Autor Beitrag
FinnO
ontopic starontopic starontopic starontopic starontopic starontopic starofftopic starofftopic star
Beiträge: 1331
Erhaltene Danke: 123

Mac OSX, Arch
TypeScript (Webstorm), Kotlin, Clojure (IDEA), Golang (VSCode)
BeitragVerfasst: Mi 04.01.12 01:52 
Moin,

mir ist beim Antworten gerade aufgefallen, dass wenn man unten in der sog. "Thema-Überblick"-Box die Quelltexte ausklappt, die Zeilen in der Spalte "Zeile" nicht mit den Zeilen des Quelltextes Übereinstimmen. Sieht recht nutzlos aus.

zeileverrutscht

LG
Einloggen, um Attachments anzusehen!
Martok
ontopic starontopic starontopic starontopic starontopic starontopic starofftopic starofftopic star
Beiträge: 3661
Erhaltene Danke: 604

Win 8.1, Win 10 x64
Pascal: Lazarus Snapshot, Delphi 7,2007; PHP, JS: WebStorm
BeitragVerfasst: Mi 04.01.12 02:05 
SUUUPER, jetzt holt Firefox die alten Darstellungsbugs vom IE7 wieder aus der Schublade oder was?

Lass mich raten, das wird pro Zeile die etwas fettgedrucktes enthält "falscher"?

_________________
"The phoenix's price isn't inevitable. It's not part of some deep balance built into the universe. It's just the parts of the game where you haven't figured out yet how to cheat."
Boldar
ontopic starontopic starontopic starontopic starontopic starhalf ontopic starofftopic starofftopic star
Beiträge: 1555
Erhaltene Danke: 70

Win7 Enterprise 64bit, Win XP SP2
Turbo Delphi
BeitragVerfasst: Mi 04.01.12 04:59 
Hi,
Also ich bin hier mit FF 9.0.1 online, gerade frisch installiert ohne jegliche Addons, und der Fehler tritt hier definitiv nicht auf, auch nicht bei längeren Quelltexten. BTW ist im Screen der Code zu klein, die Markierte Zeile mit procedure ist bei mir schön brav Nummer 53 (Ich denke doch es geht um dieses Topic). Deshalt
Ich kann den Fehler also nicht nachvollziehen.
mfg Boldar
jaenicke
ontopic starontopic starontopic starontopic starontopic starontopic starontopic starofftopic star
Beiträge: 19272
Erhaltene Danke: 1740

W11 x64 (Chrome, Edge)
Delphi 11 Pro, Oxygene, C# (VS 2022), JS/HTML, Java (NB), PHP, Lazarus
BeitragVerfasst: Mi 04.01.12 09:27 
Du hast aber auch in der Suche beim Ausklappen in der Beitragsansicht geschaut? Dort sieht es bei mir im FF zu Hause genauso aus. (Welche FF Version weiß ich nicht.)

Tja, soweit ist es gekommen:
Jetzt kopiert der FF nicht nur das Aussehen des Browsers nahezu 1:1, sondern auch das der Webseiten. Vielleicht wollte man da mal von nem anderen Browser kopieren. :mrgreen:
baka0815
ontopic starontopic starontopic starontopic starontopic starontopic starontopic starofftopic star
Beiträge: 489
Erhaltene Danke: 14

Win 10, Win 8, Debian GNU/Linux
Delphi 10.1 Berlin, Java, C#
BeitragVerfasst: Mi 04.01.12 10:53 
Kann es nicht vielleicht einfach daran liegen, dass bei dir eine andere Schriftart verwendet wird, die etwas anders skaliert?
FinnO Threadstarter
ontopic starontopic starontopic starontopic starontopic starontopic starofftopic starofftopic star
Beiträge: 1331
Erhaltene Danke: 123

Mac OSX, Arch
TypeScript (Webstorm), Kotlin, Clojure (IDEA), Golang (VSCode)
BeitragVerfasst: Mi 04.01.12 11:31 
user profile iconMartok hat folgendes geschrieben Zum zitierten Posting springen:
Lass mich raten, das wird pro Zeile die etwas fettgedrucktes enthält "falscher"?


