This group is announcement-only. You cannot post messages because you are not a manager.
Description:
Vlerq is a research project about column-wise data storage, native persistence, and scripting.
|
|
|
Shelved
|
| |
The vlerq.org website and this mailing list are being shut down. Vlerq is a research project which continues to remain very near and dear to me. It has gone through countless iterations over the years, with occasionally some tangible results in the form of working code. I'm still exploring ideas, but I'm finding that the open source... more »
|
|
|
TclVlerq 1.8.0
|
| |
It's time to take Vlerq forward again. - moved LuaVlerq from vlerq/trunk to vlerq/branches/v7 in svn - added a new vlerq/trunk with a new pure-Tcl implementation Don't expect much from this initial public release, it's just to establish a baseline for some Tcl code I've been working on recently. -jcw... more »
|
|
|
making a case for views
|
| |
I've uploaded a draft of my upcoming EuroTcl2008 presentation, titled "Why arrays (and dicts) are great, yet totally inadequate." at: [link] It's an attempt to get people to think a bit more about structured data. All of this is really a re-hash of things, which most people on... more »
|
|
|
The MetaJet, Jive, jerq, MetaJet roller coaster
|
| |
A few years back I thought about writing a version (or at least
subset) of Metakit in Java. This is because all my applications are
now developed in Java using Jacl (Tcl in Java) and Swank (my Tk like
toolkit written in Java). I really wanted to be able to make use of
Metakit directly in the applications (as opposed to using it in a... more »
|
|
|
Rigs are re-usable Tcl modules
|
| |
A new definition: a "rig" is a re-usable Tcl script. Nothing earth-shattering, just an attempt to introduce a *few* conventions so as to encourage writing Tcl code in a modular and loosely-coupled manner. We all do this in one way or another I'm sure, but IMO this has not led to a sufficient level of communality so far... more »
|
|
|
new view model
|
| |
Update from the Vlerq front. Or is it the Metakit front? Read on... It looks like it is possible to construct a more flexible view model on *top* of Metakit. The key differences are: * support for optional values (0 or 1 occurrence) * support for repeating values (0 or more occurrences) * heterogenous subviews, i.e. views where each subview has a different... more »
|
|
|
Ratcl example (sort of)
|
| |
Here's a little task I recently had to go through: remove all duplicate files in a directory based on their MD5 digest. First the way I solved it, pretty trivially: 1) create a tcl script, called "thinmd5s.tcl": while {[gets stdin line] > 0} { set md [lindex $line 0] set fn [lindex $line 1]... more »
|
|
|
views in pure Tcl
|
| |
Hiya, I've been experimenting with YAVI (Yet Another Vlerq/View Implementation), and it's oodles of fun. Attached are three source files, run the demo with "tclsh8.5 views.tcl". This is the tip of a new iceberg btw, stay tuned ;) -jcw
|
|
|
Ratcl's notation
 
|
| |
Faced with once again the need to embed a new Vlerq implementation in Tcl, I've been pondering on notation again. In v4, Ratcl has the "pipeline" notation, which is quite neat at first sight: view $v op1 ... | op2 ... | op3 ... That can get a bit unwieldy for long pipelines, so there is also: view $v do {... more »
|
|
|
Vector-based XML parsing?
|
| |
I just found an XML API and parsing technology called VTD-XML. It appears to have a much nicer interface than either DOM or SAX-style APIs, and definitely is faster for most purposes than either (SAX is faster if all you're doing is looking for a particular token, but grep will do better if that's all you want) -- but more importantly, it... more »
|
|
|
