Magic Web Fairies

It’s been awhile and since I just sent the URL of this “blog” to someone, I feel obliged to have a new entry.

Plone conference videos are up: http://plone.org/events/conferences/seattle-2006/presentations/session-videos

Absolutely see Eben Moglen’s keynote. You’ll be glad you did.

In other news, right now I’m interested in microformats. Alyisa mentioned them the other day. I hope it’s not going to be one of those things that seems like a good idea and we all talk about but never really manifests. But this is really an aside and what I’d really like to rant a bit upon is the separation of the spirit of Web 2.0 from the buzzword signifiers.

What I’m seeing, and this is not a surprise in any way, that as the IBMs and Suns of the world embrace blogging and podcasting and Second Life and such everyone wants to have a “Web 2.0″ site. And when they say that they really mean exactly and specifically blogging or podcasting or RSS or video. And that’s not quite it, is it?

When I think Web 2.0, my brain does this…

Markets are conversations

Conversations build relationships

Relationships organically provide sales

I find that I’m not necessarily thinking about a specific Web application, scripting language or technology. I’m thinking about connection and about language. I’m thinking that the Taco Bell CEO “our food is safe” TV commercial is more Web 2.0 in spirit than some corporate podcast shows which are Web 2.0 in “format.” (Although TB still missed out on some other Web 2.0 opportunites - namely YouTube.)

As the push becomes widespread to Web-2-ify sites, it’s critical that we go back to the spirit of this as our guide so we don’t just add more noise in an inherently noisy cyber market.

Practical application

As a Web developer, if you have any input at the strategic level for your organization, here are a few questions you may want to ask pre-scoping that will help keep the project aligned with the spirit of Web 2.0 while reducing your risk by level-setting what that really means.

Are we truly open to listening to customers or are we viewing this as another mechanism to speak “at” them? If the quiet expectation is around the second part, recommend that they produce more white papers and Webinars instead. Seriously.

Will we be able to execute on this input? It’s not enough to get the input, you must have a process in place in which you not only respond to the input but also actually improve your product as a result. This part of the relationship takes work and time. In fact, it takes a designated resource - don’t let this essential component slip out of the scoping phase for the project (more on resources below).

Will we be providing a needed service or useful information for the customer or potential customer? Isn’t that an age old question? If the plan is to distribute press releases throught RSS, tell them not to be disappointed if subscription numbers are low. Content is King in Web 2.0 with regards to aggregation. Nothing, absolutely nothing, is more important than madly useful information. Oh and let them know… everyone loves the delicious snack that is the “how to.” I’ll take one “How To Achieve X with Product Y” over 30 press releases any day. Remember that in Web 2.0 land, people have the power to filter information and will choose the feeds that provide the most value to them.

What tone will we be speaking in? Always be on guard for “corporate speak” sneaking in - it’s a hard habit to change and it is no longer an effective method of communication. It’s critical that bloggers and podcasters understand from the beginning that they should speak in an informal tone (informal does not mean unprofessional - there is a balance). And if article content to be served through RSS is just another marketing-toned set of white papers you may have difficulty finding an audience.

What are the expectations at the management level around metrics? This is important to address up front. Relationships are not quantifiable and you may not be able to draw a direct line from your blogging efforts or podcasts to sales. At best perhaps, you may be able to show that you have assisted with pre-qualification but even that will require that your Web people have an exceptional relationship with the company’s sales force.

Do we have the resources to maintain a Web 2.0 project? These projects are labors of love and require a long term committment to be successful. They will require new or expanded roles. Blogs: Who will blog? Who will train the bloggers? Who will moderate comments on our blogs? Who will derive value from the comments? Who will keep up with the ever expanding feature add to this technology? Who will search the blogosphere for comments about our company? Who will respond to those comments? Podcasts: Who will decide the editorial direction? Who will host the podcast? Who will record? Who will mix? Who will we interview? Who writes the interview questions? Who arranges the recording schedule? Online Chat: Who will be available to answer chat? Who will follow up? Articles (with comments or ratings): Who will respond to comments? Who will qualify social tagging as editorial guide?

So those are just a few things to think about. They certainly don’t add up to all the questions I’d ask during scoping but they can provide enough level setting to get everyone on the same start page. If you really want to give people a culture shock, have them read the Cluetrain Manifesto as it seems the more we talk about Web 2.0 “technology”, the less we talk about these underlying concepts. But it’s the concept or spirit of this place in cyber space that make it important.

