MLP - The Multi-Lingual Publishing Pack

els · 2012-10-02 16:50:11

jens31 wrote:

never mind. im stupid. but lucky.

Care to share what it was? Remember, nobody who is smart enough to use Txp and MLP is stupid, and it might be helpful for others some time.

jens31 · 2012-10-02 17:01:53

thanks for the warm words, els :)
edited my post once again..

els · 2012-10-02 17:02:44

Thank you :)

russs95 · 2012-10-06 10:14:25

I can’t wait til the new MLP is ready! I’ve been using it extensively for years now managing up to a dozen languages and as many collaborative translators. Alas, before finding this forum I tried installing the current MLP on my Textpattern 4.5 site— and it totally, irrevocably crashed it. :-S Good thing it was just a test site! Keep up the good work and keep us posted!

Last edited by russs95 (2012-10-06 10:15:41)

jstubbs · 2012-10-07 01:00:03

It is quite frustrating not to be able to upgrade to 4.5 or 4.5.1 because of MLP, glz_custom_fields and other important plugins. MLP itself is quite buggy though and needs either some love and care or a simplified version for core IMO.

colak · 2012-10-07 05:04:15

jstubbs wrote:

It is quite frustrating not to be able to upgrade to 4.5 or 4.5.1 because of MLP, glz_custom_fields and other important plugins. MLP itself is quite buggy though and needs either some love and care or a simplified version for core IMO.

+1

martinsvihla · 2012-10-09 07:27:40

Hello there,
I have two old TXP sites for an upgrade right now, both of them use MLP. I have to decide if to upgrade TXP to 4.4.1 or wait for MLP compatible with the new TXP 4.5.1. Is there any guess when the new MLP will be available? It would help me deciding:).
Many thanks and regards,
Martin

els · 2012-10-09 19:02:16

It is not for me to say when the update will arrive, but it’s being addressed. If I were you, I’d wait a little :)

Bloke · 2012-10-10 00:46:57

All

A long time coming, thanks for your patience. For those of you awating a 4.5-compatible release, please feel free to download MLP 4.5.x v01. It fixes a number of deficiencies and lapses that have been creeping in over the most recent few Textpattern versions. The remaining known issues have been added to an equivalently-named file in the zip package so please check there before reporting a bug (although anybody who has patches or supporting evidence for the known issues should please raise them here as normal). This version is not backwards compatible with prior versions of Textpattern.

Please note that the l10n and gbp_admin_libraray plugins have changed and will need upgrading, along with the usual uploading of the relevant files from the lib folder.. The language files haven’t changed so you can leave those as you have them.

Mostly, the fixes have been cosmetic to bring the pack (sort of) in line with the altered admin interface, but there’s one new addition. If you have wet_native installed you will notice that the language select lists on the admin side (which are now floated right, top and bottom) disappear. This is to reduce conflict, because overriding the admin-side language when the site administrator has already set your admin-side language using wet_native means the two plugins wrestle and the MLP selection loses out. To alleviate head-scratching, the simplest option was to remove the ability to change language under these circumstances.

Also, as a general note after trawling through the thousands of lines of code and nefarious hacks required to bring this pack to fruition, I’m going to make it my mission to improve the core so we can reach a day where MLP ceases to be a hacky mod and can be realised as a pure plugin. We’re already some of the way there (the pack just doesn’t take advantage of all of the new core stuff yet) but it’ll take some doing, so please bear with us. I think multi-lingual content is important enough to warrant the effort.

Many thanks to Els for the gracious use of her time and server to diagnose, test, report, debug and beta test this version. It would have been a lot more of a struggle without her dedication.

Now go play :-)

tye · 2012-10-10 06:13:08

I don’t use this plugin (yet)…. but good work señor bloke…. it will make lots of people happy :)

Destry · 2012-10-10 08:19:24

Bloke wrote:

I’m going to make it my mission to improve the core so we can reach a day where MLP ceases to be a hacky mod and can be realised as a pure plugin … I think multi-lingual content is important enough to warrant the effort.

+1. You are a giant among dwarves, Stef.

Moving out of the mod zone and into plugin makes multilingual content possible for a lot more people, and thus the CMS a lot more attractive. And as I understand it, and despite MLPs few constraints at the moment against the greater set of callbacks in new Txp, the MLP is one of the more elegant pieces of multilingual code out there, in any CMS. So evolving MLP to plugin status would be tremendous, and something to sing from the mountain tops. If it stops there, we should all be happy.

