ARCHIVED POSTS
< Earlier Kibitzing · PAGE 950 OF 1118 ·
Later Kibitzing> |
Sep-26-16
 | | chessgames.com: There is no problem at all; the game in question was rejected not because of the $1 but because it has been in the database for many years: Fischer vs M Shifrine, 1964 |
|
| Sep-26-16 | | zanzibar: <chessgames> that's more like it, and would indeed have been quite surprising otherwise. * * * * *
For those unfamilar with PGN NAG's(*):
$1 = !
Fairly standard stuff, and easy to parse/strip out. (*) NAG = Numeric Annotation Glyphs
https://en.wikipedia.org/wiki/Numer... |
|
Sep-26-16
 | | WannaBe: ⨀ |
|
Sep-27-16
 | | OhioChessFan: Per putting the Best Games of 2016 link on the home page, not a big deal, but wouldn't that ensure the best games will all be early in the year? I realize the WCC doesn't enter into what I'm saying, but still. The early games get the normal traffic during the tournament or match they were played, then show up (at least indirectly) every day the rest of the year on the Best Of link on the home page, drawing more visits, and create a self perpetuating spot in the top 10 to the detriment of games later in the year. Not a big deal to worry about, but my mind works in strange ways sometimes. |
|
Sep-27-16
| | chessgames.com: OCF, not the homepage per se but rather this page: Notable Chess Games. You are right, these lists change over time, and the early games have a kind of head-start. If the most mind-blowing masterpiece of 2016 happens to be played in December, it will be unlikely to appear on the top-10 list until sometime in 2017. Cleverly, the notability algorithm now makes allowances for the head start effect. We don't just compute a 'score' for each game and then list the highest ten. These scores have to first be weighted in proportion to the logarithm of their ages. |
|
| Sep-27-16 | | zanzibar: <CG> fixup needed: Louis Stumpers (kibitz #7767) Maybe we should have a standard depot for these minor nits? |
|
Sep-28-16
| | chessgames.com: <Maybe we should have a standard depot for these minor nits?> We do, the Biographer Bistro. In this case WorstPlayerEver did a fine job by posting the info on the page. (Although due to the unusual nature of that page it is destined to be buried and forgotten very soon, so thanks for pointing it out before that happened.) The best thing, I think, is that editors keep an eye out for that kind of thing and then act on it when they thing appropriate. Whether WorstPlayerEver's info is appropriate to add, I don't know. Remember that Chessgames Bios focus on the chess career of the individual and only mention other accomplishments as a side note, if at all. The biography of Napoleon Bonaparte is a humorous example. Note that we have more to say about his chess interests than that coup d'etat and the domination of Europe. |
|
| Sep-28-16 | | zanzibar: Ah, sorry <chessgames>, I was specifically referring to a depot for mangled links in the previous post. I should have made that clearer.
https://en.m.wikipedia.org/wiki/CCIR_(selcall)
should be
https://en.m.wikipedia.org/wiki/CCI... But do we really have to educate all <CG> users about <& l p a r ; ... & r p a r ; % 2 0> trickery (or would that be witchcraft)?
Or can we finally get a simple escape sequence syntax in place, one that always works without too much thought? As far was what to include/exclude from Stumpers' bio, I'll leave that mostly to <Tab>'s discretion. .
. |
|
Sep-28-16
| | chessgames.com: <Or can we finally get a simple escape sequence syntax in place, one that always works without too much thought?> Not that I'm against some kibitzing improvements, but I really think the small set of users who would be aware of the escape-tricks would be the same set of users who know how to use ∥ and other tricks. However, since wikipedia specifically has adopted the habit of parens-inside-URLs, perhaps some improvement to the code could make it not be fooled by that practice. After all, there's a big difference between (http://www.someurl.com/blablabla.html)
and
http://www.someurl.com/blablabla_(example).html
But the software is rather dumb to the difference. If the software merely was able to not be tricked in the second case, there really wouldn't be much problem. |
|
| Sep-28-16 | | zanzibar: <... but I really think the small set of users who would be aware of the escape-tricks would be the same set of users who know how to use ∥ and other tricks.> Apparently not, at least in <WPE>'s case (who I don't think is a numpty). Lots of people blithely cut-and-paste links, it's natural - that's why I suggested tracking them in a depot. <CG> already has the data, and could do a quick scan to filter out 50-100 examples of a link followed by a paren (or other punctuation) - and then do a rough count by eye of how many are mangled links vs. people taking advantage of <CG>'s "convenience" features. I'd be interested in that simple metric. |
|
Sep-28-16
 | | juan31: chessgames.com: maybe it's my tablet (not applied before) the pages appear gradually, from top to bottom |
|
Sep-28-16
| | WannaBe: <juan31> Sounds like your internet (connection) speed, were you on your data plan or on WiFi? Do you know how fast your data plan (3G, 4G) or your WiFi connection is/was? |