Went to lunch with Myroslav Opry, Robert Nagel (Houston) and Jonah who’s last name I can’t spell. We ate Indian buffet and Myroslav was traumatized. He’s having trouble finding things he would consider edible. Where can you get chicken soup around here? He’ll eat that without experiencing abject terror I think.

I whined more about stats and audios at Myroslav who has got to be tired of that by now. Discussion revolved around podcasting, travel, food, newsmastering, and most interestingly of displaying RSS feeds in Plone pages (a source of personal pain) .

I was of course late for Nate’s talk so didn’t have a plug for the stupid Dell with the crappy battery so had to actually WRITE DOWN my notes. Meaning that this one may be even less coherent that usual because my handwriting has been illegible for years now. Sorry Nate.

1:30 Multimedia and Podcasting with Plone

Nate Aune

Melodeo will deliver audio to mobile phones. (How cool is that? Is this a service. I need to look at this.)

Podcasting is growing significantly (he had lots of stats and I wrote none of them down. Must get presentation later.)

Plone OoB

• File content type is no good for audio files - no metadata
• RSS is only 1.0

ATAudio and qRSS2Syndication helped quite a bit.

Actually I’d like to plug qRSS2Syndication here as we did use it with great success ATAudio pains not withstanding.

Nate mentioned TalkBMC on the success story slide. (Had to tell him afterword that we ran screaming away from ATAudio. I really really hope to find a little bit of time to talk with him further about our woes over this one. Maybe it will help on the new product they are working on.)

He showed a screenshot of TalkBMC’s front page and mentioned in passing about how all the audio sites in his examples had very customized skins. (Oh the pain. Poor Wald at QuintaGroup.)

With Zope 3 they are finding that they can make the multimedia products more extensible and more usable. Zope 3 enables:

• Plone should be smart enough to detect if file type is audio or video
• Should be able to plug in media formats
• Components can be reused outside of plone

Latest product is Plone4ArtistsAudio and the alpha has just been released. Do not deploy on production sites.

New product uses python ID3 library parcer to extract metadata. (The example looks way better than ATAudio looks.)

Edit will allow you to edit ID3 data & will write back to file. (Looks like it’s full ID3v2 support now.)

They’ve adapted a folder content type for a nicer display. Select Activate Media to get a “track listing” view. (Looks OK - could be fleshed out a bit.)

Question about m4a feature support. Answer - Not

yet in this new product.

Question about the id3v2 problems of ATAudio. Answer - They have been resolved in Plone4ArtistsAudio.

You can create podcasts in smart folders so you can set critera and pull them together from different parts of the sites. Again you’d choose activate media to give this a nice view. (hmmm… how can we use this for EL?)

RSS2 in use for this product is using basesyndication from Zope3. Does not support iTunes metadata but that will be included. Quills also uses basesyndication from Zope3.

Question: Can you add description to the podcast? Answer: Not yet but you will be able to in the final product.

A bit on the challenges of large files. (I wrote a page of notes on this which I’m hesitant to type in because of the horrible flash backs it induces. Oh well.)

• Upload through browser is sketchy (hahahaha! Scott are you laughing yet? Falling out of your chair? Cursing the universe? yeah. )
• ZODB is not designed to store large files. (Although the rumor is that zope 2.11 will have BLOB support. How good will this really be?)

Recommendations for uploading

FTP (hahahahahaa! oh you’re killing me)

WebDEV (buggy in IE and on Macs)

Various other suggestions of which Tramline which bypasses Zope & uses Apache sounded familiar and may be what QG did for us to tide us over until we got our MP3 distributor.

Why storing media on an actual File Server is Good For You

• Avoid ZODB bloat
• Expose to third party steaming software
• Easier to backup or move
• Transcode video files

The improvement proposale at Plone.org is PLIP #154.

Roadmap for the products is at plone4artists.org

New product should not create new content types but should use existing Plone content types and just be smarter.

They’re working on Plone4ArtistsLink which identifies remote services lik egoogle, odeo, youtube. (I’m actually pretty interested in keeping an eye on this.)

Also thinking about community feedback products around commenting, rating and tagging.

—————————————–

I could rant but I won’t. Need caffeine.

11:20 Building a Humane CMS with Plone

Joel Burton

Evidently, this is the session to go to if you are interested in humans.

