Go to main content

Textpattern CMS support forum

You are not logged in. Register | Login | Help

#13 2011-01-23 13:27:22

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

Re: [mention] 2010 Open Source CMS Market Share Report

jsoo wrote:

hcgtv wrote: Our new mantra “No PHP in our templates!”

It is one of Txp’s distinctive features — and an important one. Kind of geeky for a mantra, though :)

Just threw it out there, but the emphasis should be that there’s no raw PHP in Textpattern’s templates. I see a coming backlash at WordPress for having PHP in their templates. It’s not going to come from their hosting operation (WordPress.com), but from users downloading at WordPress.org. These users, a bit more knowlodegeable, are getting tired of getting their sites hacked, security update after security update, and now we’re seeing malicious code in templates, ala MySpace debacle.

Whatever clever slogan we can come up with, we should use it on as many of our own sites to promote how safer the web is if you use Textpattern.

Offline

#14 2011-01-23 14:03:50

els
Moderator
From: The Netherlands
Registered: 2004-06-06
Posts: 7,458

Re: [mention] 2010 Open Source CMS Market Share Report

thebombsite wrote:

If everyone would just Google for textpattern about 20 times per day

Honestly Stuart, I was just about to post exactly this… noticed your post just in time ;)
But surely this is what caused the bad reviews, of course we never google for textpattern, we don’t need to :) Someone should tell them to reconsider their methods: no Google searches == best CMS.

Offline

#15 2011-01-23 14:19:29

jsoo
Plugin Author
From: NC, USA
Registered: 2004-11-15
Posts: 1,793
Website

Re: [mention] 2010 Open Source CMS Market Share Report

hcgtv wrote:

… users downloading at WordPress.org. These users, a bit more knowlodegeable, are getting tired of getting their sites hacked, security update after security update, and now we’re seeing malicious code in templates, ala MySpace debacle.

In fairness, a hypothetical malicious Txp template could include baaaaad code in a <txp:php> block, there to catch the unwary. Yes, raw PHP is disabled by default, but it’s still an opening. Some unwary users might have enabled PHP to work with some other template or for some other reason. Perhaps we should exclude or at least strongly discourage templates containing raw PHP, that is for official or quasi-official templates (e.g. Textgarden).


Code is topiary

Offline

#16 2011-01-23 14:22:49

maverick
Member
From: Southeastern Michigan, USA
Registered: 2005-01-14
Posts: 975
Website

Re: [mention] 2010 Open Source CMS Market Share Report

On the more serious side, I don’t like seeing Txp being lableled “@ risk”. None of us do. But we all seem to understand that we’ve got to take reviews like this w/ a grain of salt.

Their metrics are just that – THEIR metrics. Their interpretation is just that. THEIR interpretation. That’s okay. I seriously doubt their report will make or break any of the software reviewed. Shoot – one of the other top CMSes (imo) out there is Expression Engine, and it didn’t show up anywhere. Why? It’s not open source. Until this year Moveable Type was not included. Why? They said it was primarily a blogging platform. (Yet WordPress was included . . . go figure). Other solid, growing, quality CMSes are not there – because they weren’t viewed as candidates for the top 20. It’s just their take. I might choose other metrics and methodology, arguably more accurate, but there’d be flaws to it.

I wonder what is worth learning from. For example, their metrics showing:

  • Retention. If your drop out rate is close to, equal to or greater than your adoption rate . . . . ouch.
  • people who’ve never heard of Txp. Scary high. No wonder no one googles Txp. You’ve got to know what to google first.
  • the number of links to Txp (their measure of good will). It was one of our better showings.

Last edited by maverick (2011-01-23 14:24:20)

Offline

#17 2011-01-23 15:26:03

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

Re: [mention] 2010 Open Source CMS Market Share Report

jsoo wrote:

In fairness, a hypothetical malicious Txp template could include baaaaad code in a <txp:php> block, there to catch the unwary.

Yes, of course, you can make anything malicious. But our Pages and forms don’t need PHP, where in WordPress’s case, it’s a neccessity.

