leviramsey

I don't think anyone is talking about allowing posters to post arbitrary HTML.  Have you read my posts at all?

As it stands, every user of this site is posting HTML.  It's what CKEditor sends to the server.  It's what users who don't use CKEditor (either by disabling Javascript or via a mythical GreaseMonkey script to disable CKEditor) are sending to the server.  This is true regardless of what permissions are set.

EDIT to add: the scrubber does appear to close every tag opened.

A Friday nighter?

Friday morning, UK time

NFL is legally prohibited from playing games on Friday night (more properly, legally prohibited from showing them on TV within 100 miles of a high school football game).

Quick Reads

Between the two of them, Winston and Mariota have thrown a total of only 49 regular-season NFL passes, so of course it's far too early to draw any meaningful conclusions about the fate of either franchise or they're respective saviors. That said, it's hard to imagine how early returns could have been any more different. Mariota's Titans scored touchdowns on five of their first six possessions, and the reigning Heisman Trophy winner finished with 13 completions in 16 attempts for 209 yards, throwing for four touchdowns with no interceptions. Winston, on the other hand, struggled badly. His very first NFL pass was intercepted, and he finished up 16-of-33 for 210 yards with a pair of touchdowns to go with two picks. 

By our numbers, Mariota was not the best passer of the week, nor was Winston the worst. Both quarterbacks, though, made very powerful impressions, for good or ill. Was Mariota's game the best debut in the DVOA era? Was Winston's the worst? It's hard to say for sure, but the answers right now appear to be "quite possibly" and "no, but it sure was awful."

Time constraints stopped us from checking every rookie debut in our database, so I looked at the 55 quarterbacks between 1989 and 2014 who threw at least 200-ish passes in their rookie campaigns (Eli Manning threw 197 passes in 2004, so I went ahead and included him), then looked at how they fared in their first game with at least 15 pass attempts. (Those same time constraints stopped me from checking rushing data.) Not surprisingly, most rookies struggled in their first extended playing time -- 31 of those 55 finished below replacement level. This wasn't always the first game of their careers. In Matt Ryan's case, it wasn't even his first start. Ryan went 9-of-13 for 161 yards and a touchdown against Detroit in Week 1 of 2008, good for 37 DYAR, but for the purposes of this study I'm counting his debut as his 13-of-33, 158-yard, two-interception game against Tampa Bay seven days later. None of those 55 passers, though, matched the 155 passing DYAR Mariota posted against Tampa Bay. (In fact, Mariota's very first game will likely wind up among the top ten rookie games in DVOA history, a list that currently features Russell Wilson in first place.)

As for Winston, though few quarterbacks have gotten off to a worse start, he was a far sight better than the worst debut we have seen.

Very good rookie debuts: Mariota (2015), Garcia (1999), Griffin (2012), Newton (2011), Ramsey (2002), Freeman (2009), Sanchez (2009), Collins (1995), Manuel (2013), Young (2006)

Very bad rookie debuts: Weeden (2012), Plummer (1997), Winston (2015), Tannehill (2012), Stafford (2009), Banks (1996), McNabb (1999), Gabbert (2011), Walsh (1989), Ryan (2008)

(No, that's not a mistake in Young's statline. His performance is graded on a very steep curve because all his plays came with Tennessee down by multiple touchdowns in the second half, plus he had no interceptions or sacks.)

It's very important to note that the numbers for 2015 do not yet include opponent adjustments. We don't yet know how good the Tampa Bay secondary will be this year, so we don't know which way Mariota's numbers should be adjusted, or how strongly. We do know, though, that Tampa Bay was 23rd in pass defense last season, so we can reasonably guess that Mariota's debut might fall behind those of Jeff Garcia or Robert Griffin when all is said and done. Of course, Tennessee was 26th in pass defense in 2014, so the most likely result for Winston is that his numbers will also drop, though it's hard to imagine his debut would ever look worse than that of Brandon Weeden.

(Keep this in mind, by the way, when reviewing the quarterback tables below. Pittsburgh, New Orleans, and Chicago were all in the bottom 10 for pass defense DVOA last year.)

That's not good news for Winston or Bucs fans, but just how much does one game tell us? After all, look at the names in the "Very Good Debuts" table. Griffin, Josh Freeman, and Vince Young are some of the great flash-in-the-pans of recent times. Kerry Collins lasted well over a decade in the NFL, but was top-five in DVOA or DYAR just once (in 2002), and he's best known for the worst single-game playoff performance we've ever measured. Cam Newton is doing the best he can with a Conference-USA team in Carolina, and Jeff Garcia's career is nothing to be ashamed of, but the Tennessee will be disappointed if Mariota isn't clearly the best passer in this group when all is said and done. 

