For me thesis, I want to suppress all badbox message until it is somewhat finished to concentrate on the more pressing warnings. As suggest here https://tex.stackexchange.com/a/164889/81936 I tried using the silence package. This did not change anything. I guess because badbox is actually neither a warning nor an error.

As suggested here https://tex.stackexchange.com/a/50850/81936, I added


to my header. This also had no effect. The structure of my thesis looks like this

%many chapters that throw badbox messages

I set the \hfuzz stuff in the header.tex file.

  • what messages do you get? (you could set hbadness=10000 to silence more) Nov 11, 2015 at 14:24
  • I have set hbadness=10000. I stil get bad box message. E.g. Overfull \hbox (16.9164pt too wide) has occurred while \output is active \OT1/cmr/m/sl/10 CHAPTER 5. ADVANTAGES OF GAUSSIAN PROCESS PANEL MODELING \OT1 /cmr/m/n/10 35 [] and Overfull \hbox (200.03493pt too wide) in paragraph at lines 647--648 [][] [] Nov 16, 2015 at 15:26
  • overfull box warnings are do not report "badness" so are not affected by that, you could set \hfuzz to some value > 17pt to silence that message, if you have set it as you show but still see the message something is setting it back. Hard to say without a real example to reproduce. Nov 16, 2015 at 16:43
  • 1
    most likely you (or a package you are using) have used \sloppy or \fussy both of which set \hfuzz to less than 16.9pt. Nov 16, 2015 at 16:46
  • @DavidCarlisle this is also what I suspected. Therefore, I placed \hfuzz=10000pt after all packages have been loaded. Apparently, this does not solve the problem. Is there a way to declare hfuzz as a constant, such that changing its value will either be permitted or throws an error? Nov 16, 2015 at 19:46

1 Answer 1


While reducing some warnings while drafting is probably good, silencing badness 10000 (infinitely bad) or overfull 200pt (probably off the page) boxes seems like it is taking that idea to extremes but if you want to do that set


to set the primitives to their maximum effective values, then do


to replace those commands by registers so that any later settings by other package code will have no effect.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.