So a user, with limited knowledge, has to trust the template creator that this loop for posts is ok:

<?php
get_template_part( 'loop', 'index' );
?>

Not to mention the 176 lines that is loop.php, which is full of PHP that even a seasoned programmer has to read carefully so as to come to the conclusion that no bad code resides in it. Here’s just a sampling:

<?php else : ?>
		<div id="post-<?php the_ID(); ?>" <?php post_class(); ?>>
			<h2 class="entry-title"><a href="<?php the_permalink(); ?>" title="<?php printf( esc_attr__( 'Permalink to %s', 'twentyten' ), the_title_attribute( 'echo=0' ) ); ?>" rel="bookmark"><?php the_title(); ?></a></h2>

			<div class="entry-meta">
				<?php twentyten_posted_on(); ?>
			</div><!-- .entry-meta -->

	<?php if ( is_archive() || is_search() ) : // Only display excerpts for archives and search. ?>
			<div class="entry-summary">
				<?php the_excerpt(); ?>
			</div><!-- .entry-summary -->
	<?php else : ?>
			<div class="entry-content">
				<?php the_content( __( 'Continue reading <span class="meta-nav">&rarr;</span>', 'twentyten' ) ); ?>
				<?php wp_link_pages( array( 'before' => '<div class="page-link">' . __( 'Pages:', 'twentyten' ), 'after' => '</div>' ) ); ?>
			</div><!-- .entry-content -->
	<?php endif; ?>

While a Textpattern user just has this to contend with:

<txp:article form="article" limit="10" />

Here’s the article form:

<h2 class="date-header"><txp:posted /></h2>
<div class="post">
<h3 class="post-title"><txp:permlink><txp:title /></txp:permlink></h3>
<div class="post-body">
<txp:body />
</div>
<p class="post-footer">
<em>Posted at <txp:posted format="%I:%M %p" /></em>
</p>
</div>

So it just comes down to educating the public, which I intend on doing by showing a WordPress template side by side with a Textpattern template. You’d be amazed at how many less lines of code a comparable TXP template has than a WordPress template.

Last edited by hcgtv (2011-01-23 15:33:48)

Offline

#18 2011-01-23 16:15:05

jsoo
Plugin Author
From: NC, USA
Registered: 2004-11-15
Posts: 1,793
Website

Re: [mention] 2010 Open Source CMS Market Share Report

I’m not up on current templating implementations. Any reason that this:

<?php
get_template_part( 'loop', 'index' );
?>

can’t become a tag?


Code is topiary

Offline

#19 2011-01-23 16:46:48

Bloke
Developer
From: Leeds, UK
Registered: 2006-01-29
Posts: 10,568
Website GitHub

Re: [mention] 2010 Open Source CMS Market Share Report

maverick wrote:

Hey – you’re not going to have ALL the fun! I did it. I put Textpattarn at Risk!

Hehehe. Thanks for the solidarity. Nothing like a good bit of self-deprecating irony is there :-)

Since we’re equal partners in TXP’s perceived downfall, I’m going to make amends and start to undo the damage. I consider myself — at least to some degree — absolved. Your turn!


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

#20 2011-01-23 19:18:36

els
Moderator
From: The Netherlands
Registered: 2004-06-06
Posts: 7,458

Re: [mention] 2010 Open Source CMS Market Share Report

Nice one, Stef!

Offline

#21 2011-01-24 09:34:49

Destry
Member
From: Haut-Rhin
Registered: 2004-08-04
Posts: 4,753
Website

Re: [mention] 2010 Open Source CMS Market Share Report

I think there are some good takeaways from the report, and it would be foolish to laugh it off as a fluke study. It’s not about Water and Stone’s interpretation, nor are their statistics particularly spurious. On the contrary, it’s a huge undertaking, a good effort, and a wonderful contribution to the industry, despite the fact they might hope to get some work from it too. Can’t blame them for that. If anyone here could produce a more accurate analysis showing Txp in a better situation, do the community a favor and deliver it.

While the survey does miss a technical section in the report that might otherwise address Textpattern’s many great aspects like security, simple semantics, lightweight build, lack of PHP in the templates, and so forth, technical specs was not what the report was about. The report was about Market Share.

If you don’t care about market share, then go back to sleep. But if you’re remotely interested, then I’d suggest reading pages 5-17 (if you didn’t do it) about the methodologies employed, because you all seem to think this was a study done on Google stats alone, and it was far from it. One of the sources of data was a survey to which over 4000 participants replied to. That’s real people giving real answers, not interpretations of google data solely. Other interesting (and valid) sources were used too.

Walk with me while I recap a few things about Txp.

Due to the large number of participants, this year’s survey showed a remarkable diversity…While most of the participants in this year’s survey came from small firms…

3 The survey was promoted by (1) press release; (2) announcement to the various projects involved in the survey; (3) email to last year’s participants; (4) announcements on PacktPub.com; (5) announcements on CMSWire.com. We would like to thank Packt Publications and CMSWire for their continued support of this project.