Our inspiration should be less is more
Plone is packed full of features and no site needs them all.

Profile
Focusing on semi-regular content contributor. People who value ease of use over all else.
- Won’t learn much about Plone
- Won’t use system often enough to remember everything
- Ease of use over “bells and whistles”

Goals
Fewer options for content placement
correct choices about content types
power user options just for power users
less geekish complexity
sane workflow/security

How to structure a site
Member’s folder
One way to organize site is by member folder. (We turn this off on all our sites except one internal site but no one uses folder as a place to do more except keep personal notes in their folder)
Very “my documentsy”
Bad URLs
What happens when ownership moves?
Is useful for community sites

How to structure site (2)
Folder like content
Requires users to have permissions here
useful when one person/dept. manages items of that type
example: /events folder

How to structure site (3)
by project / authority
/chemistry
/biology
Location should follow security
generally, the most useful
type-specific folders within
/chemistry/events

Question: users don’t care about urls
we have to be the advocate for them because we understand bad things happen when urls are renamed or changed.

Content in the right place
Restrict what’s addable in folder using “restrict”
link is at bottom of list. if you do only one thing this is the single best thing you can do. Select both addable type and preferred types. the addable type will go under a more more in the list and the preferred types will be listed. Change the descriptions type to fit your users. This will help people understand what the “more” options mean.

this helps with the “no you can’t add the wrong thing” idea

Content in the right place 2
Create specialized folder types
- cloning type in portal_types
- fast and easy
- Creating new Archetype based on other
- more flexible for future
- “Staff Directory” is clone of “folder” (more obvious in search results)
- can be done in archetypes
- can paste in types tool

portal_types, folder type
allowed content types - this is even more important than the adhoc restriction

Question: if you change strings, you will have to retranslate things

Only Folderish stuff at top of site
Never have more than 7 concepts in a place
No individual items, single news items. Only folders go at the top. restrict that only folders can be added at top level

you can restrict what’s addable at the plone site level

Removing “My Folder”
Few sites really want user folders
Just delete the “Members” folder - ther’s no side effects. My folder link goes away. The member folder can be recreated later. You can go under the member data tool and change what you call it (like “staff”)

Specialized Factory Forms
- “”Make more like me”
Helps you direct someone to a certain place and create a certain content type

Obvious choice for content types
“Give things the right title
You can change the title of the type under portal_types. Under the hood it will still have the “programmer name” but to users it will be renamed. If you do this do a catalog refresh. This is where the description is. You get a tooltype mouseover from the description as well.

Same Thing; Different name
Users find “file” confusing
Make separate types:
word doc
powerpoint doc
PDF Doc
This fits peoples brains plus allows separate searching
In the types tool, there’s a field for icon. put image in portal_skins custom

Edit Forms: General
Avoid multiselect boxes
- many people don’t know how to select multiple with control
- list of checkbox is better
- If multiselects are required, explain how to select multiple

Edit forms: dates
People loathe drop-down menus - slower, limited range of dates
Replace with add-on product, DateBox
- Allows textual entry in any form
- Plus ideas like “yesterday”
- immediately turns into canonical format

Edit forms:keywords
most useful tool for organizing sites
Move keywords onto edit form
properties is a ghetto for meta-data.
Call keywords tags to address user fear factor
schemata = default to move

Customizing kupu
Can customize with xml
But easier to do in CSS
#kupu-subscript-button {display: none;}

Remove Button from Kupu
you don’t want definition list, full-screen button (full screen button is esp. confusing)
You might not want tables, left, center, right justify

Kupu style menu
Can add classes per type-of-content - para styles option for wigits

Kupu drawers and your types
New types of content are not discoverable in kupu. go to kupu setup under resource types and ad portal types

Dependent items
Users don’t understand that the image isn’t “in” document
RichDocument keeps images/files inside document

Kupu has ReffieldText type that will catalog images

Power users
make a role for it

Undo
Too complex for most users (I remove this for all but admins)

Properties tab
If you remove keywords then you don’t need it for any but power users

Display menu
Should be removed for power users or changes in the types by giving only one display type

Default page
A tricky concept for many users. often having a body field for folderish item is easier. then you can turn of ability to select default type

Add menu
“restrict” may not be useful. Modify constrain types permission should only be given to power users.

Actions menu
allows people to delete copy and paste. Rename allows people to break links.
Don’t change in permission - some products assume rename capability
Chane in portal_actions and hide in menu

Cut/copy/paste
people can find this confusing. Can add a duplicate option instead.