Indeed.
jaenicke
ontopic starontopic starontopic starontopic starontopic starontopic starontopic starofftopic star
Beiträge: 19272
Erhaltene Danke: 1740

W11 x64 (Chrome, Edge)
Delphi 11 Pro, Oxygene, C# (VS 2022), JS/HTML, Java (NB), PHP, Lazarus
BeitragVerfasst: Mi 04.01.12 11:38 
FF 10 zeigt das Verhalten übrigens ebenfalls.
baka0815
ontopic starontopic starontopic starontopic starontopic starontopic starontopic starofftopic star
Beiträge: 489
Erhaltene Danke: 14

Win 10, Win 8, Debian GNU/Linux
Delphi 10.1 Berlin, Java, C#
BeitragVerfasst: Mi 04.01.12 13:44 
FF 9.0.1 zeigt das Verhalten bei mir nicht, daher meine Vermutung bzgl. der Schrift.

Evtl. auch ein installiertes Add-On?
FinnO Threadstarter
ontopic starontopic starontopic starontopic starontopic starontopic starofftopic starofftopic star
Beiträge: 1331
Erhaltene Danke: 123

Mac OSX, Arch
TypeScript (Webstorm), Kotlin, Clojure (IDEA), Golang (VSCode)
BeitragVerfasst: Mi 04.01.12 14:21 
Moin,

habe keine AddOns, keine Schrift verändert.

LG
jaenicke
ontopic starontopic starontopic starontopic starontopic starontopic starontopic starofftopic star
Beiträge: 19272
Erhaltene Danke: 1740

W11 x64 (Chrome, Edge)
Delphi 11 Pro, Oxygene, C# (VS 2022), JS/HTML, Java (NB), PHP, Lazarus
BeitragVerfasst: Mi 04.01.12 14:30 
user profile iconbaka0815 hat folgendes geschrieben Zum zitierten Posting springen:
FF 9.0.1 zeigt das Verhalten bei mir nicht, daher meine Vermutung bzgl. der Schrift.
Dann zeig doch einfach mal einen Screenshot wie es bei dir aussieht. ;-)
Dann lässt sich das ja vergleichen.
Boldar
ontopic starontopic starontopic starontopic starontopic starhalf ontopic starofftopic starofftopic star
Beiträge: 1555
Erhaltene Danke: 70

Win7 Enterprise 64bit, Win XP SP2
Turbo Delphi
BeitragVerfasst: Mi 04.01.12 15:27 
Also hier siehts so aus:
Moderiert von user profile iconKha: Überbreites Bild entfernt, s. Anhang
Einloggen, um Attachments anzusehen!
jaenicke
ontopic starontopic starontopic starontopic starontopic starontopic starontopic starofftopic star
Beiträge: 19272
Erhaltene Danke: 1740

W11 x64 (Chrome, Edge)
Delphi 11 Pro, Oxygene, C# (VS 2022), JS/HTML, Java (NB), PHP, Lazarus
BeitragVerfasst: Mi 04.01.12 16:30 
In dem Screenshot bist du ja an einer anderen Stelle. Bei mir ist es in der Suche so.

// EDIT:
Und in der von dir geposteten Topicansicht auch sehe ich hier. Aber das war zu Hause nicht so.


Zuletzt bearbeitet von jaenicke am Mi 04.01.12 18:02, insgesamt 1-mal bearbeitet
Martok
ontopic starontopic starontopic starontopic starontopic starontopic starofftopic starofftopic star
Beiträge: 3661
Erhaltene Danke: 604

Win 8.1, Win 10 x64
Pascal: Lazarus Snapshot, Delphi 7,2007; PHP, JS: WebStorm
BeitragVerfasst: Mi 04.01.12 17:00 
Hat jemand der Betroffenen ein Firebug da?
Dann deaktiviert mal in common.css, Regel .codekey (Zeile 197) die "font-size: 12px" Angabe.

Wird's dann besser? Wenn ja, haben sie zwar keinen IE-Bug kopiert, sind aber dafür die ersten die diesen Bug (bzw dessen IE-Fix) nicht verstehen, was dann im Endeffekt genauso falsch ist ;)

_________________
"The phoenix's price isn't inevitable. It's not part of some deep balance built into the universe. It's just the parts of the game where you haven't figured out yet how to cheat."
jaenicke
ontopic starontopic starontopic starontopic starontopic starontopic starontopic starofftopic star
Beiträge: 19272
Erhaltene Danke: 1740

W11 x64 (Chrome, Edge)
Delphi 11 Pro, Oxygene, C# (VS 2022), JS/HTML, Java (NB), PHP, Lazarus
BeitragVerfasst: Mi 04.01.12 18:01 
user profile iconMartok hat folgendes geschrieben Zum zitierten Posting springen:
Dann deaktiviert mal in common.css, Regel .codekey (Zeile 197) die "font-size: 12px" Angabe.