If you want to influence the conclusions of reports like this, you have to be a source of information, or influence the sources themselves. It’s not going to happen automagically. Next time, be a little more proactive and participate in that survey that contributes to the percentage representing Txp. And if you don’t think you can, take a look at the “Size of Firm” chart on page 6. Firms of 1 person (e.g., freelancers) were 21% of survey participants [from more than 4000 participants — p. 10]. People like you and me, theoretically. That’s a damn good percentage for the solo group, but clearly that percentage needs Txp representation.

ADOPTION RATES

A function of downloads, installations and third party support. (page 7)

Downloads:

On page 8 they say downloads should be one of the most telling sources of data to inform about adoption, but it’s surprisingly difficult to gather. And then the real funny stuff on page 9…

Information was unavailable on the following systems: … Textpattern

Not to point blame, but if you’re in charge of a project, you need to do more than just develop. You also need to respond to the industry. In this case, when they came knocking for your data (a courtesy, actually), you didn’t give it to them. You can’t blame the report for failing you.

Installations (pages 10-11):

Survey data and third-party support data was used to determine installations, which is an even stronger metric of adoption than downloading (suggesting follow-through). The table on page 11 shows Textpattern second from last. And then the painful truth…

We have concerns about the representativeness of this data set. We feel the survey data is likely slanted towards the projects with more active communities, as those communities made an effort to publicize the existence of the survey to their members.

Et voila !

This is an honest statement that recognizes an imbalance to bigger communities. But you can’t blame them for what they have to work with. Increasing market share is a positive feedback process when all factors are optimal. The more you promote, the more the community grows. The more the community grows, the more promotion occurs and the more people are likely to take surveys like these.

Third-party Analysis (pages 12-13):

Two sources were used: W3 Techs (8 Dec 10) and BuiltWith (30 Nov 10).

But…

The W3Techs system does not assess … Textpattern (p. 12) . The BuiltWith results did not show occurrences of …Textpattern (page 13)

Again, they can’t give a score if Textpattern is not represented in the sources. Too bad for Txp. More missed points. Clearly one thing to do here is find out why Txp is not recognized by W3 Techs and BuiltWith.

Third Party Support (pages 14-17):

This looks at developers and publishers to make the inference that the more of these there are, the more interest there is in the system.

For developer support, they used Elance and Guru as sources:

  • Elance: 32 devs, an increase of 129% from 2009.
  • Guru: 24 devs, an increase of 26%

While those are positive increases, other systems had percentages up to 740%. This data suggests that if you provide professional services for Textpattern and want to help promote Textpattern in the metrics of developer support, register yourself with Elance or Guru (it might even help your income).

Books in Print (page 16-17):

This is probably a fairly weak metric. Nevertheless, WordPress, for example, has 174 books in print (compared to Txp’s single title) and 74 were published in 2010. That seems like a waste of paper to me, but an incredible show of community support nonetheless. I’m sure there are some more books to be written about Txp, and especially after Txp 5 is on the market. (I have a funny story about Txp and books, actually, that I’ve never revealed.)

BRAND STRENGTH

This is a very important part of the analysis, and there were many metrics used, but I’ll only go over a few.

Search Engine Ranking (page 19):

They used a specific set of search terms in this metric (not names of projects). I wouldn’t lend to much weight to this one. “Content Management System” is a string of keywords with a lot of baggage hanging on it. The only way to rise to the top is to have a shitload of content online comparative to the big competition. In my opinion, that’s a fools game. It’s much better to focus on producing great content and not worry about search engine ranking. Why? For one thing, if your content is great and structured well, Google will pick up anyway, and for the right reasons. Useful links to good content. But more importantly, there are studies showing that web users are changing the way they find information on the web; no longer is it the Google search form (which for a while now only returns a lot of crap in the top page results produced from content farms), but rather from human-curated content sources like this one, this one, and anywhere else someone takes the time to list and share the best resources about a given topic. Not only do people find curated information more helpful than Google search, but Google will also pick up the curated resources and improve it’s offer too.

(P.S. Time to update that wikipedia page with branding and whatnot, yeah?)

Google Page Rank (page 21):

This seems to come from a rather complicated algorithm or something, and they admit…

the value of PageRank as a meaningful measure of relevance is doubtful. There are well documented cases where PageRank and search ranking do not correlate well.