Contents tab
for some users this may be confusing. Under actions to hide

Sharing tab
should only be shown to power users
or offer a greatly simplified sharing tab

smart folders
too tricky for ordinary users to make
create sample smart folders users can duplicate and modify

keywords
stock sample keywords (make private doc with starter keywords, plus prevents keywords being taken away)
user PloneKeywordManager to clean up keyword space
(will this work with simpleblog?)

http://plonebootcamps.com/resources
————————–

Another session I’m uploading so far after the fact that I hardly remember it. I think we’re getting near brain explosion. Happily Joel’s presentation is available so I need to get that.

Wow I am way behind on blog entries. Been keeping them in notepad and analog notepad. Had to skip the 9am talk due to hotel chaos (have to change rooms) so got out late. Hope it was taped.

OK so on to the 10:30  talk

A Sneak Peak at Plone 3.0

Alexander Limi

Plone 3: Sneak Peek

Release Q1 2007 March hopefully

Various usual disclaimers. Entire presentation could be sci-fi

#1 versioning
history of mods, reverting revisions
differences bet revisions
per-type policies - you can decide which type gets versioned
replaceable backend

#2 in-place staging
let’s you work on one piece of content while another is “live”
buildt on top of cmfeditions
simple UI

#3 locking
uses WebDAV semantics
“stealable” locks - if doc was locked some ago and is no longer appropriate
locks are automatic on any edit action
tells you who locked it and how long ago
(can you make stealable admin only)

#4 Easier sharing
curent sharing page too painful, too long
New approach is streamlined, nice checkboxes

#5 Link integrity
tracks what internal resources are relying on each other
warns you if you try to delete an image that is used in a page, linked to, or in other way used (cool!)

Warning: Do you really want to fuck up your site? (OK it doesn’t really say that.)

#6 Generalized previous/next
if you are in a folder or album and you want to have previous / next navigation you can enable this. How are documents linked together I wonder?

(Note: all of this is production code and needs a little work)

#7 Fieldsets
Tons of tabs in plone, hard to add things. split up border under edit to have categories of information. looks sweet.
no page load when you go over each option. doesn’t go back to the server at all. completely accessible

#8 content rules engine
UI for admins to respond to events
examples:
move to folder on add
send email - can send email - you want to get notification when content is added or deleted
change workflow
pluggable (new events, new rule elements)

you have a rule system in plone where when certain events happen you can trigger other events.
UI in general works without coding but you will likely have to do some coding for more than the basic things

#9 portlets engine
UI for managing portlets
Portlets per:
folder, user, group
infrastructure to write advanced portlets

portlets in plone right now are just templates. you have to write lots of logic inside the portlet. this lets you manage the order of the portlet and you can list things like all items that have been changed since a certain date. (we need this)

#10 indexing support for Word/PDF
Not sure of state of this. if you put a word doc in it’s searchable and parseable.

#11 OpenID support
decentralized login/identity system
lets you use your URL as a login
in use by sites like Livejournal, technorati
plugins exist for wordpress, mediawiki, etc
Backend by companies like Verisign
available today and shipping with Plone 3

Defacto standard for centralized logins. You can use with 2.5 if you want it. It will ship iwth 3.0 but you’ll have to add product.

#12 Workflow improvements
more bundled workflows
workflow control panel

most people can’t tell that plone workflow is not just hardcoded but is actually quite flexible. Nice UI for this. You can remap workflows and tell it which state should correspond to the new state in the new workflow.
workflows that will ship: community workflow, web publ, internal, external

#13 Plone 3 and Zope 3 are in love
All new features using zope 3
Plone 3.0 includes Zope 2.10 + Zope 3.3
depends on CMF 2.1

#14 Better markup support
Wiki syntax support for all markup:
Even Word/OpenOffice
“wiki flavor, no wiki aftertaste” - add some elements of wiki that are nice.