While one good game doesn't guarantee Mariota a spot in Canton, Winston's fate isn't sealed by one bad game, either. In fact, you could make a strong case that the quarterbacks with Very Bad Debuts, as a group, were better players than their Very Good Debut counterparts. Ryan and Donovan McNabb are/were both legit stars, better than anyone in the Very Good club. Matthew Stafford and Jake Plummer are/were a tier beneath them, but both gave their teams a chance to compete for the playoffs year in and year out. The Dolphins believe Ryan Tannehill can reach or surpass that level as well, thus his $96 million extension this past offseason. As for Brandon Weeden and Blaine Gabbert, well, I'm sure they're nice guys.

Best QB performances of the week

1.  Brady - You'll often hear about a quarterback "getting into a rhythm" or "figuring out the defense" or some other fancy way of saying "completing a bunch of passes in a row," but you won't often see a performance like Brady's against Pittsburgh last Thursday. You have probably heard that Brady at one point completed 19 passes in a row at one point. Now, there was a DPI and a sack mixed in that stretch, but whatever, that's cool. The thing is, 19 straight completions actually undersells how effective Brady was in the middle portion of this game. In today's NFL, many completions are actually bad plays. In 2014, only 57 percent of all completions gained first downs. With that in mind, I now inform you that in one stretch of this game Brady picked up 13 first downs in 14 dropbacks. And three of those were touchdowns! Brady completed 13 passes in that stretch for 136 yards, and also drew a DPI for 28 yards. His only "bad" play in that sequence was still a 3-yard gain on first-and-goal from the 9, and he threw a touchdown on the next snap. Brady had 20 total first downs on the day, but he wasn't completely infallible. He had six plays with more than 10 yards to go for a first down and failed to convert any of them, though he did have some successful plays, such as a 9-yard gain on first-and-20 and a 12-yard gain on second-and-15. Still, it looks like forcing the Patriots into long yardage is the only way to get them off the field.

2. Palmer - Bruce Arians loves him the long ball. Palmer threw 11 deep passes against New Orleans (tied with Ben Roethlisberger for most in the league this week) for 69 DYAR (second behind Nick Foles). He actually went 0-for-5 on passes that traveled at least 24 yards downfield, but in that 16- to 23-yard range, he went 4-for-4 for 86 yards and a touchdown, plus two DPIs for 40 more yards. The Cardinals also got big plays out of short passes, as a pair of completions to Darren Fells and David Johnson gained 4 yards in the air and 99 yards after the catch. (By the way, tight end looked like a giant weakness for Arizona this year, but it wasn't Sunday -- Fells also had a 12-yard gain on second-and-7 and a 17-yard touchdown.) The other key for Arizona was staying out of ultra-long yardage situations. Palmer never had more than 11 yards to go for a first down, and on those three throws with 11 to go, one was converted on a pass to John Brown, and another went to Jaron Brown for an 8-yard gain on second down.

3. Mariota - Perhaps the most amazing part of Mariota's day is that his first two passes were both incomplete passes to Bishop Sankey near the line of scrimmage. And then he apparently figured this whole football thing out. His next pass was a completion to Delanie Walker for 22 yards on third-and-10, and then the next play Mariota hit Kendall Wright for a 52-yard touchdown, and that was basically he that. He was sacked twice, and he did have one failed completion (a zero-yard gain to Sankey on second-and-20), but otherwise it's hard to find a flaw. The first five third-down plays of his NFL career: 22-yard gain with 10 yards to go; 22-yard gain with 6 yards to go; 4-yard touchdown; 1-yard touchdown; 16-yard DPI with 6 yards to go. He had one third-down play after that, a sack for a loss of 13, but by that point the Titans were up by 35 points in the third quarter.

4. Rodgers - There's nothing mind-blowing about Rodgers' overall numbers. He ranks this high mostly because he had no sacks, fumbles, or interceptions, and he was killer in the red zone, with all three touchdowns coming within 13 yards of the goal line. The Bears more or less limited the damage done by Rodgers' short passes, but he killed them with midrange and deep throws. He threw eight passes 10 or more yards past the line of scrimmage, completing five of them for five first downs (including a touchdown) and 113 yards. Another of those deep throws resulted in a 34-yard DPI.

