After the Bar Closes

| 38 Comments

Some of you may not know this, but the Panda’s Thumb does have a forum on AntiEvolution.org: After the Bar Closes.

For those of you missing the Bathroom Wall, you can post your pearls on the forum.

38 Comments

It just hasn’t been the same since evopeach was banned.

bah. just invite Blast into some of your discussions.

Maybe we could give Blast some more tips on ID friendly scientists,e.g. http://www.prop1.org/thomas/peacefu[…]rgy/how5.htm He must be the best thinker the DI has, they need all the help they can get. Penrose is a funny bird isn’t he ? Thought had a thought, “glass bead game” sort of thing. Quantum gravity intelligence ?

Why not just change the title of this thread to “The Bathroom Wall”?

Now the interest over the Dover Trial has subsided, the need to present a decorous public face has lessened, so why not bring it back? It would reduce the level of extraneous comments in topical threads that, for instance, Jack Krebs was complaining about recently, and there would be somewhere one can ask stupid and naive questions. Try doing that in “After the bar closes” and you may as well talk to yourself.

It just hasn’t been the same since evopeach was banned.

I’ve noticed the improvement, too.

I’m doing some hacking on the IkonBoard software at the AE BB. There are a number of features that the programmatic posting function I’ve made will make possible. We’re looking into setting up an AE BB topic for every post here, so off-topic comments can be easily moved over to the AE BB thread. I’m also considering automated addition of news alerts from Google. Other suggestions would be appreciated.

that sounds great, Wes. How do you find time to do all this?

another suggestion:

is there a way to change the tag error checking so it redirects to the same kind of error window we get if we try to enter two posts too close in time?

that way, if there are coding errors, we can fix them without having to retype everything.

thanks again

Reed does the bulk of the MT hacking. But I think that comment-handling hacking is getting near the top of his to-do list, so I will certainly mention the “preview” window on error hack to him. It is something that has been discussed for a while.

that way, if there are coding errors, we can fix them without having to retype everything.

If you have a decent browser, you should be able to hit the back button rather than retype.

It is something that has been discussed for a while.

I’m surprised tha it wasn’t done that way in the first place. As it is, it’s like a compiler that takes a program with syntax errors and turns it into an executable that prints the error messages, rather than printing the error messages. Or a spelling checker that replaces your entire document with just the misspelled words. Even making the Post button do nothing if there are errors would be better than the way it works now.

define “decent browser”, and how does that decent browser cache submitted form data?

There is such a thing as Preview, we have that button down there for a reason.

I’m surprised tha it wasn’t done that way in the first place.

It’s harder than you might think.

1. You have to modify the blog software such that it understands validating comments for their format.

2. You have to construct a parser that not only knows where the error was, but can guess what caused the error in the syntax.

3. You have to figure out how to construct the error message such that users understand where they went wrong.

4. You have to find time when you are not teaching, researching, writing, or trying to finish your dissertation to implement, test, and support the feature.

etc.

There is such a thing as Preview, we have that button down there for a reason.

And I use it, but that isn’t a reason to post syntax errors to the board.

1. You have to modify the blog software such that it understands validating comments for their format. 2. You have to construct a parser that not only knows where the error was, but can guess what caused the error in the syntax. 3. You have to figure out how to construct the error message such that users understand where they went wrong.

No, you don’t. You have to do that to produce more meaningful error messages than what we get now, but that’s not the request. The request is to treat Post like Preview when there’s an error, preferably with including the error message (the same one that now gets posted) somewhere on the page.

4. You have to find time when you are not teaching, researching, writing, or trying to finish your dissertation to implement, test, and support the feature.

I understand why it takes time to make, test, and install changes – which is a good reason to put in extra effort to get it right in the first place. What I said was “I’m surprised tha it wasn’t done that way in the first place” – that way being to treat Post like Preview upon error, not to highlight the error in the text or anything like that.

P.S.

You might consider posting the software somewhere, making it open source. You would still have to coordinate changes, but you wouldn’t have to code (all of) them.

that would depend of course on whether the kernel for the software is already open source or not.

I doubt Reed wrote it from scratch, after all.

I’m sure better open source software for this kind of forum already exists, but totally reconfiguring PT to use it, while maintaining the presence and order of all the previous archives would be quite difficult.

I’m sure Reed would welcome any help in modifying the existing software. I don’t think my limted html or asp scripting or database skills would be useful tho. If you have the skills, you could send him an email and offer to assist?

in the meantime, what browser were you suggesting again? Mine does not maintain submitted form data in a cache anywhere.

The request is to treat Post like Preview when there’s an error, preferably with including the error message (the same one that now gets posted) somewhere on the page.

That’s problem #1. MovableType is not constructed with the idea that a comment or post may be malformed. It doesn’t validate comment formats or provide any hooks to do so. I’ve requested this feature but they haven’t done anything.

You think it is easy. Fine, download the MovableType software and construct a patch for me to apply that provides such a plugin interface.

I had a limited amount of time before the semester started and teaching ate up my spare time, so I had to prioritize my hacks.

At one point I suggested removing the post button, but was overruled by the other crew members.

MovableType is not constructed with the idea that a comment or post may be malformed. It doesn’t validate comment formats or provide any hooks to do so

thanks, Reed, that answers my question.

define “decent browser”, and how does that decent browser cache submitted form data?

Anything but IE? Firefox and Konqueror on my linux box do this – most likely anything built on the Mozilla code base does, as well as other high quality browsers like Opera. As for how … do you want a listing? The back button accesses a stack of data structures; those data structures include the form data, associated with the form it belongs to.

