Work at SourceForge, help us to make it a better place! We have an immediate need for a Support Technician in our San Francisco or Denver office.

Close

#2 3.72 beta startup/shutdown times are longer

Beta 241
open
nobody
None
3
2008-11-06
2008-11-06
Arantor
No

JohnB (Atlanta) added:
Emerald Editor 3.72 beta takes significantly longer to start up and to shut down than Crimson Editor 3.70.
This impairs usability only slightly, but it IS noticable enough to be annoying.

DATA COLLECTION

10 trials each with stopwatch.

Both editor versions configured to:
- NOT create new document on startup
- reload last working files on startup
- NO last working file; the editor will start up with no files.

Editor versions installed in separate directories.

AVG. STARTUP TIMES:
v3.70: 1.16 sec.
v3.72: 5.81 sec.

AVG> SHUTDOWN TIMES:
v3.70: Too fast to measure by human user. Virtually instantaneous.
v3.72: 1.21 sec.

_________________________________________________________________________________

Pvt_Ryan added:
We are aware of the issue.. its to do with disk i/o i believe.

When the user config was rewritten I dont think phil realised that the disk i/o would be so high. I am working (albiet slowly) towards a complete rewrite of the config reading/writing which should fix the issue.

This thread is for reference:
http://forum.emeraldeditor.com/index.php?topic=401.0

Thanks for addig this to the tracker, i thought it was already here, but just checked and its not.. :)

_________________________________________________________________________________

Guest added:
It might be also related to the floppy access bug - at least the startup?
I have my doubts that any config fiddling could be so noticably slow even on almost 3Ghz Opty...

_________________________________________________________________________________

Pvt_Ryan added:
not its nothing to do with the floppy access bug. thats a seperate bit of code.

processor speed has no bearing on this.. its disk I/O..

so if you have a 10k rpm raptor drive you will not notice it as much compared to someone that is using a laptop with a 4.2K rpm HDD.

Basically the code accesses the config file once for every option. and the same for saving options.

_________________________________________________________________________________

Guest added:
Ugh, well, thanks a lot for explaining me this.
Any chance for fix soon?

And still it is much faster on my machine (eg. the slow-down is much less noticable on 3Ghz Opty that on 1Ghz old Athlon). Perhaps the disk IO stuff get completely cached by winblows and then occur only within memory?

_________________________________________________________________________________

Pvt_Ryan added:
basically, when i get a chance to rewrite it, it will

1. Read ini into memory
2. set preferences/settings
3. on close rewrite ini to be sure all settings are caught.
[Delete] [Edit Comment] Visible

_________________________________________________________________________________

Guest added:
Sounds good :)
Fingers crossed! :-D

_________________________________________________________________________________

Phil added:
Should be fixed now.

_________________________________________________________________________________

JSR York added:
Hi

I found this bug because I noticed the start-up time for Crimson Editor 3.71 was much worse when I moved to a new, bigger, faster PC. My old PC 2.4Ghz, 760Mb, XP SP2 would start up in about 2 seconds. On my new dual core 3Ghz PC, XP SP3 it takes about 25 seconds. I see this either if I copy over the Program Files\Crimson Editor tree from the old PC, or if I rebuild CEDT 3.72 using Dev Studio 6 on my new PC.

On my new PC 3.70 starts in under a second.

John Robinson
York, England

Discussion