5. Romo - First-half Tony Romo has cable. He went 18-of-24, but gained only 146 yards with eight first downs (including a touchdown) and an interception. Second-half Tony Romo, though subscribes to a popular satellite dish television company. He went 18-of-21 (!) for 210 yards, with 14 first downs (!!), including three touchdowns. He also threw an interception, but he did gain 16 more yards on a DPI. In particular, Second-half Tony Romo was dominant on first down, going 11-of-12 for 127 yards and nine first downs, including two scores. First-half Tony Romo was 23rd this week with -18 DYAR. Second-half Tony Romo had 151 DYAR, better than anyone except Philip Rivers.

6. Roethlisberger

7. Dalton

8. Taylor

9. Ryan

10. Rivers

11. A. Smith

12. Foles

13. Kaepernick

14. Mallett

15. Brees

16. Bradford

17. Fitzpatrick

18. Tannehill

19. Cousins

20. J. McCown

21. E. Manning

22. Luck

23. Carr

24. Cutler

25. Stafford

26. Newton

27. McGloin

28. Manziel

29. Wilson

30. Hoyer

31. Bridgewater - Bridgewater actually moved the ball pretty well on his own side of the field, but he was impotent in anything resembling scoring range. Inside the 49ers 45, he went 3-of-9 for 8 yards (seriously) and no first downs, while getting sacked three times for a net loss of 30 yards. That is -1.8 yards per play on 12 dropbacks. He also had no luck, and I do mean no luck, keeping drives alive. On third and fourth downs, he went 3-of-6 for 11 yards and no conversions, with three sacks for a net loss of 38 yards. That is -3.0 yards per play on nine dropbacks.

32. P. Manning - Manning had a very narrow strike zone in this game, and that strike zone came 6 to 10 yards past the line of scrimmage. In that range, he went 10-of-13 for 99 yards and nine first downs. Anything shorter than that, he went 13-of-21 for 58 yards with one first down and one pick-six; anything deeper, he went 1-of-6 for 18 yards and a first down. He was actually decent on third downs, going 10-of-16 for 94 yards and seven first downs, though he also had the pick-six and a sack. The problem was, he always needed those third-down conversions, because on first down he went 5-of-11 for only 23 yards and one first down.

33. Bortles - Credit where it's due: Bortles was really good on second down, where he went 8-of-12 for 72 yards and seven first downs, including a goal-line touchdown. Unfortunately, he was less successful on third and fourth downs, going 4-of-9 for 35 yards and three first downs, with three sacks. Both of Bortles fourth-down passes were targeted at Bryan Walters. Because when you absolutely need to keep a drive alive, you want to throw to a guy who couldn't stick with Seattle's receiving corps. The best weapon for Jacksonville might be the running back screen. Bortles threw four passes to Denard Robinson, T.J. Yeldon, and Corey Grant behind the line of scrimmage, resulting in three first downs and a 5-yard gain on first-and-10.

34. Winston - Making the worst of a bad situation: Winston had four plays with at least 20 yards to go for a first down, and was sacked three times. Well, at least he managed to throw incomplete on third-and-42, and no I am not making that up. Both of Winston's interceptions and one of his two fumbled snaps came in the first half. He was better after halftime, going 10-of-21 for 140 yards and seven of his 11 first downs.

35. Flacco - Between sacks and negative completions, Flacco lost yardage four times -- just as many times as he completed passes for for first downs. (He did add a fifth first down on a 2-yard DPI.) On Denver's half of the field, he went 3-of-9 for 13 yards, with one interception and no first downs, plus that one DPI. Throwing to the short right side of the field, he went 5-of-9 for 5 yards (not a typo, I promise you) with no first downs and a pick-six, plus that one DPI. On first downs, he went 8-of-11, but only gained 38 yards and no first downs. You get the idea.

