ning vs erlhive

[Joel Reymont]

I had an epiphany this morning when I remembered ning.com.

I checked out the site, again, and think a social network could be a
great base for The Erlang Journal... except I would need a good
designer to come up with something like this:

http://www.openwineconsortium.org/

Then again, I don't know PHP, like the looks of it or know how much
social interaction you guys are looking for. I think the journal is
about learning, ideas and code sharing.

Rather than host OpenSocial widgets, I think I'd rather give you the
ability to experiment with Erlang code, on my backend infrastructure,
without needing to set up Erlang hosting of your own. Ulf Wiger's
ErlHive is just the tool to make this happen.

What do you think?

Thanks, Joel

P.S. Scheduled programming will continue once I get some sleep and,
hopefully, sample Erlang Web code from Francesco Cesarini.

--
http://twitter.com/wagerlabs

[grant michaels]

I think that without a doubt, the ErlHive concept is the most interesting ... that being said, Erlang-Web seems to be documented best ...

best personal regards,
 
-[ grantmichaels ]-

[Eli Liang]

I suggest that whatever is technology is selected, the final goal should be something easy enough to use and come with clear documentation on how to interact with the journal and sample code. Anything less will put people off on participating with the journal.

I would further suggest that any project can be one which can develop a life of its own. If the journal system is itself the first journal project, there should be some limits put around it so that it's not still in progress in February and that we've moved on to other learning opportunity, even if the journal system itself turns out entertaining enough that a few of us hang back to twiddle more with it.

In fact, I would suggest that one of the first things that should be decided is the overall structure. Is this a strict monthly journal structure where we'll have a primary project every month, and even if it is not fully complete, we move on to the next project at the start of the next month?  I recommend this model, since it will keep things fresh. Furthermore, if a person isn't interested in the project, they only need to wait a month for the next one. In some ways, this would be like an online version of Dr. Dobbs Journal then.

Lastly, I suggest something for those that hang back to twiddle with older projects. I suggest that every new project, including the journal project, come with it's own newly spawned forum/subforum *when* the journal moves on to the next month's project. That way, people who are terribly interested in continuing the work of the prior month's project and further enhancing it have a place to discuss their work on it without derailing the new subject of the month. Alternatively, if there is a concern that having a separate forum/subforum for every project might fragment the journal community too much, perhaps just have one other forum called something like "Older Projects" which could serve as a shared forum discussion of all projects from prior months.

[Joel Reymont]

On Dec 3, 2008, at 1:07 PM, grant michaels wrote:

> I think that without a doubt, the ErlHive concept is the most
> interesting ... that being said, Erlang-Web seems to be documented
> best ...

I meant ErlHive in the sense of a framework that allows execution of
arbitrary, even malicious, user code. I'm still going to use Erlang
Web but ErlHive should let me put up a web Erlang prompt, have users
type Erlang code into it and execute it.

--
http://twitter.com/wagerlabs

[Joel Reymont]

On Dec 3, 2008, at 1:28 PM, Eli Liang wrote:

> I suggest that whatever is technology is selected, the final goal
> should be
> something easy enough to use and come with clear documentation on
> how to
> interact with the journal and sample code. Anything less will put
> people off
> on participating with the journal.

I agree. I haven't figured out a format for the journal yet. For now I
just post the code to http://github.com/wagerlabs/thinkerlang/tree/master
and explain it in separate email messages.

It would be awesome to have a comments system like that of http://www.djangobook.com
but that's far out.

> If the journal system is itself the first journal project, there
> should be some limits put around it so that it's not still in
> progress in
> February and that we've moved on to other learning opportunity, even
> if the
> journal system itself turns out entertaining enough that a few of us
> hang
> back to twiddle more with it.

Agreed. My goal is to be done with the first version this week or next.

> In fact, I would suggest that one of the first things that should be
> decided
> is the overall structure. Is this a strict monthly journal structure
> where
> we'll have a primary project every month, and even if it is not fully
> complete, we move on to the next project at the start of the next
> month?

Correct. Something new every month. I thought of something new every
week but I'm still working on the journal part time and so this may
not be possible. I'll try to post sizable chunks of interesting Erlang
code weekly so that we can take it apart.

> Lastly, I suggest something for those that hang back to twiddle with
> older
> projects. I suggest that every new project, including the journal
> project,
> come with it's own newly spawned forum/subforum *when* the journal
> moves on
> to the next month's project.

Makes sense. The good thing about Ning is that it comes with groups,
discussions, forums, comments, etc. etc. etc. No need to re-implement
all that and you can post videos and photos if you want to. The bad
thing, IMO, is that it's PHP.

I'm not a PHP hater by any means, it's just something I haven't used
yet and I have heard plenty of bad things about it.

Thanks, Joel

--
http://twitter.com/wagerlabs

[Matthew Kanwisher]

Ning has so much stuff out of the box you rarely if ever have to write php code, you can just skin things or bring in open social widgets. Hell the who thing has apis everywhere so the parts you want to add you could go write in Erlangs ;)

[Joel Reymont]

On Dec 3, 2008, at 3:03 PM, Matthew Kanwisher wrote:

> Ning has so much stuff out of the box you rarely if ever have to
> write php
> code, you can just skin things or bring in open social widgets. Hell
> the who
> thing has apis everywhere so the parts you want to add you could go
> write in
> Erlangs ;)

Here's a list of issues in no particular order...

The middle column is too narrow, how am I supposed to post code?

How do I add code highlighting to displayed code?

How do I add Markdown syntax to the default editor widget and how do I
tell it the kind of code that I'm posting so that the highlighter
would work?

--
http://twitter.com/wagerlabs

[Matthew Kanwisher]

To make it larger you can just add a new page and then you don't have the bars on one side. As far as adding code highlighting, there is nothing built in, but if you already have something that creates html you could easily have it post to ning using their apis or probably just a good old fashion form post to where their blogging editor does ;) or there is always just copy and paste. Not a great answer to that.

~Matt