MovableType is not constructed with the idea that a comment or post may be malformed. It doesn’t validate comment formats or provide any hooks to do so. I’ve requested this feature but they haven’t done anything.

Are preview, post, and check spelling actions completely canned functions provided by MT, or do you have some control over how they are performed? If the latter, I can see why this is such a problem. I hadn’t realized MT was so inflexible.

that would depend of course on whether the kernel for the software is already open source or not.

I was only referring to the software specific to PT, but I pretty much assumed that the underlying code was OS, because volunteer projects like PT are almost always based on such.

Are preview, post, and check spelling actions completely canned functions provided by MT, or do you have some control over how they are performed? If the latterformer, I can see why this is such a problem.

Anything but IE?

I use Netscape, and can’t do it by hitting the “back” button.

Something I don’t understand: why didn’t we see any of these syntax error messages before the switch to KwickXML? Wasn’t that just a change to the parser, not to the blog software?

In order to support nested quotes, lists, and other such things, KwickXML has to use an XML parser which generates error messages. I capture the error message and return them to the user. This allows people to get and idea of what went wrong when they preview. (The other option is do nothing and let the parser die, causing loss of data.)

I consider this a good design decision since it prevents commenters from messing up the the XHTML of these pages.

Anything but IE?

I use Netscape, and can’t do it by hitting the “back” button.

why do i feel like we are one step away from a windows vs. linux war?

let’s head this off at the pass, shall we?

I consider this a good design decision since it prevents commenters from messing up the the XHTML of these pages.

Ok, here’s a crazy idea: wrap the user’s broken text in code tags and append that to the error message.

why do i feel like we are one step away from a windows vs. linux war?

Because you mistake talk about browsers for talk about operating systems, or assume that a critical comment about one piece of a vendor’s software is criticism of all their software? Netscape, Firefox, Mozilla, and Opera all run on both Linux and Windows. I only mentioned Linux because that’s what I’m running at the moment and I didn’t want to assume that Firefox acts identically on Windows, and I didn’t want to boot it up to find out.

let’s head this off at the pass, shall we?

We hadn’t even left town.

but would that work if there aren’t any hooks to use to intercept a “broken” message to begin with?

gees, man, get a clue. i was hoping the head it off comment would be a clear statement that this isn’t worthy of discussion here, and you go ahead and bait me with overgeneralized commentary about what my assumptions are.

is it possible for you to tone down your approach a bit?

The MT software places the raw comment into the database without verifying its format. On a subsequent rebuild, the MT software retreives the comment from the database and invokes the KwickXML formatter which attempts to render KwickXML in XHTML.

If there is an error in the comment format, the formatter returns an error message instead of the rendered text. That is why you see error messages in comment. Those users didn’t preview and fix their errors.

One way to fix the problem of syntax errors being posted to comments would be to require that a post first be previewed before posting. Will Moveabletype’s software allow the post button to be initially suppressed?

Yes, the post button can be removed, and that was my initial plan when we went to KwickXML, but the rest of the crew tabled the idea.

i think Reed already addressed that just a few posts above. From what he said, it sounds like others objected to removing the post button.

i can see why that would be, as in busy discussions waiting for preview and post, rather than just post, can get a bit tiresome.

I have to admit, that’s the main reason i rarely use preview, and usually just copy my text before i post it.

I suppose more patience would be warranted.

but would that work if there aren’t any hooks to use to intercept a “broken” message to begin with?

The broken message isn’t intercepted by MT, but it is intercepted by Reed’s formatter; as he just said, he captures the error message from the XML module. At that point, he can attach the text (which he just passed as an argument to the XML module) to the error message, surrounded in code tags. (That could still result in broken XHTML if the text contains a /code tag, but another XML parse could check for that).

gees, man, get a clue.

I’m not the one who is clueless. You “tried to head off” something that wasn’t happening and wasn’t going to happen, as you would have realized if you had paid more attention. I suggested that any browser but IE was “decent”, and Lenny noted that it doesn’t work in Netscape either. To the degree that Netscape is associated with Linux rather than Windows, that would tend to head off any battle (of which there were no signs in the first place), not lead to one.

you’re still baiting. I’m not biting. can we end this now?

you’re still baiting. I’m not biting. can we end this now?

But you just DID bite. If you don’t want to talk about something, then DON’T TALK ABOUT IT. Sheesh.

hmm. actually i guess i do want to talk about something. i posted it in a different thread:

http://www.pandasthumb.org/archives[…]n_found.html

perhaps it would even be better to start a completely ot topic on the subject of our posting behaviors over on the ABC?

I for one have no interest in the subject, and would certainly be happy for you to stop making any comments at PT about mine.

On the Royal Adelaide Hospital Intranet which I’ve designed and wrote I force the user to do a preview before posting up a message on the home page. Step one lets the user enter in their post Step 2 displays their post If they make any changes then “Post” changes to “Preview” and Step 2 is preformed again. Step 3 happens when you “Post” and they are given a msg indicating success or failure of the message being added. The only real error that happens at that point is a loss of connectivity between the web server and database server.

Forcing a preview I found works great for us.

Forcing a preview I found works great for us.

see the discussion about removing the post button above. Reed tried it and was denied.

see the discussion about removing the post button above. Reed tried it and was denied.

I read that already. I was just sharing my experiences with this style of implementation.

gotchya.

About this Entry

This page contains a single entry by Reed A. Cartwright published on November 13, 2005 7:58 PM.

Bongo for Bird Biogeography was the previous entry in this blog.

Lecture Planned on Intelligent Design is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.

Categories

Archives

Author Archives

Powered by Movable Type 4.361

Site Meter