the forum over features in recent days is what drives me to do what I do in what limited time I have, so thank you everyone who has been involved. I’ve said it time and again, the people here are what make Textpattern more than just software and it’s why I love contributing.

I wish there was a project manager and that they were more visible.

Want a job? :-) I’m serious…


The smd plugin menagerie — for when you need one more gribble of power from Textpattern. Bleeding-edge code available on GitHub.

Txp Builders – finely-crafted code, design and Txp

Offline

#15 2015-06-24 15:01:31

gaekwad
Admin
From: People's Republic of Cornwall
Registered: 2005-11-19
Posts: 3,252

Re: Textpattern vs WordPress Based on the Real Using Experience

hcgtv wrote #291975:

How many times have I asked how to get more involved – “Go commit code, we’ll think about”, is the response – time marches on, years go by, you’re right, crickets.

Credit where it’s due, though – the switch to GitHub has opened up the commit process up significantly, especially since now even non-developers like me can poke around and make changes. I had a 6-month tenure on cdnjs to get to grips with git in a real-world situation, and that helped with understanding how it all plugs together.

Bert, Chief Bullshit Wrangler ;)

You should team up with Destry, you’d be formidable.

Offline

#16 2015-06-24 15:18:38

gaekwad
Admin
From: People's Republic of Cornwall
Registered: 2005-11-19
Posts: 3,252

Re: Textpattern vs WordPress Based on the Real Using Experience

Bloke wrote #291976:

Seems a bit low to me given the number of sites I personally know about, but maybe (as I said to Adi who dug up the above link), if that’s the extent of its use then I (for one) need to do less procrastinating and more… something else, to further Textpattern’s cause.

Textpattern ≠ Stef Dawson. Textpattern ≠ Robert, Stef, Phil and Jukka. Textpattern is you four, the moderators, the community, the weight loss spammers and everyone outside of here that uses and abuses Textpattern on their sites.

Case in point: let’s assume the 4.6 roadmap is mostly accurate – as I write this, there are 41 open issues and 60 closed issues. That’s almost two thirds of the way there, right? Some housekeeping on the issues could probably close a few more to make it more than two thirds of the way to release.

I understand you’ve had some iterations of unlimited custom fields code that weren’t working as well as they could, so were pulled. That’s understandable – but here’s the thing: I haven’t see any code that you’ve done. Maybe I’m not in the right club for it, but I would think posting some code here or on a branch with a request for comment might get some equally smart people looking at it. I know ruud, etc and some other non-developer minds have insightful feedback on many things code-wise. Hell, if it doesn’t work well enough, just etc_query All The Things and have done.

Despite what my plugins look like, I’ve always been an advocate of small is beautiful.

And that’s totally fine, and FWIW I agree, but this Wordpress comparison is nonsensical at the present time.

According to w3techs, there’s a gradual decline in Textpattern’s use and WordPress use is gradually increasing – no surprise with the latter. I want that first graph to level off, then tick up again.

For me, the more important metric is how we’re comparing to ourselves with each version.

This.

Want a job? :-) I’m serious…

Let’s talk offline. I’ll email you.

Offline

#17 2015-06-24 15:59:46

Bloke
Developer
From: Leeds, UK
Registered: 2006-01-29
Posts: 9,839
Website

Re: Textpattern vs WordPress Based on the Real Using Experience

gaekwad wrote #291987:

I haven’t see any code that you’ve done.

No, mea culpa. I have the old version still in a local repo somewhere which I’m cherry picking into the new one. Suppose I could incrementally commit to that and add it as a branch to master, I just wanted to get it into something that looked halfway decent and looked like it might work before leading everyone up the wrong path and having to scrap tonnes of code.

But maybe just publishing what I have so far, its many warts and all, is the best approach. Lemme get it in a state that at least has the previous iteration mostly in place (a few days I would think) and I’ll make it public.


The smd plugin menagerie — for when you need one more gribble of power from Textpattern. Bleeding-edge code available on GitHub.

