Over the past few weeks, I’ve witnessed many discussions throughout the WordPress community pertaining to the version number of WordPress. Many think that because of the reworked user interface and the large number of features that this version (2.7) will contain, this should allow the software to be bumped up to version 3.0
I asked the WP-Testers mailing list to see if I could get a word from either Ryan or Matt to tell me if 2.7 would remain or if it would be bumped up to 3.0. Jane Wells chimed in saying that Matt doesn’t want to skip version numbers anymore, and that there will be ten increments between integers, so 2.9 > 3.0, not 2.9 > 2.10. Yes, 2.7 is a major change, but the numbers are based on timing, not volume or significance of change.
So there you have it. WordPress 2.7 will be, WordPress 2.7. 2.8 will arrive then 2.9 then 3.0 etc. The feature set has no context when giving it a version number label.
While were on the topic of version numbers, it’s important to note that typically, versions such as 2.7, 2.8 are major releases that usually contain new features or UI improvements. The incremental releases e.g. 2.7.1, 2.7.2 usually contain bug fixes as well as security patches. One last thing to note regarding version numbers. Each major release of WordPress is named after jazz musicians the team admires.
I hope this clears up some of the confusion regarding WordPress versioning.
Let’s hear it for Monica Zetterlund . . .
What is the rationale behind the “ten increments per integer” doctrine? It is insufficient to say “that’s the way mathematics work”, because
a) we’re talking version numbers, not mathematics
b) then there is no reason for the decimals at all, we could just name it “27”, “26” and so on
c) in “real math”, you can’t have two sets of decimals, as in “2.6.4”
I for one would increment the second decimal for minor fixes and security patches, the first decimal for major updates to the existing code, and the integer only when the entire codebase is reworked (from scratch or by an extreme overhaul)
Version numbering is based on a standard set within the project managers and, or developers. It can derive from decimal numbering partly or completely. To say one way is better is going in to a discussion of preference and has no relevance.
(and another comment because I forgot to check the “Notify me of followup comments via e-mail” checkbox)
Ohhhh,
So thats the fact behind 2.7
Keep it up MA.TT 🙂
I have to say, I really don’t see any reasoning for their view. The overwhelming majority of open source projects don’t operate with this type of version numbering system.
It doesn’t matter what other projects are doing. Only WordPress. If the developers say this is the way it is going to be, then that is the way it is. The decision had to be made at some point and it was. At least it wasn’t as bad as WordPress 1.5.1.1.
I agree, but it hasn’t always been done this way with WordPress, so we are losing consistency.
The point is that a standard has been recently established after various changes. It may change again, but it is really what the lead development team decides.
Ubuntu works much the same way, as does OS X (10.1, 10.2, 10.3).
Actually, Ubuntu numbering is based on year and month of the release. They don’t do incremental numbering.
Anyway, I like the WordPress number system. Is clean and easy to follow. I also like the fact that there will be no 2.10 WordPress. Each version number parts should go up to 9, that way you can give your users possible timeline of releases. Or we will be stuck in 2.x revisions for a long time.
Mmm… I don’t really love the way the numbers change, still wished it was 3.0, on the other hand, I like the jazz musician names in every version.
I can’t decide if I blame you Jeff or not for these discussions. I’ll let you know Friday.
Yeah, be sure to call in on Friday so you can yell at me over the air about this post haha! All I did was try and kabash the discussion within the community that 2.7 should be 3.0. I think I did that but look at the resulting discussion hah!
Which is why I’m conflicted. It isn’t your fault people like to debate the merits of a standard. I’ve done it with WordPress in the past, before I realized it is futile.
Then again, you always shoot the messenger.
Completely useless discussion, in my humble opinion.
What does it matter what the version numbering system is?
Agreed. Virtually every project has their own version-numbering scheme, and so it’s gotten to the point where the numbers themselves are irrelevant. Might just as well call this version “Taco,” and the next one “Burrito.” As long as users know there’s a new version and what the changes are, who cares about the rest?
Not a huge fan of doing things that way. It renders meaningless the delimitation between the first and section numbers.
The major version indicates a major feature set release or complete rework. The minor version represents a feature addition or modification. The build represents a revision to that minor version. That’s just standard practice.
The way you are doing it now, it might as well be, as someone mentioned above, version 26 going to version 27.
Information that matters most. 2.6 is a headache right now. WordPress could get a lot of better, I hope.
I thought this was common sense for everyone.
‘Common sense for everyone,’ hasn’t been invented yet.
I just invented that. 🙂
Common sense does not exist …
“At PsychicSoftware dot com we build the software that you want. Call us when you know what you want and we will have it built for you …
Why is no one calling? Seems as though everyone has an idea of what ‘everyone-wants’ … then can’t decide what they want.
No wonder business sucks!”
As long as there is consistency this numbering convention is perfectly fine.
Got to love version numbering discussions. Try the linux kernel numbering scheme. The old linux kernel scheme had even number = stable while odd number = development. That would cause to occasional gottcha.
