Comments on: What is the RebDev project?
As I mentioned in Better Developer Communications, improving developer communications for R3 is critical. By "critical" I mean that we will not do a bigger (in terms of users) release of R3 until we have a way to easily and accurately communicate about it.
Nutshell summary:
RebDev is the messaging (BBS/chat) backend for DevBase.
When DevBase was released, it was pointed out that it would be problematic without a messaging method attached to the storage mechanism. RebDev is that method.
The long story:
As you may or may not know, I've been writing BBS/chat/IM systems since the early 1980's, so I have a few strong opinions about how they should work. IMO, there are very few good systems out there. (The one I use most often, AltME, even lacks many of the features I require.)
Last month, I wrote RebTalk, an experiment using R3 for the interface. It proved that an R3-based GUI was easy to write and fairly stable. (We've improved text-edit since then too.) The back-end was just LNS (REBOL/Services).
Mid-December, I met up with my friend Ron Everett, and spent several hours talking about associative databases. If that name sounds familiar, recall that Ron made a presentation at the REBOL Devcon in Paris.
Ron has created what to me has been a dream project since my neurophysiology days (late 1970's): a true homogeneous associative database system. It is not based on tables, but on links between concepts, the neurons of the database. It's an amazing technology. Basically, it is a storage mechanism like the brain, and I understand the implications of that statement, and I don't say it lightly.
Anyway, the discussion got me inspired to rewrite the RebTalk backend, using some IOS associative/tagging techniques. The result is a very lean, extremely flexible messaging server that should do quite well as the "BBS" needed to properly support DevBase. (In essence, RebDev is really the start of Altissimo.)
So, RebDev has been up and running on the server for about a week now. I continue to test it... getting it ready for your usage. And, yes, it will be accessible via:
| R3 shell | This may not seem useful to you, but it is very useful to me, as I'm often logged via SSH to some remote location.
|
| R3 GUI | Small, simple, easy. And, helps us test R3.
|
| R2 | We (community) can write a back-port to it for shell, and GUI if you want that too (maybe via RebGUI to make it easier).
|
| HTML in mobile CSS | Mainly for access from my (and your) iPhone or other mobile device.
|
| HTML in pretty CSS | For web users who prefer interfaces that look like PHPBBS.
|
| RSS | You users know who you are.
|
| Email | Because a lot of folks still use it.
|
The R3 shell and HTML mobile will be done first. Others will follow.
Conclusion:
So, there you have it. And, the architecture is beautiful and simple. It is exactly what I've wanted and should satisfy the needs I posted in my earlier blog... quite well in fact.
I will be saying a lot more about it very soon.
17 Comments Comments:
Steeve 27-Dec-2008 22:27:16 |
Very interesting way.
Could be that the future RDBMS ?
- an associative data structure handler
it's remember me the IDE Plex(obsydian) in the nineties.
it used widely the concept of triples (tuples) to modelize applications and databases.
| Edoc 28-Dec-2008 13:30:26 |
The associative database sounds a bit like the semantic web (e.g. RDF or the Web Ontology Language, OWL http://www.w3.org/TR/owl-features/).
I've thought that it might be powerful to be able to associate word variants in REBOL according to sets of rules instead of as synonym equivalents. An associative db sounds like an interesting idea. | RobertS 28-Dec-2008 13:35:55 |
Is that the Ron Everett of www.spoke.com ? | RobertS 28-Dec-2008 13:42:19 |
http://www.relevanttechnologies.com/ ???
Some Rebol DevCon note mentions Relavance (sic) | RobertS 28-Dec-2008 13:44:54 |
Nope. That must be a neologism of sorts: "relavance"
Ron Everett seems to be behind
http://www.associativesolutions.com/relavance.php
| RobertS 28-Dec-2008 13:49:01 |
Relavance is apparently a trademark of ASC
Strikes me as a misguided Gallicism ... comparable to "Oracel" instead of "Oracle". Whatever. Sme people still think "Rebol" is a typo for "Rebel" ... | RobertS 28-Dec-2008 13:54:19 |
One last note: the acronym ADBMS is used for "Active DBMS" already ... ADMS is also used for Advanced Database Mgmt Sys ... life without a TLA ... | RobertS 28-Dec-2008 14:00:49 |
I can't resist: if you only consider the importance of the concept of "relevance" in advanced data systems then the choice of the trademark appears as silly as "Oz", "Mozart" or "Alice" to name only 3 of my favorites from the Saarland AI group.
I can only hope we will not be adopting this latest as trade jargon. As someone who works in Curl, I know a silly trademark when I have to use it in a web search. | RobertS 28-Dec-2008 14:19:55 |
In case you think this is trivial folly, consider the following.
OZ permits constraint resolution which is important in generating musical harmony. Had the OZ team called their new language "Dorothy" instead of "Alice", anyone seaching on "Oz Dorothy" could not add " -musical" as a filter. Only worse might be an IDE for Mozart called "Amadeus"
Consider any future search for "relevance relavance" and the typo's for "relevance" that such a search will regurgitate.
