Author: Andrei Fortuna
Date: 15:03:19 12/06/00
Go up one level in this thread
On December 06, 2000 at 16:59:48, Roy Eassa wrote: >Quick feedback: > >1. It asked for a debugging DLL: msvcrtd.dll. I only made it work by making a >copy of msvcrt.dll and renaming it to what was asked for. sorry about that >2. Need a bit of instructions on use. What is meant by the folder called >"Unsorted"? I don't know how to sort it. And what's the folder called "Thrash" >for? all new messages are placed in "unsorted" "thrash" is where all messages go when you delete them kind like a safety net and of course when you delete something from the folder "thrash" it's gone for good. all messages are stored in just one big file named "bigfile.dat" deleted messages from thrash aren't deleted at once from this file. when you have deleted lots of messages and want to reduce the size of the file you can do File->Compact Database which will generate a new bigfile.dat without the deleted messages. >3. User must double-click where the Windows standard (and convenience) calls for >single-clicking. E.g., to expand an outline on the right side, or to show the >message selected on the left side. Single-click please! right now it's double click for left side and single click for right side I did it with double click on the left side because you might want to select lots of threads from the left pannel to delete and I didn't want all for each thread you select to have it in the right pannel. >4. Choosing "Rebuild Search Tree" (what does that do??) on just a single day's >worth of messages made it lock up my PC for about 90 seconds. What hardware configuration do you have ? Did you have many other messages in the database or you just unpacked my program to a directory and added that day of messages ? If the case is the latter, it should be very fast, unless you have a much slower computer than myself (amd k6/2 300Mhz 64 mb ram). And to answer your question : the search tree is a list of all the words used in all the messages you have in the database, for each word I keep pointers to the articles that contain that word. So a search for the articles that contain one or more words is very fast. The so called search tree (it used to be a tree in the beginning, but I changed to a list afterwards and didn't change the name) makes searches very fast but with a price : each time you do an update to the database it must be updated. Else applying filters or searching text won't work on the newly added messages. The search tree is the most memory consuming part of the browser, so I strongly recommend to rebuild the search tree after each new 10000-20000 messages added to the database. On my computer it was using about 46 mb of memory (from which about 36 used only by the search tree) when I told him to rebuils the search tree for a new 60000 messages archive. >Otherwise, looks nice overall! Thanks ! Next version will contain a html viewer for messages, which will allow me to display chess diagrams like in ccc. Did I mention colors inside messages and various fonts ? >On December 05, 2000 at 18:28:54, Andrei Fortuna wrote: > >>Greetings ! >> >>I'm happy to announce the release of an Offline CCC Browser, written by me in >>the past weeks. Among its goals are : the ability to store all CCC messages from >> the beginning of CCC, very fast update of the archive, super fast search among >>messages, ease of use by storing messages inside internal folders, automatic >>message filtering. >> >>It's all in there except that the filters interface is a bit limited/awkward and >>the search command allows you to seek just one word. Consider it a beta, it's >>been tested but not long enough. >> >>URL : http://www.geocities.com/andreifortuna in the downloads section.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.