This goes back to what I was talking about with crap content from content farms getting high ranking. Gaming google is the devil’s game.

Project Site Popularity (page 22):

This was a look at textpattern.com specifically “with the goal of determining which project site has the most traffic.” Data was measured via Alexa, which provides a measurement of a site’s popularity relative to other sites. The lower the value, the higher the ranking. Textpattern fairs pretty badly.

The biggest problem here is not with how .com looks, but rather with it’s content and IA. It’s mostly static back-slapping that’s never going to generate hits if it doesn’t change; certainly not repeat visitors, which are just as important as first-timers if not more so. The blog—the one thing that .com can effectively use to turn things around a bit—is underutilized. I think I mentioned something similar before. Let’s see… hmm-hm-hm… Oh, yes! Here we go under the blog part. That’s kind of long, so let me extract it out:

Ideally…that blog would be the Daily Bugle for everything Textpattern… Someone from the community with a penchant for blogging regularly about community happenings should be recruited as .com’s blog Editor. I could easily imagine bringing in more TXP MAG-type content in the blog channel. That would be refreshing, actually.

Content is not the developers’ fault. What Textpattern.com needs is a real content strategy, and a willingness to adopt and maintain it over time. Devs can’t do that. So they need to let someone step in on that and take control of the content planning, communication and social community management channels that are owned by Txp. Not an easy thing to do, nor an easy person to find (having the qualifications), but if Txp wants to improve in the site popularity metric, content has to be usable (including well organized and clear), fresh and engaging everywhere. Static ROT (redundant, outdated and trivial) content will assuredly kill a website’s popularity with humans and machines alike.

(P.S. It’s a damn shame the TXP Mag is defunct. I can’t believe nobody has adopted that. If .com was smart, it would adopt the mag. I would love to do it if I had more time, but if a group of people wanted to do it together and share the load, and put content before design, I’d be game to help.)

Share of Voice (page 26):

This is essentially social media channels (bookmarking, twitter, facebook, etc), including blogs. The only channel Txp is respectable in is bookmarking which probably relates with Delicious (whose own future is now uncertain). In every other social channel, there’s relative silence about Textpattern. This doesn’t surprise me considering this Web 1.0 bulletin board is everyone’s preferred hideout.

Abandonment (page 34):

This shows there’s a strong trend of Txp users leaving Txp, somewhere around 80% of those responding to the Textpattern part of the enquiry. That’s people telling them in a survey they tried and left Txp. Period. Like a hole in the water bucket. Wouldn’t it be nice to know why they left? What could you say/offer to people to get them to tell you? (Example of missing but needed community management strategy.)

There’s many other metrics, but you can look at the report yourself. Let me jump ahead to the page Michael mentioned at the head of this thread (page 50)…

We listed Textpattern as a Project at Risk in the 2009 Report.26

26 In that report, we noted that the one bright spot was social media mentions. What we discovered in this course of this year’s research was that the social media statistic reported in 2009 was distorted by a Textpattern widget that included links to Textpattern in every post. In other words, the one bright spot we thought existed was in fact rather less persuasive.

I’d sure like to know what the “widget” is they are talking about, but that footnote says two things to me. First that social media is (and we should already know this) a powerful influencer with respect to brand recognition and search engine rankings. Second, the widget, whatever it is they’re talking about, was the wrong way to go about promotion. Sounds a lot like trying to game Google and that’s a pretty dumb thing to try. Promotion needs to be done with smart content and human effort. Google likes both of those.

CONCLUSION

The reports concluding remarks about Txp are hard to hear and accept, no doubt, but this should be an eye-opener for this project. Shrugging it off as irrelevant might just lend weight to their forecast (certain doom). You may know/think differently about Textpattern, being one of the few fanatics..er users, but those reports are very influential to people in the wild; people who make CMS decisions for their company and clients based on information they find. There’s clearly a lot of work to do if anyone cares at all about the outside having a better impression of Textpattern, and Txp gaining a stronger market presence as a result.

But there’s the rub, the community has to actually care. I know you care about Textpattern itself, simply because it’s running your sites or you’ve invested a fair bit of time into the development or whatever, but I don’t think many here actually care beyond their own use of the system. Nobody really seems to care about community growth, and that, at the very root of it, is what’s causing the downward trend in this graph. (Btw, when was it Dean bowed out of Textpattern? Wasn’t it about first quarter 2005?)

