Re: [Texmacs-dev] S7 vs. Guile

[TeXmacs]

Dear Max,

Thanks for the news; that looks quite promising indeed.

It is curious that compiling the manual is slower,
because I don't think that this involves many Scheme invocations...

If start-up time gets cut by four, then that is indeed very interesting,
and a good reason to switch to S7.

Best wishes, --Joris


On Mon, Jan 04, 2021 at 07:47:36PM +0100, Massimiliano Gubinelli wrote:
> Dear all,
> 
>  I've managed to find few subtle bugs and now TeXmacs/S7 seems to run fine 
> (there is still a bug I'm aware of but you can give a look if you want). 
> Anyway, I've still to reorganise the changes and test it more. You can find 
> it here:
> 
> https://github.com/mgubi/texmacs/tree/s7/src 
> <https://github.com/mgubi/texmacs/tree/s7/src>
> 
> in the README.md file you find also the results of running some r7rs 
> benchmarks against S7 and Guile 1.8, Guile 3.0.4. Based on them I still 
> expect to be able to have better performance of TeXmacs/S7 vs. 
> TeXmacs/Guile1.8.
> 
> Currenlty the startup time for TeXmacs is 4x faster, but compiling the full 
> user manual (~120 pages) takes 20 sec for S7 and 15 sec for Guile 1.8. So 
> there is still some performance problems. 
> 
> This is still experimental, not ready for everyday use in any way, but looks 
> promising.
> 
> Best,
> Max
> 
> 
> 
> > On 4. Jan 2021, at 11:29, Massimiliano Gubinelli <m.gubinelli@gmail.com> 
> > wrote:
> > 
> > Dear all,
> > I've been making some progress on testing S7 with TeXmacs. Now performance 
> > is good and most editing works ok but I still have some bugs which I have 
> > to correct. Boot time is ~4x faster. As soon as everything works correctly 
> > I plan to run some more benchmarks to see how much we can gain. 
> > 
> > In the meantime I finished to run standard Scheme benchmarks for Guile 
> > 1.8.8, Guile 3.0.4 and S7 on my machine. The one which is slower is Guile 
> > 1.8.8 and sometimes the benchmarks kills the process because it is taking 
> > too much time (CRASHED). These are the results (in seconds). 
> > 
> > Best
> > Max
> > 
> > 
> >     test    S7      G1.8.8  G3.0.4
> > 
> >     browse  24.27   80.32   12.060597               
> >     deriv   25.194  61.39   18.581995               
> >     destruc 52.077  CRASHED 7.143701                
> >     diviter 9.685   77.85   15.453743               
> >     divrec  11.803  78.55   17.41294                
> >     puzzle  27.716  191.35  18.086531               
> >     triangl 33.931  98.16   8.519252                
> >     tak     12.925  134.2   4.757643                
> >     takl    20.968  CRASHED 9.456034                
> >     ntakl   17.073  CRASHED 9.516082                
> >     cpstak  103.358 221.03  59.444873               
> >     ctak    44.139  CRASHED CRASHED         
> >     fib     10.218  195.78  12.090909               
> >     fibc    25.799  CRASHED CRASHED         
> >     fibfp   1.885   45.98   22.001634               
> >     sum     6.637   281.63  6.866215                
> >     sumfp   2.499   105.1   42.058511               
> >     fft     32.198  CRASHED 7.685201                
> >     mbrot   24.403  CRASHED 50.086067               
> >     mbrotZ  18.556  CRASHED 67.011491               
> >     nucleic 19.946  67.46   15.347245               
> >     pi      NO      CRASHED 0.564552                
> >     pnpoly  17.981  CRASHED 24.886723               
> >     ray     20.455  CRASHED 18.51229                
> >     simplex 46.344  CRASHED 13.895531               
> >     ack     10.572  CRASHED 8.413945                
> >     array1  11.483  160.88  9.241778                
> >     string  1.714   1.82    1.872806                
> >     sum1    0.47    1.63    4.427402                
> >     cat     1.187   CRASHED 28.396944               
> >     tail    1.188   CRASHED 9.821691                
> >     wc      8.266   57.91   16.963138               
> >     read1   406     0.95    5.804979                
> >     compil/ 41.155  CRASHED 5.149011                
> >     conform 51.031  CRASHED 10.508732               
> >     dynamic 22.736  69.58   7.374259                
> >     earley  CRASHED CRASHED 9.489885                
> >     graphs  127.611 CRASHED 23.026826               
> >     lattice 139.275 292.7   15.937364               
> >     matrix  72.073  CRASHED 9.881781                
> >     maze    23.258  CRASHED 4.70391         
> >     mazefun 19.51   129.61  9.664338                
> >     nqueens 55.11   CRASHED 19.372148               
> >     paraff/ 31.424  CRASHED 4.24542         
> >     parsing 39.443  CRASHED 10.687959               
> >     peval   29.677  98.91   15.644764               
> >     primes  7.73    39.33   7.521318                
> >     quicks/ 93.996  CRASHED 13.252736               
> >     scheme  71.462  CRASHED 15.142413               
> >     slatex  32.069  48.96   45.047143               
> >     chudno/ NO      CRASHED 0.306648                
> >     nboyer  39.274  151.42  5.10214         
> >     sboyer  31.537  168.81  4.755798                
> >     gcbench 20.54   CRASHED 3.511493                
> >     mperm   173.33  CRASHED 10.650118               
> >     equal   781     CRASHED CRASHED         
> >     bv2str/ 10.782  CRASHED 4.489627        
> 

> _______________________________________________
> Texmacs-dev mailing list
> Texmacs-dev@gnu.org
> https://lists.gnu.org/mailman/listinfo/texmacs-dev