Wird's dann besser?
Ja, dann sieht es korrekt aus. ;-)
Martok
ontopic starontopic starontopic starontopic starontopic starontopic starofftopic starofftopic star
Beiträge: 3661
Erhaltene Danke: 604

Win 8.1, Win 10 x64
Pascal: Lazarus Snapshot, Delphi 7,2007; PHP, JS: WebStorm
BeitragVerfasst: Mi 04.01.12 21:00 
Okay, dann muss zu der IE7-Browserweiche noch eine Fx9-Browserweiche.

Hintergrund: im IE7 werden span-Tags, die italic sind ein Pixel höher als die Zeile drumrum. So ziemlich alle Highlighter generieren für die daher eine Höhe die exakt ein Pixel kleiner ist als "normaler" Text. Richtige Browser haben das bisher ignoriert/weg-overflowed/wasauchimmer.
Firefox nimmt das jetzt wörtlich und macht die ein Pixel kleiner, interessanterweise auch bold. Das bricht uns hier das Genick.

Ich werd mir den mal beschaffen und einen Testcase bauen, aber ich garantier für nix.
Spaß am Rande: in Fx9 schwanken die Zeilenhöhen zwischen 14px und 15px, in allen anderen Browsern sind sie konstant 16px :wall:


EDIT: äh wie jetzt?
ScreenShot-5059

Also ne Extension? *grübel*
Der Affe ists nicht, sagt user profile iconFinnO
Einloggen, um Attachments anzusehen!
_________________
"The phoenix's price isn't inevitable. It's not part of some deep balance built into the universe. It's just the parts of the game where you haven't figured out yet how to cheat."
Regan
ontopic starontopic starontopic starontopic starontopic starhalf ontopic starofftopic starofftopic star
Beiträge: 2157
Erhaltene Danke: 72


Java (Eclipse), Python (Sublimetext 3)
BeitragVerfasst: Mi 04.01.12 21:46 
Gerade Firefox neu heruntergeladen und installiert. Erste Seite aufgerufen:
ff

Edit: ich hatte ihn vorher noch nie auf dem System (Windows 7 Home Premium) installiert.
Einloggen, um Attachments anzusehen!
Martok
ontopic starontopic starontopic starontopic starontopic starontopic starofftopic starofftopic star
Beiträge: 3661
Erhaltene Danke: 604

Win 8.1, Win 10 x64
Pascal: Lazarus Snapshot, Delphi 7,2007; PHP, JS: WebStorm
BeitragVerfasst: Mi 04.01.12 22:01 
Welches OS haben die betroffenen? Bisher stammen alle Meldungen von Win7 x64.

Also: bitte mal testen, ob das bei euch geht und mit User-Agent-String und ggf Edgemonkey und dem Wert der Einstellung Design->"Codeblöcke als monospace anzeigen" zurückmelden!

ausblenden volle Höhe For your convenience: some sourcecode
1:
2:
3:
4:
5:
6:
7:
8:
9:
10:
11:
12:
13:
14:
15:
16:
17:
18:
19:
20:
21:
22:
23:
24:
25:
26:
27:
28:
29:
30:
31:
32:
33:
34:
35:
36:
37:
38:
39:
40:
41:
42:
43:
44:
45:
46:
          A:= NextR(i);
          if Tokens[A].Kind=tokEmpty then
            TE_FunctionCall(tmp.GetObject).Arguments:= nil
          else
            TE_FunctionCall(tmp.GetObject).Arguments:= Tokens[A].Expr;
          Tokens[A].Expr:= nil;
          Tokens[a].Kind:= tokVoid;
          Tokens[i].Expr:= tmp;
        end;
      end// Zeile 10

    //finally, operators
    // Note: this wastes some extra iterations per Prio-Class, just ignore that, okay?
    for A:= 0 to high(Expressions) do begin
      for i:= L to R do
        if (Tokens[i].Kind=tokOperator) and
           (Expressions[Tokens[i].OpIdx].P=Expressions[A].P) then begin
          eid:= Tokens[i].OpIdx;
          rr:= NextR(i);
          ll:= NextL(i); // Zeile 20

          if rr<0 then
            raise ESyntaxError.CreateFmt('Position %d: Operator has no RHS',[Tokens[i].Pos]);

          if Tokens[rr].Kind<>tokExpression then
            raise ESyntaxError.CreateFmt('Position %d: expected expression, found %s',[Tokens[rr].Pos, Tokens[rr].Value]);
 // Zeile 27
          if (Expressions[eid].Cls = TE_Subtraction) and
             ((ll<0or (Tokens[ll].Kind<>tokExpression)) then begin
            tmp:= TE_Negation.Create;//Zeile 30
            tmp.RHS:= Tokens[rr].Expr;
          end else begin
            tmp:= Expressions[eid].Cls.Create;

