[Linkbat-devel] Linkbat data model, content, etc.

[James M. <lin...@ji...>]

Hi All, 

this is the second half of a conversation I have been having with Shanta 
McBain who is working on the presentation of the data (among other things). 
She reminded me that we need to have these discussions on the list and not in 
private. 

I have created a linkbat-writers mailing list to discuss actual content. Most 
of the people on that list will not be all too interested in the technical 
details of the actual implementation and I don't think you developers are all 
that interested in what goes into which article. However, I'll let you be on 
both lists if you want. 

As another reminder, please, please, please send me comments, questions, etc 
about the data model as soon as possible. 

On Monday 18 November 2002 17:41, Shanta McBain wrote:
> > Here are the content pages for the tutorial. With the data files and code
> > that I sent last week, you should have enough to get a working demo
> > running using eXtropia.  The actual data won't change much and the format
> > of any CSV file should be *basically* the same. Even if the order of the
> > fields is different or there are more fields, the basic principles are
> > the same.
>
> I will over the course of the next few days convert your data to MySQL.
> Hope to have the FAQ done tonight.

Keep in mind that what is currently in the CSV files is only a small fraction 
of what will be there when we get all of the KUs defined.

> > So, where do you want to go from here? Do you want to be responsible for
> > the data extraction and presentation? Just the presentation? I have one
> > person (Luan) already working on the conversion from the existing CSV to
> > XML and two people who were interested in the coding from the other
> > direction. However, they have not responded since their original email. I
> > am going to ping them again to see if they are still interested. We also
> > have a couple of new editors and content writers.
>
> XML is a subset of eXtropia views. I will have to show her how to put
> her code into the views used to display the content of the MySQL tables.
> Once the cvs is converted there is no longer any need for them.
> Except if you what others to install the system on their server which I
> don't see a need for.

There is a  misunderstanding here and I am not sure on which side. Luan is 
currently working on the code to generate the XML file from the existing CSV 
files. The other two are planned to write the code to covert the XML to 
either new, expanded CSV or insert them into a database. Once the XML files 
are generated their will a need for CSV files, in general.  We need to 
provide the option of running the system without an SQL database or some 
other database other than MySQL. Although we can initially work toward this 
first public version being on MySQL, I don't want code written exclusively to 
import to MySQL.  
 
> > If you are doing the presentation component, could you also come up with
> > a decent CSS concept? What I did was a kludge and doesn't look good on
> > all systems, particularly MS Internet Exploder 5.x. With 6x it's okay,
> > but 5 gives me problems. We could tell people to upgrade, but I think it
> > is just as much a problem with the styles as it is with IE5.
>
> You will not have too much luck accommodating IE except to give them a
> complete view done just for them. This is one reason that I have avoided
> fancy dancy. My suggestion here is to code to HTML 4 get Bobby
> certification and encourage users to upgrade to a gecko based browser
> such as Netscape or the leading edge Mozilla by providing links to them.
> Possibly a page giving reason why they should abandon IE for a net
> standard browsers.

If we can get it looking decent on Mozilla-based, Konquerer, Opera and any 
thing else that runs on Linux, plus fairly decent on IE 5/6 then I am happy. 
We have people using IE visiting the site. As long as no functionality is 
missing and something looks goofy, then I am happy. We cannot afford to 
alienate people by saying the site only works on non-MS browsers. That's 
shooting ourselves in the foot.

> > I just what to re-emphasize that I don't want to turn this into an
> > eXtropia project. That is, a project to get linkbat work explicitely with
> > eXtropia. It should be completely independent of the front end we choose.
> > I understand your enthusiasm with the product, but as much as possible
> > the presentation layer of linkbat needs to be portable to any system.
>
> eXtropia is portable to any system that has Perl. The data can be in any
> file or SQL. So if your code can access the files or SQL then you can
> display the contents. eXtropia takes care of authentication, Sessions,
> encryption, DataAccess (file, MySQL, PosGress, Oracle etc), translation,
> Logging, to name a few. This is part of the base eXtropia distribution.

Not eXtropia has to be portable, but linkbat. eXtropia handling the data 
access concerns me in that we will be dependant on eXtropia. 
If eXtropia access an "api" (one we define or is provided by someone else) 
that goes out and gets the data from MySQL, CSV, or whatever, then there is 
less of an issue. However, eventually (not necessarily in the first 
iteration), I want the Data Access to be presentation independant or at least 
have the ability to make it so. Obviously if it is MySQL and someone know how 
the tables are defined they can access the data. However, I do not see that 
as the same thing as us providing the "api". 