We know there are many great things to appreciate about Textpattern, but mere existence of these things isn’t enough, it needs to be promoted, vigorously, if you care.

Offline

#22 2011-01-24 10:19:08

Bloke
Developer
From: Leeds, UK
Registered: 2006-01-29
Posts: 10,568
Website GitHub

Re: [mention] 2010 Open Source CMS Market Share Report

Exactly. Thank you for your considerable talents and views at distilling the fact to the important parts.

In the same vein, our Wikipedia entry will always have the warning about ‘This article needs references that appear in reliable third-party publications’ unless something changes. We’re not going to get that all the while we’re insular and focused on developing in our own little worlds, and nobody outside the community is writing about us. (btw, I don’t have a wikipedia account and even if I did and updated the entry it’d likely be rejected by the content gestapo because of my bias and involvement with the project. I could register a pseudonym I suppose if nobody else wants to alter the page and brand it accordingly).

As you rightly say, people need to step up and offer editorial services. I’m willing to do whatever it takes to get more community involvement. If that requires a way for people to write about TXP on .com or in TXPMag then let’s do it. To facilitate that is just a simple tech change; to actually do it requires willing participants. Neither is a step change, just an evolution of roles to open up what we have to more a more decentralized base.

For TXPMag, a group of people makes a lot of sense given the amount of free time a single person has. I can do some of it if you wanna join forces with a few others: I’m already a bit stretched so I’d be more of a sleeping partner (heck if I spend any more time in TXP land my wife’ll probably leave me) but I’m more than willing to play my part in promotion. It just needs someone with better leadership skills than I have.

btw, this:

In this case, when they came knocking for your data (a courtesy, actually), you didn’t give it to them.

If I’d seen any communication I’d have responded. Perhaps the contact form is going to someone else or it got missed somehow? *shrug* Should probably investigate that pronto!

Since I’m not in the industry know-how (not being the world’s greatest designer, nor a professional developer by trade) I’d never heard of Elance, Guru, W3 Techs, BuiltWith, etc. If you think it’ll help I can register with the latter two, and pretend to whore myself for cash with the former two. I’d probably never respond to a tender because that sort of thing doesn’t interest me in the slightest, but if merely representing myself on there and listing TXP as a skillset helps give it visibility then I’ll do it; and I’d suggest others here do likewise.

[ no use relying on my web site for clients: it’s useless ROT (certainly the ‘T’ part!) and I don’t use Twitter or Facebook personally, primarily because I prefer to ridicule myself privately on my own space ]

But when it comes down to it I’m a do-er not a planner. I have the leadership skills of two badgers taped together and dumped in a mineshaft. If someone more knowledgeable than me tells me how to make a difference I’ll damn well make a difference. For goodness sake, someone — anyone — with a good eye for projects, get your sheepdog whistle out and herd the flock! Baaaaa.

Last edited by Bloke (2011-01-24 10:25:33)


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

#23 2011-01-24 10:22:49

colak
Admin
From: Cyprus
Registered: 2004-11-20
Posts: 8,776
Website GitHub Twitter

Re: [mention] 2010 Open Source CMS Market Share Report

For TXPMag, a group of people makes a lot of sense given the amount of free time a single person has.

This should also go for welovetxp which had no entries since Aug 2010


Yiannis
——————————
neme.org | hblack.net | State Machines | NeMe @ github
I do my best editing after I click on the submit button.

Offline

#24 2011-01-24 11:19:05

Destry
Member
From: Haut-Rhin
Registered: 2004-08-04
Posts: 4,753
Website

Re: [mention] 2010 Open Source CMS Market Share Report

Note to devs: I was not speaking to you specifically. That was a community broadcast. :)

Everybody in this community needs to contribute to market share. Some already do, and some provide the lion’s portion, but a lot of others don’t.

@Bloke: So true about wikipedia. Editors there are like border guards. I was just using that as an example of a curated piece (barriers aside). What might be good is another Google knol in English. And a revised wiki home page, etc. :)

I know what you mean about spread thin. We all do. That’s why I said it’s not easy finding those people. But for the record, I’m working on some things in that direction. It’s just a question of time. They include: an editorial style guide for documentation (which could be used for .com, theoretically), an audit of the FAQs (have to start somewhere), and an audit of the Forum (which would aid migration, among other things). I’ll make these public at some point as soon as they’re a bit further along, or enough so that others can help. You, my friend, keep doing what you are doing, which is tremendous!

Offline

Board footer

Powered by FluxBB