|
Sep-28-16
| | juan31: Gracias <WannaBe> now its ok. |
|
Sep-28-16
| | WannaBe: <juan31> De nada. =) |
|
Sep-29-16
| | chessgames.com: <zanzibar> Good news, I have made some code changes that stop mangling wikipedia (etc.) URLs as described. My test suite is:
+++++++
https://en.wikipedia.org/wiki/Frank...(chess_player) (http://www.reddit.com/r/chess)
(https://en.wikipedia.org/wiki/Frank...(chess_player)) (lorem ipsum http://www.reddit.com/r/chess)
+++++++
The code parses all of those URLs correctly, even the one where somebody put parens around a wikipedia link. However the code is not live on this server, it's on the development server. Look for it in a week or so. My regexes look like this
$blob =~ s/(?<!\()(https?:\/\/[^\s"<>]+)/makelink($1)/ige-
;
$blob =~ s/\((https?:\/\/[^\s"<>]+)\)/'('.makelink($1).')-
'/ige;
I think with enough cleverness I can combine it into one regex but even as it stands, it seems to work fine. |
|
| Sep-29-16 | | zanzibar: <chessgames> I'm always a bit slow with complicated regexp's, not to mention my Perl is a bit rusty, so this reply isn't at the detail level... These incremental improvements should help improve the situation as it stands today. I still have a general dislike for having to chase down one special conditon after another, so I'm maintaining advocacy for a general syntax allowing link and verbatim formatting akin to WordPress and/or ChessTempo style formatting... |
|
Sep-29-16
| | chessgames.com: <I'm always a bit slow with complicated regexp's> Don't worry, that's why we have me. $blob=~s/(?<!\()(https?:\/\/[^\s"<>]+)/makelink($-
1)/ige;
$blob=~s/\((https?:\/\/[^\s"<>]+)\)/'('.makelink(-
$1).')'/ige;
A thing of beauty, isn't it? (But don't try to speak it aloud; it might summon Cthulu.) To translate that into English:
First, exclamation points are no longer URL terminators like they used to be. The number of instances somebody says "I love www.lichess.com!!!" are very few compared to people posting Google Maps and certain Instagram pages, etc. So that's become a relic and won't bother us anymore. But more relevant is now the close-paren is regarded as an URL terminator if-and-only-if the http:// part was preceded by an open-paren. That means that while I can still say things like (http://www.lichess.com) I can also link directly to a Wikipedia URL and it won't do that "very stupid thing" that's caused all those broken links. I totally agree with you that a quote-delimited method or some other foolproof means would be nice, but even nicer is if you could just plop URLs in place and have them work without any further fuss. So once we get that working for 99% of the cases (and I think we are there) then we can focus on the remaining 1%. |
|
| Sep-29-16 | | WinKing: <chessgames.com: A thing of beauty, isn't it? (But don't try to speak it aloud; it might summon Cthulu.)> http://blog.flicks.co.nz/wp-content... |
|
Sep-29-16
| | WannaBe: Worse yet, it might summon deadites (google ash vs evil dead). |
|
| Sep-29-16 | | zanzibar: Did you know that a Goat shows up in the Lovecraft mythos?! I didn't know that. |
|
| Sep-29-16 | | zanzibar: <I totally agree with you that a quote-delimited method or some other foolproof means would be nice, but even nicer is if you could just plop URLs in place and have them work without any further fuss. So once we get that working for 99% of the cases (and I think we are there) then we can focus on the remaining 1%.> Perhaps we are there, let's see how these changes work out then... But I'm use to the following two systems:
<Wordpress>:
http://www.laptoptips.ca/wpd/wp-con... and
<ChessTempo>:
http://chesstempo.com/chess-forum/i... (btw- I hate the emoticons and scrolling marquee. IIRC those features can be turned off). Both offer up a set of toolbars to help "plop" your edits in place, and have 100% certainty what gets done with them. This approach might be a bit too heavy for <CG>, as the die has already been cast. But, on the other hand, having [url][/url] and [pre][/pre] escapes could be added without impacting how people currently use <CG> if they don't want to use such new features. |
|
| Sep-30-16 | | Abdel Irada: ∞
<zanzibar: Did you know that a Goat shows up in the Lovecraft mythos?!I didn't know that.>
You just have to be careful saying her name, lest someone think you're uttering a racial slur. ∞ |
|
| Sep-30-16 | | Abdel Irada: ∞
I do have a serious request, though.
Could we have a "Search [kibitzes] by forum" option added to "Kibitzing Search"? I think it would be enormously useful. ∞ |
|
| Sep-30-16 | | zanzibar: I second <Abdel>'s request. |
|
Sep-30-16
| | chessgames.com: <Could we have a "Search [kibitzes] by forum" option added to "Kibitzing Search"? I think it would be enormously useful.> If I understand you, not just chess forums, but any page. Search Bobby Fischer's page for the word "Kissinger", things like that. I agree it would be a big improvement. The query is easy; the presentation and user-interface is what's caused me to drag my feet on that project. Right now new coding is temporarily on hiatus for a week or so but I will keep that in mind. |
|
 |
|
ARCHIVED POSTS
< Earlier Kibitzing · PAGE 950 OF 1118 ·
Later Kibitzing> |
Kibitzing