For the sake of argument (and though quite buried in this ridiculously long thread), I’d like to say a few words on the idea of multlingualism as a core CMS ability. Even if it never happens, that’s fine. I’m not arguing that it should. But one thing that’s always bugged me about “bloat” arguments is they’re based on lines of code alone. Sure, that’s part of it, but not the full picture for certain things. The topic of language is probably the most poignant example.

Bloat is only a legitimate argument when the code in question is for a “mechanical” or “physical” extension of the system. Those are tricky words, but for example: comments, forums, third-party tool modules, etc. You see what I mean? “Bolt-on” pieces of metal for the robot.

But language isn’t a physical bolt-on thing. It’s an inherent part of the human intellect, to thought and communication, to natural expression. It’s characteristic of natural dialogue and content, it’s not a physical content type like comments, or contact forms, for example. Saying that language is bloat is antithetical to CMS design.

Textpattern already has a single language choice preference, why should having a preference for two or more languages be any different? Because most English speakers can’t speak more than one language? I mean, seriously, what is the logic? If that were the case, then logic would suggest you don’t offer any language preference in the CMS at all—everyone has to use English only because that’s the core default. Rubbish. It should be that mono-lingual people select the one language they want (as it is), and multi-lingual people select the additional language(s) they want, and the CMS goes into multinlingual mode.

Again, I’m not arguing that MLP should be core, though it should become a plugin and we should all rejoice for the milestone when that day comes. My comments are more a point of distinction, or a reminder, about how to think about what becomes core or not. Language is not something you put in the attic, or down in the garage until you need it one day. Language is an inherent part of communication, and quite important on the web. More people use their multi-lingual abilities when given the opportunity to do so easily, which, in this case, means removing the barriers. Thinking about it in terms of ‘bloat’ is fallacious, if not a lazy copout.

So how ya’ll doin’ today?

Bloke · 2012-10-10 08:33:50

Destry wrote:

Bloat is only a legitimate argument when the code in question is for a “mechanical” or “physical” extension of the system. <snip>

Good point, well made.

I have some thoughts spinning round the back of my head on how to make (or at least offer) a more seamless multi-lingual experience from core. I really really don’t want to promise anything I can’t deliver but it’s certainly something I’m keen to explore.

We can’t do it alone (heck, I’m a feeble monoglot), so what might be a good idea is for someone (probably me) to start a new thread to discuss the technical barriers that MLP currently hits (scraping through buffers to inject content / having to write a custom DB absctraction layer to inject language markers: bleurgh), how to overcome them with clever core hooks or tables, and also to hammer out exactly how publishing a multi-lingual site should work in terms of publishing and editorial workflow.

I definitely feel that Steve and Graeme’s vision of MLP is still way ahead of the curve in terms of interface. Its integration with Textpattern’s existing “Write first” workflow and the fact it can scale the interface as more languages are added is fabulous. For sure, some aspects can be improved and with some cool jQuery, various bits and bobs could be made better and the interface tightened. So let’s take this kind of discussion out of the bowels of this thread and try to shape what a multi-lingual Textpattern experience should be; which bits should be core, which should be plugin territory and then how to actually go about achieving that.

Sound like a plan?

Last edited by Bloke (2012-10-10 08:36:44)

Textpattern CMS

Textpattern CMS support forum

#2029 2012-10-02 16:50:11

Re: MLP - The Multi-Lingual Publishing Pack

#2030 2012-10-02 17:01:53

Re: MLP - The Multi-Lingual Publishing Pack

#2031 2012-10-02 17:02:44

Re: MLP - The Multi-Lingual Publishing Pack

#2032 2012-10-06 10:14:25

Re: MLP - The Multi-Lingual Publishing Pack

#2033 2012-10-07 01:00:03

Re: MLP - The Multi-Lingual Publishing Pack

#2034 2012-10-07 05:04:15

Re: MLP - The Multi-Lingual Publishing Pack

#2035 2012-10-09 07:27:40

Re: MLP - The Multi-Lingual Publishing Pack

#2036 2012-10-09 19:02:16

Re: MLP - The Multi-Lingual Publishing Pack

#2037 2012-10-10 00:46:57

Re: MLP - The Multi-Lingual Publishing Pack

#2038 2012-10-10 06:13:08

Re: MLP - The Multi-Lingual Publishing Pack

#2039 2012-10-10 08:19:24

Re: MLP - The Multi-Lingual Publishing Pack

#2040 2012-10-10 08:33:50

Re: MLP - The Multi-Lingual Publishing Pack

Board footer