Slashdot Mirror
The Software Internet Database
Larry points out an interesting project called The Software Internet Database, the goal of which "is to compile the largest database of software titles and credits on the Internet. This includes all types of software such as operating systems, security, financial, mapping, browsers, video editing, games, word processing, and more. They have made a good start but still need more titles. Please take some time to submit missing titles so that they may compile the database faster." It would be informative to have a subway-map overhead view to know which of these are still available from their makers,have been folded into other products, or are now abandonware.
sourceforge.net, freshmeat.net, tinyapps.org, packetstormsecurity.org, etc, etc have been around for a number of years and do this task well. I checked out that site, and it looks like they are attempting to use a vanilla CMS for the task of managing the titles... You really need some custom software to do it right (check freshmeat.net for an example of a good custom interface).
To find out how many degrees of separation I am from software written by Kevin Bacon?
Makes picking a name for a new project much easier, if you have the names of what exists in what genre.
It only has somewhere between 100-200 titles... closer to 100. I can't even tell how many exactly, because their search tool is broken (can't go beyond the first page).
Search for an empty string (like I did) and it dumps the entire list.
Also their idea of web design leaves much to be desired. Who the hell stuck that black and gray logo on a white and off-white page? If they changed one to match the other it would look much better.
This seems like a good concept. The search feature on sourceforge and other software sites always seems a little dated. Searching for video editing software brings every file with the words "video" and "editing". Not really very helpful. For example, It took quite some time to find some OCR software on there two weeks ago. Sorting and arranging these programs by type should streamline the process.
I tried a few searches for projects I've worked on. No matches. Then I tried some searches for more popular software packages.. No matches for things like "Linux" or "GCC". What's this doing on slashdot? Nice idea but it's got a way to go.
Is there supposed to be some kind of incentive for us to contribute to this effort? Will the info always be freely available? How can we be sure of that? What are they planning to do with the info? What is the business? Just selling ads? How do we know this isn't another Gracenote stunt? Why is this even posted here with so little info?
It's not enough to bash in heads, you've got to bash in minds. - Captain Hammer
Um...pretty crappy "database". Anyone can submit a company? Then, when you submit a "title", you don't specify the company?
Also, when you submit these things, there is no indication of what is done with the information or when it might be posted.
Later, when you try to add a version, the drop-downs contain ALL companies and ALL versions. This is a recipe for future disaster - the author should probably have tried larger data sets when planning his interface.
...or it will be spamridden. We have tucows, which was useable, but only had windows software. We have freshmeat, which is useable, but mostly free software. We have sourceforge, which hosts free software. Point is, all those software catalogs is either narrow in scope, requires registration from the individual software project, is spam ridden, or disappears.
And how would you rate programs? Ensure that links works? Whom should you credit for the programs? What if the homepage moves? Sounds like a lot of if's without good answers.
And if they are submission-based, how many will bother? How many dupes will there be? Is MS Office 2003 seperate program from MS Office 2004 for MAC? Is Firefox and mozilla discrete apps? What about the different parts of the KDE suite?
Assembling etherkillers for fun an profit
For instance there are only 83 titles in the database. And while this deficiency can be addressed by people contributing data there are design choices which I find puzzling. For instance, How is Bill Gates' height relevant to the purpose of the project? And would it not be an idea to associate names with titles via a "role", rather than just a credit? Ie, I'd be interested to know of the 200 people who worked on, say, Starcraft, who was the producer, the artist, the game designer, the beta tester etc. As far as I can tell there is no way of identifying this at present.
a world in progress...
Oh, so they have the internet on software now?
*throws away tubes*
Those who believe the Internet is private,
find their privates are on the Internet.
Slashdot is news for nerds, not information for nerds. I guess we just learned the difference. ;)
Developers: We can use your help.
Does everyone remembers when IMDb was totally free and now only paid members can comments? Such community sites tend to turn up their nose on the community which built it.
FTF Post above ...
... submit some crap ... and then this crappy search engine will return results for things like Linux, Windows, Netscape, Gimp, Photoshop, Vmware, Inkscape
... they're great idea to create this super-huge database was spoiled by the fact they realized it would take forever to enter in all the data ... Solution: post an article on /. asking for them to submit data
They have made a good start but still need more titles. Please take some time to submit missing titles so that they may compile the database faster."
Get busy
From my perspective
For this to be successful they need to develop a user friendly interface and attract non-technical users. This could easily become the IMDB for software IF it's made for everyone, not just us geeks.
Haiku for you!
Some Wikipedia articles do have extensive lists of software, but Wikipedia is not supposed to be a link list. I've seen very useful lists of software deleted from Wikipedia articles precisely because they were just link lists, not references for the article. These deletions were correct under Wikipedia policy.
There is now DOAP (description of a project) - a vocabulary / schema that allows to mark up such information.
It would be much more fun to have machine-readable links between different titles that forked one from another, etc. Uses can be starting from "maps" of software evolution mentioned in above and to other uses yet to be imagined. (Note: I do not know if DOAP allows to describe such parent-child relationships between software projects, but if such a property is needed I am sure someone will invent it).
P.S. Having information about abandonware would also be useful - but mainly if they'd also provide downloads and source code (where available). Although I doubt anyone will go to such extent to preserve abandonware.