Hi everyone, I now consider VizReader to be stable enough, and feature
complete enough (for me) that I will stop working on it.

To those of you who tried it out the first time, feel free to try it out
again at http://vizreader.com (all your info has been erased ten times over
so you have to register again). Even if you don't have an OPML to import you
won't be a sitting duck anymore as you will have the option of subscribing
to existing feeds after registration is complete :-)

I've written a little bit about the process here:
http://www.prodevtips.com/2010/03/20/how-to-write-a-big-app-in-an-obscure-lisp-dialect-you-dont-know/

I might maybe come off a little bit unappreciative in some places in that
article, however that is not really the case. Take for instance what I see
as the biggest problem with PicoLisp; the lack of existing libraries for
doing common suff. I knew before I started hacking away on VizReader that
that was the case but I wanted to complete the project anyway and any grief
I got from the fact that there are few libraries is therefore self
inflicted.

In retrospect most of the non-trivial important parts of the application
were practically written by Alex, the asynchronous imports for instance, or
the remote indexing routines, clever idx stuff and much more.

Anyway, I wanted to see for myself is PicoLisp is a valid alternative when
it comes to web development in general and the answer is YES.

Thanks for the help Alex!

/Henrik



On Mon, Nov 23, 2009 at 3:06 PM, Henrik Sarvell <hsarv...@gmail.com> wrote:

> Hi everyone, so the picolisp application I've been working on and off
> on for the past year has started to come together. It's a feed reader.
>
> It can be found here: http://vizreader.com
>
> The reasons for making a feed reader are many, but there are two that stand
> out:
> 1.) I don't like Google Reader, the GUI for instance is annoying.
> 2.) Even if I liked GR it would still be a good idea since I've
> realized that almost 100% of the time I spend on the internet - when
> I'm not googling or browsing wikipedia - is in the reader. Being able
> to control that experience is important, getting Google to implement
> feature requests is a long shot...
>
> The name comes from the fact that there are quite a lot of
> visualizations that can be employed to make sense of the information,
> but most of all for discovery, especially the link maps. In this
> respect the Twitter link map has been really helpful. It wasn't
> supposed to be like that from the beginning but I got hooked on
> visualizing at some point down the road. The visualizations are
> displayed through shockwave files so the flash player is a must if you
> want to view them.
>
> At this point there are no community functions but everything is
> primed for this if the reader gets a few regular users.
>
> Feel free to try it out, some pointers first though:
> 1.) Don't test it just because it's a picolisp application, the feed
> reader should be a really important application for you, just like it
> is for me. At this point I don't want a lot of unused cruft i the DB.
> 2.) If you use Twitter then vizreader will be especially useful as the
> system doesn't differentiate between microblogging and normal
> blogging, only the display of the posts are different, a difference
> that I enjoy a lot.
> 3.) If you like it please don't tell anyone who don't fit the
> description in #1, at this initial stage it is a good thing with only
> serious users.
>
> Some instructions/information:
> - Separate tags with commas, yes tags can contain spaces (I hate when
> they can't).
> - If you update your twitter feed note that the interface you end up
> in will refresh the feed every 60 seconds, it will of course go
> towards your daily API limit.
> - The whole database is updated 6 AM central US time, using the
> application for the next 30-60 minutes will be painful as there will
> be frequent write blocks in place. I will look into the new pubsubhub
> protocol to see if this procedure can be eliminated in the future.
> - If a feed is selected for reading and it hasn't been updated in the
> last two hours it will be updated, this can take some time, yet again
> the subhub thing could solve this problem in the future.
> - There are no error messages when external sources time out.
> - So far only OPML files exported by GR and a desktop app called
> FeedReader has been tested to work. The xml.l library is used to
> import them so they have to be perfectly formatted.
> - The positive/negative graphs/features should be considered a joke at
> the moment since they simply use word lists to determine the amount of
> negative or positive words in an article. They only respect first
> level negations, "not good" will be interpreted as a negative
> statement but "not very good" will slip through and be interpreted as
> a positive statement.
>
> TODO:
> - Show error messages when external sources time out.
> - When a feed is added the feed list should update right away instead
> of having to reload the home page.
> - Update the number of unread posts in every feed in a certain
> category/tag when the category is "opened" in the list to the left.
> - Profile setting that controls behavior when a feed is clicked,
> update always, never or after an arbitrary interval, it's speed versus
> freshness which is important enough that it should be controlled by
> each user.
> - Get it to work on the 64bit version, it is currently running on the
> 32bit version.
>
> I hope some of you will test it in a serious fashion and like this
> first draft despite the rough edges!
>
> /Henrik
>

Reply via email to