New formats supported:
Markdown
Textile
Making it easier to select between them

#15 Ajax (yay!)
In-place editing
Inline validation
Improved UI and widgets
Less trips to the server
Still fully accessible
Making Plone more efficient

—————————-
Announcements:

Plone.net should be working now. Login is fixed.

picture outside at lunch

—————————–

Questions:
Large file handling?
You don’t want to put large files inside zodb.
Won’t be solved by 3.0 because zope 2.11 won’t be shipped in time.

large file handling thing is hard to get customers to pay for or invest in.
solution is close and supporting those people with resources and $ would go a long way.

pre-release testing
they will issue some alpha releases. do not use on production servers!
Perhaps next week an alpha will be out
the best testing is done by companies with sites with a lot of data, make a copy and trying a migration

how much testing has been done on migrations
2.0 to 2.1 switched all content types over to archetypes
2.1 to 2.5 few problems unless running specialized LDAPs
(generally a trivial migration unless you were running CMF member. but this is no longer a problem because of Remember and Membrane)

What will be deprecated in 3.0
not too much is being removed. gruf support is out for 3.0 so you’ll have to use PAS for that.

Myroslav asks something that I can’t hear
Limi says there was a screw up with 2.5 where a small script was removed and almost all 3rd party products failed. He says that will not happen in 3.0.
Additionally, Themes will not break when you update to 3.0. Adding some CSS for dynamic Ajax UI.
There will be new elements that you will need to potentially style.

Is there anything in 3.0 for placeless content… can you reuse content in several locations?
No. Perhaps can be a part of 3.5. Need a standardized way of doing these taxonomies and handling content in more structured ways than putting them in folders.

——————————–

So those are my notes from the talk which I can hardly remember now. I have no hope whatsoever that they are coherent but I saw a vid camera so we can all watch it later. Limi is hilarious. He and Joel Burton are tied for my favorite Voice of Plone.

3:30 A Needle in a Haystack: Discovering relationships in your content

Ben Saller

Haystack - relates similar content, summarizing incoming info, visualize relationships between content.

Language has inherent ambiguity and can be misinterpreted. We figure out meanings by keeping a lot of context. We are efficient at navigating webs of possible meanings & selecting likely interpretations. We are able to recover from errors and reevaluate prior choices.

Computers and language - Google considers search an AI problem. (ha! bonus points for using “grok” in a sentence.)

Haystack1 does probabilistic analysis of individual tests. Look at a given text figure out a term frequency distribution & extract likely keywords. Relates items by common keywords. Maps well into Plone because we can suggest or automatically apply extracted keywords to content.

Haystack2 was written to contextualize the keyword output of they Haystack1 frequency analysis.

HS2 uses your corpus, your data, to aid in its determination of what are vital concepts.

WordNet groups English words into sets of synonyms called synsets. It’s a huge database.

MultiSemCor takes a large text corpus and a lot of stat analysis and tags. Can figure out use of terms used in context.

With WN you can take articles say about angels and devils and connect them as both being about spiritual beings.

(Sigh. Another person who makes itty bitty slide graphics. I really need to start sitting up front with the good students.)

This type of system isn’t going to give you 100% accuracy. Maybe 80% or so. What you can get is a top level view of your content’s relationships.

Practical applications - personalization (affinity, matching) , coverage (which concepts are underrepresented in your corpus, visualization (hierarchies & clusters)

Issues - if any phase fails it can disrupt later phases. It’s harder to go back and reevaluate old info based on new (but Haystack will try). WordNet lacks coverage of proper nouns and many domain specific senses.

——————-

OK another one without any proofing. Did run spell check for once so there should be less pain overall.

After the LinguaPlone session I talked with Donna Snow and Amy Chan of Disney a bit about the various woes of blogging and Plone. Amy and I went to lunch at a fantastic Thai restaurant and talked mostly about Plone but I also tortured her a bit with my geek Anime/Video game obsessions. She was an excellent sport about that. Tom - you did a podcast with a coworker of Amy’s at Disney about their Plone / Vignette implementation.

2:20 Open Source Best Practices

Munwar Shariff

I’m in this session because this is the person who mentioned during yesterday’s Plone and the Enterprise case studies that he implemented a Plone site for 15 million users. So anything he has to say after that statement, I definitely want to hear. He’s the CTO of Cignex who have been involved in over 50 Plone/Zope/etc. open source projects. He is co-author of Plonelive book.

#1 Understand business users & requirements

• Get the big picture first?
• Who are going to use the system?
• Why?
• What do they expect?

It is all about expectations management.

This phase has to be completely non-technical. Send someone who doesn’t know Plone at all.