            tmp.RHS:= Tokens[rr].Expr;
            if not Expressions[eid].Unary then begin
              if ll<0 then
                raise ESyntaxError.CreateFmt('Position %d: Operator has no LHS',[Tokens[i].Pos]);
              if Tokens[ll].Kind<>tokExpression then
                raise ESyntaxError.CreateFmt('Position %d: expected expression, found %s',[Tokens[ll].Pos, Tokens[ll].Value]); //Zeile 40
              tmp.LHS:= Tokens[ll].Expr;
1           end;
2         end;
3       end;
4     end;
5   end;



ausblenden volle Höhe Delphi-Quelltext
1:
2:
3:
4:
5:
6:
7:
8:
9:
10:
11:
12:
13:
14:
15:
16:
17:
18:
19:
20:
21:
22:
23:
24:
25:
26:
27:
28:
29:
30:
31:
32:
33:
34:
35:
36:
37:
38:
39:
40:
41:
42:
43:
44:
45:
46:
47:
48:
49:
50:
51:
KeinHighlighter;
end
end
  end
  end
    end
    end
  end
  end
  end
end//10
end
end
end
end
end
end
end
end
end
end//20
end
end
end
end
end
end
end
end
end
end//30
end
end
end
end
end
end
end
end
end
end//40
end
end
end
end
end
end
end
end
end
end //50

_________________
"The phoenix's price isn't inevitable. It's not part of some deep balance built into the universe. It's just the parts of the game where you haven't figured out yet how to cheat."


Zuletzt bearbeitet von Martok am Mi 04.01.12 22:47, insgesamt 6-mal bearbeitet
FinnO Threadstarter
ontopic starontopic starontopic starontopic starontopic starontopic starofftopic starofftopic star
Beiträge: 1331
Erhaltene Danke: 123

Mac OSX, Arch
TypeScript (Webstorm), Kotlin, Clojure (IDEA), Golang (VSCode)
BeitragVerfasst: Mi 04.01.12 22:03 
Ich muss dich enttäuschen, aber dein Code passt.

Also ich habe mal verschiedene Zeichenfolgen am Start.

komp

Ferner:

ausblenden Delphi-Quelltext
1:
2:
3:
end
    end
        end

ausblenden Delphi-Quelltext
1:
2:
3:
end
end
end


Sieht so aus:

Passt
Einloggen, um Attachments anzusehen!


Zuletzt bearbeitet von FinnO am Mi 04.01.12 22:50, insgesamt 13-mal bearbeitet
Regan
ontopic starontopic starontopic starontopic starontopic starhalf ontopic starofftopic starofftopic star
Beiträge: 2157
Erhaltene Danke: 72


Java (Eclipse), Python (Sublimetext 3)
BeitragVerfasst: Mi 04.01.12 22:10 
Seconded.
Ich habe gerade mal den Button neben den "Code aufklappen"-Button gedrückt: Die Größe verändert sich nicht. Es kann also nicht an der Scrollbox liegen.
64-bit Windows.

Edit: habe keinen Edgemonkey installiert und will es eigentlich auch nicht installieren, da der FF nachher wieder runterfliegt.
Edi2: Es kann nicht an der Länge liegen. Bei kleineren "Längen" wird der Fehler bloß nicht so offensichtlich.
Edit3: User-Agent: Mozilla/5.0 Windows NT 6.1 WOW64 rv 9.0.1 Gecko/20100101 Firefox/9.0.1
Yogu
ontopic starontopic starontopic starontopic starontopic starontopic starontopic starhalf ontopic star
Beiträge: 2598
Erhaltene Danke: 156

Ubuntu 13.04, Win 7
C# (VS 2013)
BeitragVerfasst: Mi 04.01.12 23:18 
Ich hab Firefox 10 Beta 32 Bit auf 64-Bit-Win7, und sehe weder in diesem Topic, noch in den hier geposteten Codes einen Versatz. Bei allen Textelementen (fett, kursiv oder normal) gibt Firebug eine Höhe von 16 Pixeln an. Ohne den 12px-Workaround für fetten und kursiven Text wären es auch 16 Pixel Höhe. Also eigentlich alles so, wie du es vom Firefox erwartet hast.

User Agent:
ausblenden Quelltext
1:
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:10.0) Gecko/20100101 Firefox/10.0