Hello, In Pharo 3.0 (and with Pharo 2.0), I noted an important slowness when browsing importantly populated class. For example, clicking on a method category of the Morph class, it takes about 3 seconds to reveals its methods. With Pharo 1.4 it is immediate with the same Morph class. My computer is a slow dog but it is not a good reason for this slowness. Hilaire -- Dr. Geo - http://drgeo.eu iStoa - http://istoa.drgeo.eu
Hi Hilaire, maybe profiling could show something. Nautilus could benefit from some love in making it faster :) I know that my browser, AltBrowser, is a lot faster on a slow machine (x7 faster on startup time compared to Nautilus, x3 faster than the old system browser). So it could be done. (and Spur could give us a x2 boost on top of all those times :) :) ) Thierry Le 25/10/2014 11:55, Hilaire a écrit :
Hello,
In Pharo 3.0 (and with Pharo 2.0), I noted an important slowness when browsing importantly populated class.
For example, clicking on a method category of the Morph class, it takes about 3 seconds to reveals its methods. With Pharo 1.4 it is immediate with the same Morph class.
My computer is a slow dog but it is not a good reason for this slowness.
Hilaire
On 25 Oct 2014, at 12:06, Thierry Goubier <thierry.goubier@gmail.com> wrote:
Hi Hilaire,
maybe profiling could show something. Nautilus could benefit from some love in making it faster :)
Yes, help is always welcome. There is a lot of caching going on, things should work faster after the first time.
I know that my browser, AltBrowser, is a lot faster on a slow machine (x7 faster on startup time compared to Nautilus, x3 faster than the old system browser). So it could be done.
(and Spur could give us a x2 boost on top of all those times :) :) )
Thierry
Le 25/10/2014 11:55, Hilaire a écrit :
Hello,
In Pharo 3.0 (and with Pharo 2.0), I noted an important slowness when browsing importantly populated class.
For example, clicking on a method category of the Morph class, it takes about 3 seconds to reveals its methods. With Pharo 1.4 it is immediate with the same Morph class.
My computer is a slow dog but it is not a good reason for this slowness.
Hilaire
Le 25/10/2014 12:15, Sven Van Caekenberghe a écrit :
Hi Hilaire,
maybe profiling could show something. Nautilus could benefit from some love in making it faster :) Yes, help is always welcome.
There is a lot of caching going on, things should work faster after the first time.
It is rather a flat response time: in that part, it does not get faster the more I browse and rebrowse on the same method category. Hilaire -- Dr. Geo - http://drgeo.eu iStoa - http://istoa.drgeo.eu
Maybe this one: 12535 <https://pharo.fogbugz.com/default.asp?12535> To many updates on the methods list 2014-10-25 12:06 GMT+02:00 Thierry Goubier <thierry.goubier@gmail.com>:
Hi Hilaire,
maybe profiling could show something. Nautilus could benefit from some love in making it faster :)
I know that my browser, AltBrowser, is a lot faster on a slow machine (x7 faster on startup time compared to Nautilus, x3 faster than the old system browser). So it could be done.
(and Spur could give us a x2 boost on top of all those times :) :) )
Thierry
Le 25/10/2014 11:55, Hilaire a écrit :
Hello,
In Pharo 3.0 (and with Pharo 2.0), I noted an important slowness when browsing importantly populated class.
For example, clicking on a method category of the Morph class, it takes about 3 seconds to reveals its methods. With Pharo 1.4 it is immediate with the same Morph class.
My computer is a slow dog but it is not a good reason for this slowness.
Hilaire
Le 25/10/2014 12:39, Nicolai Hess a écrit :
Maybe this one:
12535 <https://pharo.fogbugz.com/default.asp?12535> To many updates on the methods list
You certainly seems to be onto something there. 12535? 12931 as well? Looks like some have computers that are too fast to notice the slowdown ;) Thierry
participants (4)
-
Hilaire -
Nicolai Hess -
Sven Van Caekenberghe -
Thierry Goubier