#2 Documentation is critical for open source projects

Mandatory Documents

• Requirements
• Acceptance Criteria
• Detailed Design
• System Installation
• Maintenance & Upgrade

#3 Always use storyboards for better understanding

#4 Always use centralized membership system

• Centralized control
• Centralized Identity Management
• One copy of data to manage
• One person / department in-charge

Use plugins in Plone to traditional DB.

#5 Define workflow & security upfront

More than 50% of your requirements speak to workflow and security.

#6 Don’t get into the hourly rates trap

Get expertise in the initial architecture phase even if it’s very expensive

#7 Choose the right open source product

There is no one size fits all solution. It’s all about right fit & compatibility.
Make sure there is time to evaluate and comparing products.

#8 Understand the licensing model

• How do you want to use the product/project?
• Internal/Distribute?
• GPL - you must release any changes you make to the code base back to the community if you are redistributing the product.

It is all about protecting yourself from legal issues. You often use more than one open source product in your project, the legal terms change when you add other products with different legal terms.

#9 Choose hardware & OS upfront

• Shall we wait till Production time frame? Scream NO! (Yay we do this!)

• High-availability
• Performance
• Dev -> Staging -> Production - in one server definitely creates problems.

#10 Provide project status updates regularly

• All the parties involved in the project should be on the same page
• Understand issues upfront
• Get help required to mitigate risks

This is the one best practice you must follow.

(Update format: What are the things accomplished last week, what are we doing this week, what is planned for next week.)

#11 Follow coding standards

Will save money in long term maintenance

#12 Write test case before coding

• Set the expectations right with the developers
• Clear technical understanding of the system

#13 Always use file system based development

Easy to debug,upgrade, maintain, control, reuse

#14 Leverage community for testing

Release 1.0 to community and 1.1 will be a stable release with Bug fixes
#15 Define production update process

Development -> test -> staging -> production

content management: acquire, store, manage, deliver

#16 Define maintenance & upgrade process

Log file analysis, Identify broken links, fix 404s

#17 Conduct post deployment performance tuning

No matter how extensively you test the system, most will fail in actual production environment

• Understand load / concurrent users
• Get real performance numbers
• Test integrations
• Tune the system

#18 Empower end users with training

Training for content managers, admins, and developers. Make sure the system is being used.

#19 Provide project end docs

User manual, installation docs, backup/restore procedures

#20 Leverage community support

#21 Contribute back to the community

Good talk. Very dense. Really wish this one was recorded.

10:30 Vive Le Plogs (Plone Blogs)

Donna Snow

Awesome! First slide shows a Penny Arcade cartoon. (PA Volume 2 is out now - go buy it.)

Lots of introductory What are blogs, why blog stuff. I really want to talk about what people are using in Plone as their Blog add on and the pro’s and cons. Myroslav mentioned Quills yesterday but said that it’s Zope based primarily and has trade offs with SimpleBlog. QG is still going to develop out SimpleBlog. I need to get a feature by feature comparison of these along side WordPress.

OK cool! Comparison of Plone Tools!

CoreBlog2 is the one Donna is using. Likes because it has an good archiving feature. CoreBlog2 has categories (wonder how this is different from SimpleBlog’s categories). Previously was a Zope product that is now compatible with Plone. She prefers the way that coreblog organizes entries to quills. CoreBlog2 works well for a single blogger who blogs often. It has comments moderation options.

Quills good for multiple bloggers in the organization. Quills is “meaty” and has a lot of stuff that are standard to blogging. In version 1.5. Quills has RSS2 support and you can use for podcasting now.

She used SimpleBlog two years ago. She said there’s no development going on which isn’t really true as we’ve been paying QuintaGroup to continue to add features and create associated products. So I mentioned that we were paying for development but it looks like the QG adjustments haven’t been uploaded yet as when she looked up SimpleBlog for the latest version it was quite old. The original developers told her they are no longer supporting it. I’ll have to ask Myroslav in more detail about the future of this product as I got the impression they will continue to work on it for other clients as well.

Plodcasting is a blog for uploading podcast to plone. She hasn’t used it.

She as a demo of a template she made that shows the results pages from the categories. It’s a really nice look. We should steal this concept it’s far superior to the default categories search results style display that we have.

Excellent question about when do you give your client Plone as a solution rather than a blogging only solution. She says that usually she suggests Plone for clients who want to add other content to their site beyond just blogging.

Question about remotely adding entries. She thinks Quills have this and if I remember correctly Myroslav was interested in doing this for SimpleBlog.

