0

When trying to compile a Latex document with many tables, I was encountering the following problem:

Too many unprocessed floats.

In order to solve the issue, I followed the suggestion of this answer and included the following code line:

\usepackage[morefloats=100]{morefloats}

My question is: could there be any adverse side effects from using this line of code?

10
  • Did you see the other answer? tex.stackexchange.com/a/241006/73
    – Joseph Wright
    Commented Jan 6, 2020 at 15:08
  • do you have a very old latex installation? I would not expect you to get that error in any release since 2015 Commented Jan 6, 2020 at 15:26
  • 1
    on an old release the only bad effect is that you take more boxes for floats so you may run out of boxes for other things, but that mattered when there were 256 boxes but as latex uses etex by default there are 32 thousand box registers so running out is less likely assuming you have a new enough format Commented Jan 6, 2020 at 16:06
  • 1
    oh pre-historic:-) that pushes you back before 2015 so the quiet period so it's basically from around 2000 with no changes, Commented Jan 6, 2020 at 16:09
  • 1
    note there have been at least 4 updates to the package since the date of your latex (and more recent versions would not work with such an old release) so asking about any potential issues with the current version of morefloats will not really tell you anything useful. Commented Jan 6, 2020 at 16:42

1 Answer 1

3

In comments you indicated that you had a very old latex release. In LaTeX releases prior to 2015, there were by default only 256 registers of each type available. A float insert needs a register for a box (latex savebox), dimen, skip (latex length) and count. So if you allocate more registers for floats there are correspondingly fewer of these registers for other things and packages that use a lot (pstricks for example) may fail. However there is no harm in trying as if it avoids the error there are no ill effects.

In any current release there are over 32 thousand registers of each type available and so this is less of an issue. More floats are allocated by default and there is an \extrafloats command so that you can do \extrafloats{100} to give yourself 100 more if needed, no package is required, although the morefloats package is still available for compatibility reasons and uses \extrafloats internally if it is available.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .