Sage 10.9.beta3 very slow to start

[Eric Gourgoulhon]

Hi,

I've installed Sage 10.9.beta3 from a fresh git clone on Ubuntu 24.04 via

make configure

./configure --enable-download-from-upstream-url

MAKE="make -j8" make

Each time I run sage, it takes a long time (more than 1 minute) to start, either in console mode or in Jupyterlab, whereas for all previous versions of Sage, including 10.9.beta1, it took only a few seconds on the same computer. 

During this long start up time, sage is actually running makeinfo, as revealed by the "top" command. 

This long time running makeinfo also occurs each time one restarts the Sage kernel in Jupyterlab. 

Am I the only one experiencing this? Shall I open a ticket? 

Eric. 

[Eric Gourgoulhon]

Le samedi 17 janvier 2026 à 15:50:53 UTC+1, Eric Gourgoulhon a écrit :

Each time I run sage, it takes a long time (more than 1 minute) to start, either in console mode or in Jupyterlab, whereas for all previous versions of Sage, including 10.9.beta1, it took only a few seconds on the same computer. 

During this long start up time, sage is actually running makeinfo, as revealed by the "top" command. 

This long time running makeinfo also occurs each time one restarts the Sage kernel in Jupyterlab. 

The issue has been fixed by installing Ubuntu package texinfo (which provides the command makeinfo) and rebuilding Sage. 

It turns out that I had no system texinfo package installed on that computer: configure was saying:

info:                           no suitable system package; standard, SPKG version 7.2 will be installed

I've installed Ubuntu's texinfo 7.1 and now configure is saying:

info:                           using system package; SPKG will not be installed

I've then rebuilt Sage 10.9.beta3 from scratch and the problem is gone: Sage starting time is again of a few seconds. 

Side note: for previous versions of Sage on that computer, there was no need to install the system texinfo. For instance, for Sage 10.8.beta*, the SPKG info 7.0.3 had been installed and the start up time was OK. Maybe this is the sign of some issue with the current SPKG info 7.2 (added to Sage 10.8.beta8  in https://github.com/sagemath/sage/pull/41034) under Ubuntu 24.04...

Eric.