11:20 Creating and managing multilingual websites with LinguaPlone

Geir Baekholt

LinguaPlone is for managing multilingual content.

4 audiences

• Site manager/ content editor
• Site visitor
• Product developer
• Translators

Site manager - Easy to create and manage translations and work with and navigate various language versions of content.

Site visitor - Provide content in language the user understands and allow the user to override automatic choice

Product developers - Easy to integrate custom content types, hooks for writing workflows for managing translations

Enable multilingual support

from Products.Archetypes.public import *

Translators - LP has simple UI for translating content in small sites, hooks for making translation workflows

For larger deployments, XLIFF lets you translate site in XML - used for huge amounts of translations.

Recording of LinguaPlone in use - gives a tab for translation which gives a split screen and a language drop down. After you add a translation, a flag shows up in the corner and you can switch translations.

Best Practices for Multilingual sites

• Don’t mix languages. User interface and content in different languages are confusing to users.
• Use workflows to manage translations. LinguaPlone does not yet ship with example workflow - just provides hooks to create workflows. May be providing a real example in next couple of months
• Make someone responsible and create processes.

Flags as language icons work but are not technically correct and can be politically sensitive. LinguaPlone provides a fall back to a selection with words if you need it. This requires that the language list not be in English to work.

Original document is the canonical version so if you update that one you can have a workflow that tags the translations as “tainted” until they are updated as well.

You can share fields across translations - dates, numbers, locations, names s when you update canonical it updates all. languageIndependent = True

Search results on return results in the current language and results that have been marked as language neutral. You can override catalog filer to get results in multiple language Language=['en', 'de', 'no']

Plone 2.1 has automatic id-from-title generation and gives pretty, multilingual URLs.

Getting right language to users. LinguaPlone looks for:

1. Language codes in URL
2. http cookie for override
3. browser language
4. default fallback

Roadmap - kill bugs (there aren’t many bugs), minor improvement, Zope3 improvements allowing significantly streamlined code.

Question - how well does LinguaPlone deal with non-ascii characters? Since Plone works great with UTF-8 it is no problems. (Plone is moving to unicode which will work well also.)

—————–

OK no time to review this for grammar and spelling so sorry for any chaos.

Last night to my great offense (at the time) some one asked me if I was finding the conference to be too technical. In the world of l33t, a front end Web developer is usually barely considered technical and certainly never considered a real programmer (compared to say a Python person) and a female front end developer is even lower on the “geek scale”. (Don’t get me started about pay rates.) So I was wondering at the time that if I had been a male front end developer would I have been asked that question?

After the listening to today’s delightfully schizophrenic keynote, I think the answer would have been, for once, yes. One of the main things I took away was that the Plone community (meaning primarily the developers and the Plone Foundation members) have reached a level a maturity where marketing has become an important topic. Plone is stable, has a solid reputation, has a clear development direction and roadmap, has a base of 200 developers, is now trademarked and the foundation has finally gained official (U.S. non-profit status). So what is missing? Full market reach though a strong marketing message that appeals to business. It’s a big concern - speaking the business language rather than the technical language. So the question on the mind of those involved in furthering Plone’s CMS market share seems to be: Are we speaking the right language? Is the conference too geeky to appeal to businesses? And as a representative of business (BMC was listed on one of the keynote slides under Corporate adopters) I’m a logical person to ask.

So here’s a brain dump of notes from the keynote of stuff I want to remember.

Plone is now five years old!

Plone 3 quick look (there will be a talk on this tomorrow with more details)

• Versioning and staging
• Portlet engine
• Rules engine
• OpenID

Plone 3.5 ideas

• Improved member handling
• Improved media handling (good!)
• Improved doc management (BLOB support) (yay!)
• More flexible themes
• Separate content production (yay!)

Various gossip

• Limi is going to Google as an interface designer. Good for him! Good for Google!
• All Plone consultants are busy - zero capacity left.
• “Plone makes money.”

Other stuff:
As part of the need for a strong marketing message, the Plone Foundation needs a way to track the number of sites using Plone. They are exploring a mechanism to do this.

Plone Foundation needs help with stats analysis.

There’s an initiative to organize Add on products including ratings, dev resources, etc. This is what the weekend Sprints are about.

Plone.net lists providers, case studies, media mentions, Plone sites. I need to login and add our sites.