> > It will eventually need to be made available for download from
> > SourceForge, so it needs to be in a form that people can install without
> > hounding us to death. On the other hand, is there any **requirement**
> > that says we need to make things  available throguh source forge.  Hmmmm.
> > Still, eventually I would like at least the code and data files format
> > available for download. I want to adher to the open source principles
> > while at the same time not giving aware the farm (just yet).
>
> eXtropia is already available for down load and is vary easy to install.

I was thinking more along the lines of our stuff getting available for 
download. What "our stuff" is and when it would be ready for download is 
still unclear. 

> They have their own support system so one can just pass eXtropia level
> questions to them or my HelpDesk system. The part that will be:
>
> HTMLViews spacific to us

How configurable are these? It is possible for users to change the color 
scheme, for example. 

What about the layout from and administration pespective? I want certain data 
presented in certain places. Take the Linux Tutorial as an example. It's a 
3x3 table with each cell performing a particular function (i.e. displaying 
certain data). 

> The Data Files. I can export the current table contents into a cvs file
> or SQL file for import into SQL.
>
> Any ActionHandlers that are unique to our project so far I don't see a
> need for them.

Please clarify for us (i.e. me ) what an ActionHandlers really does. If I 
click on a "More Info" link to see the more info related to the current page, 
wouldn't it be an ActionHandler that generates the new page? However, since 
we are dealing with "standard" queries a "standard" action handler could runn 
the query, something like this:

select * from moreinfo_table where page_id = '1234'

Is that what we are talking about?

Let's say I want a list of topics and clicking on one gives me a list of all 
KUs that reference this topic. (See Topic Tree on the linkbat concept page). 
How would that be realized? Would that also be a simply query?

> The setup files unique to the project.

What do they look like? How are they configured?

> I only code to open source though I do accept payment for modification
> of code but the code anyone can have and use.

Open source is one reason for doing it through SourceForge. Code is not the 
issue, but rather the data (i.e. the KUs). One reason I am concerned with 
releasing the data is that I see us opening up a whole can of worms to 
support raw data. If we have a complete "product" which includes instructions 
on how to configure eXtopia, perl code, XML files, etc I don't have a 
problem. I just don't want to deal with pieces.

> > The more I think about it, the more I like th
> > e idea of extra content at the
> > various levels of "membership". In addition to the online output of
> > commands,
>
> Easy to do in eXtropia.

Out of curiosity, can you define different levels? 

> > I have been planning to include "pseudo-animated" graphics on the site.
> > Using an IFRAME that refreshes itself at regular intervals, you can get
> > some nice looking graphics. Since there is nothing on the site that needs
> > 3D life-like animations, this works well. The images are pretty small
> > (typically < 50K) so I don't see bandwidth problem. As an alternative an
> > co-worker of mine showed me a trick to split up the graphics into small
> > chunks. Large white spaces are s single byte that is blown up to fill the
> > whole area. Probably not necessary, but something to consider. At any
> > rate, these animations would probably be something for those people who
> > have registered, whereas online command output would be for "paying"
> > members as they would need to download it to their local systems. Members
> > would still have the current updates and we could potentially have things
> > like the quizes only available for registered members since I plan to
> > keep track of the user's score, completed sections and so on. Hmmm. There
> > are many possibilities.
>
> This is part of the views.

And views can be user based, right? So a registed user could get the 
animation, where an anonymous person does not? Is that right?

> > Have you had a chance to go through the concept and data description? I
> > would really like some input as quickly as possible to finalize the
> > format. Luan will need to know what the final format is before she can
> > convert the CSV files. She is already working on the code (I think), but
> > is using the "suggestions" I provided.  Her code shouldn't change too
> > much, however I would like to get moving.
>
> No I have been a bit busy and took a few hours off for myself this
> weekend. I will get on them tonight. Get Luan in touch with me and move
> this conversation to the list. If you don't have one I can create a list
> for us on my system. I have access to my lists via Perl as well. Users
> can use a web form to sign up.

Please, please, please read the concept and data model. Also, if I don't hear 
back from the others who wanted to do the XML code, then I will work together 
with Luan on it. Assuming she wants to keep going. ;-)

> I will be back in touch. Check the site tomorrow night I should have the
> FAQ converted to MySQL by them.

Will do.

Regards,

Jim
-- 
---------------------------------------
"Be more concerned with your character than with your reputation. Your
character is what you really are while your reputation is merely what others
think you are." -- John Wooden
---------------------------------------
Be sure to visit the Linux Tutorial:  http://www.linux-tutorial.info