About the only important thing is that the version number can be parsed for plugin development and automation tasks. So “taco” and “burrito” would not be fun if you have to do chores or development for a specific version.
Well there are arguments for both. Personally I prefer the Complete.Major.Minor method but I don’t suppose it really matters at the end of the day.
At least this way there are no complete or major branches that need continued support.
WP version numbers are still easy to follow compared to other CMSes and operating system like Microsoft. 😉
The naming of Microsoft OS can make one sick, especially Windows 7. Lol.
lol
3.1, 95, 98, ME (Many Error), NT, 2000, XP, Vista and 7.
After 7, what’s next? Windows Final?
No, Windows 7 XP 😛
Since Vista is being referred to as Mojave … I recommend Sahara …
That way when no-one is using it they can all be left out in the desert!
But where it does matter is in the versions that are stable and maintained. We’re up to 2.0.11 for a stable, supported version, but 2.1+ keeps changing both from a codebase point of view and from a UI point of view. Not everyone is prepared/willing/able to upgrade for every release (remember 2.7 will be the second significant admin UI overhaul within a year) which leads to a lot of insecure websites.
Ironically, WordPress 2.7 will allow you to easily upgrade WordPress.
Me to would like to be 3.0 but in other ways is better not skipping the version numbers!
I prefer no numbering at all. Instead, I’d like to see it worded as “WordPress two point seven” (maybe that’ll help confuse the bots).
The numbers are inconsequential to me so long as the software is decent. That said, seeing something like WordPress 2009 might be a little much 😛
Would you have WordPress 2009.1 for the WordPress 2.8 release? WordPress 2009.2 for the WordPress 2.9 release?
Be kind of funny if WordPress 3.0 was called WordPress 2009.3.
You bet. If WordPress 3.0 to be called as WordPress 2009.3, that will sound like March 2009.
Well, I for one have never questioned WordPress’s version numbering scheme. But it took me a while to realize that 2.9 transitions to 3.0.
To be honest, I don’t care about the way in which WordPress is versioned. whole numbers for major versions with x.x.x for security updates or bug fixes is easily understandable for me. Pretty funny to see so many people get up in arms over the way a piece of software is versioned 🙂
This is no gripe or rant just an interesting observation. It fascinates me to see that there is some sort of importance over the Numbered version that WordPress (or any other product for that matter) is on that one would be using. Surely the most important thing is that we all are using a Product that is ‘Active’ in its development. Whether WP is on version 2 or 10, we are seeing the regular moving forward of Features and enhancements and not a stagnating product.
Gareth
“WordPress to the End”
Now that you can automatically upgrade WordPress from the dashboard, (in 2.7); version numbers become less relevant.
The biggest challenge of the WordPress community is getting everyone to upgrade to 2.7
2.9 to 3.0 makes sense, keeps the momentum of WordPress.
Although users will expect 3.0 to be a major milestone release (as opposed to an incremental release) … will there be big features held back in 2.8, 2.9 until 3.0?
What big features? Automattic might have something planned, but for something big like an architecture change would probably take a year of development. I think the small incremental changes and improvements to make WordPress a better CMS and, or platform for other web sites should be done in increments that way, there is less to test.
The way WordPress can improve is better stability and an environment that is less prone to defects. Automattic already has systems in place, it is just extending upon those to improve the core of WordPress.
The only part that might be interesting is whether Automattic decides that version 3.0 is going to mean the removal of deprecated (obsolete) code and files that have been in WordPress for a while. The problem is that with the 2.x releases, people expect old stuff to just work. Version 3.0 might also mean the full inclusion of BackPress, instead of just pieces.
Just speculation on my part however.
I’d completely forgot about the BackPress project. (Found a couple of your posts to wp-hackers from March 2008)
From a developer perspective, BackPress integration should be exciting – it just doesn’t hold enough interest for the every-growing feature-hungry WordPress end-users.
Personally, I’d like to see less in the core and newer features/functionality bundled-in as plug-ins. (so we can activate/deactivate as required)
I don’t care about version numbers at all. I only care about features.
I was fallen in love with WordPress at the first look. Now I am somewhat disappointed about the direction of the development. Instead of focusing on perfomance and architecture issues developers are playing with unsignificant facelifts. With every new version demands are growing. I run site with 3K-6K visitors per day and I am unable to maintain it even with nearly default theme and basic plugins – on a dedicated Xeon server. Another site, with only 500-600 visitors per day is hosted on a common webserver and is constantly a source of trouble for the hoster – it runs all the quotas out. Why not to write the code more efficently? Why not to use some liter version of database engine, e.g. sqllite? If code is a poetry then WP poetry is slightly outdated. I am tired to install more and more caching, optimization, byte-code compilation, database caching plugins and extension to my server.
We always have trade offs, I hope that face lift made for 2.7 will not affect the performance. For sure I will not notice it, my site only have around 300 visitors a day.