Didn’t have connection in the hotel last night and gave up in exhaustion dealing with the technical support folks. Finally just crashed out. Looks like it’s back this morning so I’m going to make a few notes from last night before I head over to see the State of Plone keynote.

I ended up attending a Birds of a Feather session for Plone trainers thrown by Joel Burton. It was a great connection and people had wonderful insight about training various skill level sets in Plone. There’s going to be a mailing list for trainers through Plone.org to continue the discussion.

Talked briefly to Alan of Enfold Systems. He recommended we do a Plone / Jive integration with LDAP hookups for DBAzine. He said this would work even with our big CMFMember mess but would just take some work and a bit of pick and choose around info we are gathering in CMF Member. Since this isn’t really the kind a thing three front end developers are prone to doing, we’d need to get QG to help us out if we did this. Frankly I’m not sure I’d put this kind of energy into a site that’s still on 2.0.5 which is showing itself to be vastly outdated.

Alan also mentioned that they did a Plone / WordPress implementation for a company and offered to do a demo. Hope I can track him down today or tomorrow. The conference is packed so full it’s hard to have meeting time with people.

Anyways… running late this a.m. and have to get over to see Limi speak. More later if the conference connection holds up. (There were problems yesterday.)

Jeff treated me to a lovely lunch at an Indian restaurant. It was fun to have lunch with someone who could related to my whining about security flaws in PHPNuke and lack of SEO in SMF forum. Jeff had some problems with Multimedia (video specifically) and Plone himself so had great sympathy over our ATAudio woes which lead to the ditching of that product on TalkBMC.

1:30 Case Studies: Plone for education

This is a panel sessions of 5 guys from different universities talking about how they used Plone to solve needs specific to education. After 5 years doing Web development for UF this is bringing back a lot of memories. One of the guys just mentioned something that hasn’t changed - “IT assumes hey if you work for a university you must be smart” and hands you a folder and FTP to use for collaboration. And that’s it. lol. Budget constrains are still the same.

Ah another reminder from one of the panel members about over customizing Plone and getting stuck on 2.0.5. I was just talking about that with Jeff over lunch around that problem with DBAzine and my concerns about over customizing Plone.

2:20 Case Studies: Plone for large enterprise

Munwar Shariff

Another panel - Jeff Watts from NI is on it and I see Alan from Enfold Systems.

Jeff’s story about selling Plone inside NI is pretty amazing (ie funny if you’ve worked for a big corporation). Quite a bit different from us since we didn’t sell Plone so much as sell Blogging/Podcasting … “oh and by the way they’ll be in Plone.” We also flew under the corporate radar quite a bit since we had already transfered the HTML-based DBAzine to Plone. Something we could do because DBAzine was already hosted outside of corporate IT and we didn’t have to create any type of ROI proposal to make any changes to it.

Munwar is speaking about selling Plone to enterprises by selling the value of plone not the technology of plone. He worked on a project where plone scaled to millions of users. 15 million in fact. I’m a bit in shock over that and would like to hear more.

They just gave away an award for something around Plone implementation. I’m not sure what they’re talking about with this. But the winner got an iPod nano which is very cool.

3:30

Talking a break - there aren’t any sessions I feel I need to attend and I need to catch up on email.

4:20 Plone and Accessibility: A case study

Aaron VanDerlip

Aaron used out of the box plone initially in order not reduce existing accessibility of Plone with modifications. Migrated the site recently from 2.0.5 to 2.5 - this seems to be a running theme at the conference. So much content has been added to the site (500-600 articles) that the nav structure has to be redone. This is really sounding familiar.

He makes an interesting point about taking CMS claims of accessibility at face value without doing the manual testing that is really required to prove actual accessibility. So Aaron did a basic audit of CMF templates and says they are compliant from a code standpoint. (Bonus references to logical meaning of headers and how it they have been implemented in the default plone templates.) If you view a default plone page without the styles you can see a very well optimized display for screen readers which will skip them to links or skip them to content. Nice. Forms in the UI have the Section 508 compliant labels. So if you heavily modify and add form fields you have to remember to add these labels.

CSS pointers

display: none removes element entirely

view: hidden removes from browser display but not from screen reader

Using Kupu seems to give better accessibility than using Epoz. Not really a surprise. Looks like Kupu forces use of alt tags which is definitely a bonus.

————–

OK that’s it for the talks I’m going to today. I’m going to go and decide if there are any birds of a feather sessions I need to attend or if food/beer are calling instead.