In my experience, these sobriquet mishaps occur when advice is no sought before filing a trademark (some jurisdictions have rules to limit folly in registered names)
In code from off-shore I cope with "Persistance" already.
"persistant-id" is a variable name that tends to persist.
A dictionary of your native tongue belongs on every programmer's desktop and iPhone.
Typo's are irrellavant in e-mail and chat threads which are not placed in persistent storage. "Intaligance" likely would not be an intelligent or elegant trademark in the English-speaking business world. | Carl Sassenrath 28-Dec-2008 17:32:17 |
Q: "Could it be the future RDBMS?"
A: Difficult to know. People understand tables really well, but tables are a flattened view of a multidimensional "real-world" domain - so they become problematic over time, as the original requirements diverge and as new and unanticipated queries become needed. Assoc DB model has no tables, but links... many, many links. Much more flexible.
Personally, what I like about the assoc model is its singular "cell purity" (homogeneous). That is, a single concept handles all variations, data or metadata, or whatever.
The difficulty in the assoc model has always been performance. Modern CPU's/cores with caches help solve that problem for practical cases, but they are unlike the brain, which is massively parallel in its associative mechanism. Although I'm out-of-date on this topic, I think we still don't have that in a physical implementation (of any reasonable magnitude), but I think Ron has brought it as close as we can get in a virtual way.
BTW, I should note that REBOL, being based on "relative expressions", is highly associative because the symbols of those expressions are very late bound (the "linkage" to symbolic meaning is held off as long as possible).
So, one day, if it ever comes to be, when our computing systems are truly associative, you can look forward to a nice boost in REBOL's speed. (But of course, it's pretty good already for most solutions.) | Carl Sassenrath 28-Dec-2008 17:56:10 |
Q: "The associative database sounds a bit like the semantic web..."
A: I think they are different beasts. Well, Ron could tell you for sure, but an associative database is about pulling answers from unstructured knowledge (kind of like what we do as humans), whereas the semantic web is about structuring knowledge as part of its expression... essentially a glorified encoding method -- by building an ontology-- defining terms of meaning that provide anchors for interpretation.
So, to me, one is quite the opposite of the other, but a good assoc DB could be used as a means to interpret the semantic web (or parts of it) in a high performance way.
(BTW, since I have the opportunity here, I must say that I do not agree with the general approach of the semantic web. It's just a bit too 1986... if you remember those AI days.)
| Carl Sassenrath 28-Dec-2008 18:16:58 |
RobertS: It's www.relavance.com
Ron keeps a very low web profile for his technology. I'm not sure why. Personally, I think a great number of people would benefit from learning more about his approach and concepts.
I've nagged Ron a bit about also providing university courses on it. I think professors and students would really get into his DB. (But then, those who live in glass houses should not throw rocks, eh? That is: where's my REBOL curriculum? Guido of Python fame did that himself, and I highly respect his achievement. It's probably in part why even the esteemed MIT dropped Scheme in favor of Python for undergrad coursework - something that floored me to learn last year when I visited.)
Regarding naming, TLA's, etc... it's so hard these days. The web sucks up (and often wastes) so many useful words and combos, so that leaves many of us looking for variations in existing words. For example, "REBOL", was not my first choice; however, back when I "altavisted" it, there were only six hits at that time. Works quite well for web searches.
| Edoc 7-Jan-2009 15:33:04 |
With regard to community input, I think Stack Overflow is a pretty interesting forum. See http://stackoverflow.com/ | Jerry Tsai 14-Dec-2011 14:34 |
I am developing an Associative Database based on LazySoft's Book "Associative Model of Data" using R3. Since R3 doesn't support Rebin yet, I have to use text file as the database file format. | ecollage 17-Apr-2012 16:16:47 |
Jerry:
Any progress on your rebol AMD system? Very interested... | Supra shoes for girls 8-Jul-2012 20:55:05 |
The Supra shoes for girls Asia this training camp lineup,More than 50 players from China, Hong Kong, China, Chinese Taipei, Japan, Korea, Australia and other countries and regions,Players in the Supra shoes for girls list, we are familiar with the noble, Supra shoes for girls and consultant in the lineup, not only have Yi Jianlian, Gay and DeRozan and the other two NBA players, as well as the former coach of the United Arab O'Connor, and Chinese star Du Feng and so on. | Gucci sneakers store 10-Jul-2012 22:36:58 |
All Gucci sneakers store is made from superior material and crafted carefully and intricately to ensure the best quality. Gucci shoes are some of the finest designer shoes on today's market. As leaders in high fashion, Gucci sneakers store's line always features the latest, most stylish shoes. As they are high in demand and well-made, Gucci sneakers are always available in expensive prices. You can, however, find more affordable Gucci shoes, if you know where to look. So searching on internet is really a good option to find your favorite pair of Gucci Sneakers. You could also find some great discounts if you are lucky and finding the most wanted brand while sitting at home is really a good experience for anybody. |
Post a Comment:
You can post a comment here. Keep it on-topic.
|