Txp Builders – finely-crafted code, design and Txp

Offline

#18 2015-06-24 16:11:46

gaekwad
Admin
From: People's Republic of Cornwall
Registered: 2005-11-19
Posts: 3,252

Re: Textpattern vs WordPress Based on the Real Using Experience

Bloke wrote #291991:

I just wanted to get it into something that looked halfway decent and looked like it might work before leading everyone up the wrong path and having to scrap tonnes of code.

Perfection is overrated. It’s taken me a couple of decades to figure that out, trust me. You’re an island if you chew away on code in your surroundings, but I’m certain that people will be able to comment on your code and suggest alterations if it’s viewable somewhere.

But maybe just publishing what I have so far, its many warts and all, is the best approach. Lemme get it in a state that at least has the previous iteration mostly in place (a few days I would think) and I’ll make it public.

Excellent. Thank you.

Offline

#19 2015-06-25 01:44:12

hcgtv
Plugin Author
From: Miami, Florida
Registered: 2005-11-29
Posts: 2,721
Website

Re: Textpattern vs WordPress Based on the Real Using Experience

Disabled the Zend OpCache that is installed by default on PHP 5.6 on Debian “Jessie”.

I have two words, holy sh*t.

Textpattern 4.5.7

bert@jessie:~$ ab -n 100 -c 10 http://jessie/zendstudio/textpattern/
Document Path:          /zendstudio/textpattern/
Document Length:        10017 bytes
Concurrency Level:      10
Time taken for tests:   5.882 seconds
Complete requests:      100
Failed requests:        0
Total transferred:      1031300 bytes
HTML transferred:       1001700 bytes
Requests per second:    17.00 [#/sec] (mean)
Time per request:       588.228 [ms] (mean)
Time per request:       58.823 [ms] (mean, across all concurrent requests)
Transfer rate:          171.21 [Kbytes/sec] received
Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0    3  10.8      0      57
Processing:   267  575 276.6    502    1709
Waiting:      260  536 272.8    466    1667
Total:        267  578 275.9    504    1709
Percentage of the requests served within a certain time (ms)
  50%    504
  66%    531
  75%    549
  80%    559
  90%    659
  95%   1359
  98%   1707
  99%   1709
 100%   1709 (longest request)

WordPress 4.2.2

bert@jessie:~$ ab -n 100 -c 10 http://jessie/zendstudio/wordpress/
Document Path:          /zendstudio/wordpress/
Document Length:        9148 bytes
Concurrency Level:      10
Time taken for tests:   31.366 seconds
Complete requests:      100
Failed requests:        0
Total transferred:      943200 bytes
HTML transferred:       914800 bytes
Requests per second:    3.19 [#/sec] (mean)
Time per request:       3136.634 [ms] (mean)
Time per request:       313.663 [ms] (mean, across all concurrent requests)
Transfer rate:          29.37 [Kbytes/sec] received
Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0    4  14.6      0      73
Processing:  1014 3078 1406.1   2450    6927
Waiting:      980 2908 1328.1   2317    6569
Total:       1014 3083 1413.0   2450    6927
Percentage of the requests served within a certain time (ms)
  50%   2450
  66%   3199
  75%   3561
  80%   4385
  90%   5347
  95%   6593
  98%   6901
  99%   6927
 100%   6927 (longest request)

The OpCache was made for projects like WordPress:


			OpCache on	OpCache off
Textpattern 4.5.7	1.817 seconds	5.882 seconds
WordPress 4.2.2		6.621 seconds	31.366 seconds

bert@jessie:~$

Offline

#20 2015-06-25 08:32:22

gaekwad
Admin
From: People's Republic of Cornwall
Registered: 2005-11-19
Posts: 3,252

Re: Textpattern vs WordPress Based on the Real Using Experience

Bert, what PHP are you running on that testbed? Does the PHP 5.5+ OpCache give similar results?

Edit: ignore that, saw your sig.

(Asking for a friend)

Last edited by gaekwad (2015-06-25 12:13:58)

Offline

Board footer

Powered by FluxBB