I run sites that run the spectrum. 30 a month to 20,000 a month unique. The latter crosses 50,000 pageviews per month. No issues whatsoever. Set up caching correctly, optimize your theme, get your database straight. You shouldn’t have issues.
30K to 20K a month is not the same as 5-6K a day, yeah? If I had 800-1000 visitors a day… But I have 3-6-19-26K visitors a day, with corresponding number of pageviews. The problems is in weak and lazy MySQL database and high CPU load.
The above advice still holds true, however. I run sites that have 30,000+ pageviews. Multiple sites. On the same server.
Use seperate database servers.
Load balance.
Cache everything.
Optimize your theme (don’t make unnecessary database calls for site title etc – this does add up).
Use a php accelerator for speed.
Don’t use shared hosting for busy sites.
WordPress out of the box does use a heavy load, though this is most often database calls made by a theme or plugin(s), and a lack of caching, in my experience. It can be reduced dramatically with the above.
What’s the difference? What if it was version 1,000,000? So what? It’s the software that matters, not the version number. The team wants it 2.7, so it’s 2.7.
hubbub about nothing that matters..granted now that you’ve spelled out the rationale you gotta stick with it ( which also make sense)) but honestly what does the version number mean? So much fuss over a number that most people do not even know the meaning of…TC and keep on working you have a phenomenal piece of software
This decision doesn’t make sense on a couple of levels:
1) They’ve jumped version numbers before, over far more minor upgrades (where’s WordPress 2.4, for example?).
2) What about the maintenance of old code branches, such as 2.0? This is based on siginificant releases, so for this code maintanence to make sense, a major milstone release should be marked with a version jump.
2.7 is likely to break many plug-ins, so it would be nice if the 2.6 branch was maintained for bug fixes. If 2.7 were 3.0, this would happen. But if it’s 2.7, it won’t be considered a new branch, so 2.6 won’t be maintained. This means we’ll have a choice of living with potential future 2.6 security issues, or possibly many broken plug-ins.
Why after 2.6.3 not released to 2.6.4 ~ 2.6.9 then 2.7.0
Personally, I think they should take a leaf out of Ubuntu’s book (and many other distros for that matter) and emphasise the ‘nickname’ more than the version number. Most versioning is arbitrary any-how and having a universal system of versioning is NOT going to happen. Why ram numbers down our throats when we can simply use a good name?
Really doesn’t matter to me what they call it or why – that’s not to say I’m not interested – I just fail to see why anyone needs to get antsy about this. Its just a label, after all, the proof is always in how it works. Any idea when the stable release is expected?
Thanks for the clarification. As a non-developer user, what matters for me is the features implemented in WP and how it’s improving such as speed, loading, dealing with server resources + sql databases, plugins compatibility and so on. Nevertheless, the current numbering system make it easy to recall wether it’s a major upgrade or security/bug patch.
I noticed the ‘codename’ after of releases and it’s good stuff like WordPress Vista are not implemented 😛
When observing the broad spectrum of developer practices, it is obvious that there is no universal standard and that version numbering schemes are arbitrary. So long as the version numbers are a series of incrementally increasing numbers, it really doesn’t matter what the rationale is. I would only be concerned if, for example 4.0 was later followed by 3.9. Now that would be a problem.
I have to agree with at least a few of the views, what’s the big deal on version numbers ? My God, you’d think it was a matter of life or death. If YOU write the next program as popular as WordPress, then YOU get to use your own version numbering scheme. Until then…live with it. I find out what the latest stable version is, install it, and then worry about things that actually matter.
WP 3 will have to have Intense Debate and Poll Daddy as wholly absorbed features… and perhaps a no coder version to customize your theme.
Really 2.7 is building on the 2.5 and 2.6 changes. 1.5 and 2.0 were very different, unless WordPress is rewritten after 2.9 should be 2.10.0
I was also thinking the same that this 2.7 had to be 3.0 Then I found your blog. Now my point is, if they did this much major change, they could just release 2.7 with some minor fixes or updates. Like the present version is 2.6.3, which could be 2.7 and then 2.8-2.9 and this one 3.0.
Well, they are the developers and it’s their decision.
This entire comment thread is absolutely hilarious. One of the highlights has to be “I agree, but it hasn’t always been done this way with WordPress, so we are losing consistency.”
The ships going down fellas! WordPress’ last hoorah! heh
Well, for quite a few people the versioning methodology is just another piece that makes a code & the structure that contains it elegant. Is the versioning system elegant? Is it up to par to the code’s elegance? That’s why there’re so many comments on the versioning and the semantics of the decimals.
What does it really matter? Nobody is paying this guys for the work they are doing. They offer their product free, so let them call it whatever they like. If I don’t like the way the version number is and they don’t like my idea of what it should be, then I should go make my own FREE WORDPRESS 😉
This is a nice post, it generated enough argument to help me laugh some stress off. lol