Scid vs. PC 4.13

[Dhanish]

Thank you for the update and continued development of Scid.

Update Xfcc to handle secure connections (thanks to Andrew Hunt)

I am getting an error as below, any suggestion?

Code: Select all

couldn't open socket: invalid argument
couldn't open socket: invalid argument
    while executing
"::http::geturl $uri  -type "text/xml; charset="utf-8""  -query $xmlmessage"
    (procedure "::Xfcc::Receive" line 15)
    invoked from within
"::Xfcc::Receive $uri $username $password"
    (procedure "::Xfcc::ProcessAll" line 24)
    invoked from within
"::Xfcc::ProcessAll $::CorrespondenceChess::Inbox"
    (procedure "::CorrespondenceChess::FetchGames" line 12)
    invoked from within
"::CorrespondenceChess::FetchGames"
    invoked from within
".ccWindow.top.retrieveCC invoke "
    invoked from within
".ccWindow.top.retrieveCC instate !disabled { .ccWindow.top.retrieveCC invoke } "
    invoked from within
".ccWindow.top.retrieveCC instate pressed { .ccWindow.top.retrieveCC state !pressed; .ccWindow.top.retrieveCC instate !disabled { .ccWindow.top.retriev..."
    (command bound to event)

[stevenaaus]

Last week Alexander applied a fix to XFCC. Perhaps it'll help, but perhaps not.
Add the line indicated below.

Code: Select all

--- tcl/tools/correspondence.tcl        (revision 2075)
+++ tcl/tools/correspondence.tcl        (working copy)
@@ -3925,6 +3925,7 @@
                                # tk_messageBox -title "Xfcc Oops" -type ok -icon warning -message $NoHTTPS
                                return 0
                        } else {
+                               ::tls::init -ssl3 false -ssl2 false -tls1 true
                                http::register https 443 ::tls::socket
                                set ::CorrespondenceChess::SeenHTTPS 1
                                return 1

[Zenmastur]

Steve,

What's the chance of getting the name comparison fixed.

I'm currently de-duping a database or 70,000 games. The program removed 30,000 duplicates games using the "Delete Twin Games" feature. Unfortunately it left several thousand duplicates in due to an improper name matching algorithm. I'm currently going through the DB by hand to find the rest of the dupes. I'm a quarter of the way done and have found about 1,000 dupes so far. Needless to say, this feature could use some improvements.

When deleting duplicate games, the program fails to find the duplicates because the name has different punctuation or capitalization. E.G. O'Donnel, ODonnel, Odonnel, ODONNEL and odonnel should all be evaluated the same. Also if a first and last name are given and there is a comma or other punctuation between them it should have no effect on the match. This currently causes a missed match. It would also be an improvement if the program ignored anything inside angle brackets, braces, or parenthisis etc. in the Name field.

There are several name matching projects on github etc that would greatly improve the name match rates. I'm not sure how easy they would be to incorporate into the program but they are worth a look.

Regards,

Forrest

[stevenaaus]

Yes, Tools->Maintenance->SpellcheckPlayerNames
must be run prior to duplicate removal.

Probably SpellcheckPlayerNames could be improved to better address wayward punctuation, but it's not something i'm too familiar with... Maybe this release i will look at it if no-one else does.

S.

[Zenmastur]

Yes, Tools->Maintenance->SpellcheckPlayerNames
must be run prior to duplicate removal.

Probably SpellcheckPlayerNames could be improved to better address wayward punctuation, but it's not something i'm too familiar with... Maybe this release i will look at it if no-one else does.

S.

I do use spell check before removing duplicates. I've even modified the files to recognize different variation that are common, But this can be tedious and doesn't solve all the problems. e.g. names with initials with and no period, one period, two periods, or commas, reversed initials, reversed name order etc. There are just too many combinations to do it manually.

Anything that could be done in the next release would be a big help!

Kind regards,

Forrest

[Norm Pollock]

Standardizing name variations is very important in editing a database. I use several of my own tools which are freely available at my site, see www below.

lI use either nameList or nameSimilar to list the names that are present. Then I use nameChange to change names to a single variation per player.

However this will not help where the player uses two differently spelled variations of the second name. Case in point, "Lagno, Kateryna" and "Lahno, Kateryna".

And another big issue is two different players with very similar names who give the impression that they are the same player. Case in point, "Pavlovic, Milos" and "Pavlovic, MilosM". You don't want to combine their games into one player.

[Zenmastur]

Standardizing name variations is very important in editing a database. I use several of my own tools which are freely available at my site, see www below.

lI use either nameList or nameSimilar to list the names that are present. Then I use nameChange to change names to a single variation per player.

However this will not help where the player uses two differently spelled variations of the second name. Case in point, "Lagno, Kateryna" and "Lahno, Kateryna".

And another big issue is two different players with very similar names who give the impression that they are the same player. Case in point, "Pavlovic, Milos" and "Pavlovic, MilosM". You don't want to combine their games into one player.

Actually, I have your utilities. I use them and they do help with this specific problem. But in the end I still have to manually edit the files which is a problem on large files. 95% of the problem is capitalization and punctuation. Ideally all comparisons should be done with all lower-case and all punctuation removed. This would be a simple change if the file is ascii. Not sure about how much more difficult it would be for Unicode.

There are more sophisticated and complex ways to do this type of matching, and there is code available that can be used free of charge, but like I said 95% of the problem can be handled by a simple change in the comparison routines.

One last item,

THX for all the 40H utilities. I use then on a regular basis.

Regards,

Forrest