Best RB performances of the week

1. Carlos Hyde (49ers)

2. Matt Forte (Bears)

3. Darren Sproles (Eagles)

4. DeAngelo Williams (Steelers)

5. Chris Ivory (Jets)

Worst RB performance of the week: Justin Forsett (Ravens)

Best WR/TE performance of the week

1. Julio Jones (Falcons)

2. James Jones (Packers)

3. Antonio Brown (Steelers)

4. Travis Kelce (Chiefs)

5. Kendall Wright (Titans)

Worst WR/TE performance of the week: Steve Smith (Ravens)

4th: 0.1%

5th: 0.1%

6th: 0.2%

7th: 0.6%

8th: 1.1%

9th: 1.6%

10th: 2.5%

11th: 3.7%

12th: 5.3%

13th: 7.1%

^^^ upper quartile

14th: 9.0%

15th: 10.0%

16th: 11.2% == median

17th: 12.0% == most likely table

18th: 11.9%

vvv lower quartile

19th: 12.1%

20th: 11.5%

Relegation chances

Sunderland: 52.3%

Stoke: 48.7%

Liverpool: 42.8%

Bournemouth: 42.3%

Villa: 35.5%

Newcastle: 32.5%

Norwich: 28.2%

West Ham: 7.7%

Southampton: 2.8%

Spurs: 2.5%

Chelsea: 1.7%

Palace: 1.6%

WBA: 1.0%

Swansea, Leicester: 0.2%

From examination of the HTTP request headers, it's clear that CKEditor transmits HTML to the server, which is then scrubbed by the server to the known-safe subset of HTML.  The question is then whether the source button in CKEditor bypasses scrubbing.  Without inspection of the HTTP traffic, I can't answer that question (if an admin/mod wants to check...).

From documentation for IPB 3.4 with CKEditor, it seems that whether the source button is displayed for everybody is configurable independent of HTML permissions, though it defaults to Source iff HTML.  Assuming that the IPB permission governs the server-side scrubber, then this means that everyone can get the source button.

IFrames?

One possible workaround this suggests, if you're willing to take advantage of the subset of HTML, is that if you can prevent CKEditor from loading (either via disabling Javascript or via GreaseMonkey), you get a raw textarea wherein you can post HTML.  Disabling Javascript disables the quote mechanism, but a tailored GreaseMonkey script might be able to preserve quote behavior.

Alternatively, if admins/mods are normally restricted to the same subset of HTML (e.g. no blockquotes), then it would seem that all users could be given permission to post HTML.

<img src="http://i.imgur.com/MFmtqol.jpg">

A little experimentation shows that if Javascript is disabled, any user can post at least some restricted subset of HTML (tested to work: bold, italic, link, image; tested not to work: tables, blockquotes), so it's not the editor which is enforcing HTML restrictions (which is good news in about eleventy-billion ways).

This should be indented

On a related note, any idea why no jQuery or similar client-side library for doing DOM manipulations?

Have cleared cookies and cache.

I can post with Javascript enabled if I go into page debugging and disable the "display: none" styling for the textarea with name "topic_comment_$TOPIC_ID" (the TOPIC_ID being found in the URL www.villatalk.com/topic/${TOPIC_ID}-${TOPIC_TITLE} ... e.g. this topic is ID 13921. I'll do a little digging through the Javascript and submit a patch.

test

The new layout is practically unusable on a large (wide) screen, especially in a browser (Chromium 43 on Linux) which doesn't allow the tab bar to be on the side.

Posting does work in Chromium.

Possibly related issue: cannot edit posts, with or without javascript

Browser: Opera/12.16 on Linux (which is listed as a supported browser by the editor which this site uses)

All that's visible is the file upload box.

Posting does work in Chromium.

Predictions for every regular season game made. My expected win totals (make playoffs/in the hunt):

AFC East

Patriots: 10.9

Bills: 10.1

Jets: 7.8

Dolphins: 6.3

AFC North

Bengals: 9.2

Ravens: 7.1

Steelers: 6.7

Browns: 4.6

AFC South

Texans: 10.8

Colts: 8.4

Titans: 6.0

Jaguars: 5.2

AFC West

Broncos: 13.6

Chiefs: 8.3

Chargers: 8.2

Raiders: 6.5

NFC East

Eagles: 10.6

Cowboys: 7.7

Giants: 6.7

Redskins: 5.3

NFC North

Packers: 10.8

Vikings: 7.3

Lions: 6.5

Bears: 4.4

NFC South

Panthers: 8.6

Saints: 8.5

Falcons: 7.8

Bucs: 7.3

NFC West

Seahawks: 13.7

49ers: 9.8

Rams: 7.1

Cardinals: 4.4

That's a hell of a detour.  Would you not be better flying in to Montreal first, spending whatever time you want there and then flying to New York to start the road trip from there?  I'm doing something similar (but a lot shorter) in October except it's (fly in to) Toronto, (train down to) Hartford CT & (road trip to) Charles Town WV.

It's not that big of a detour: a couple of hundred miles on a three thousand mile trip.