Download

Try now.

Both are problemartic, but the second (teste2.zip) has some problem with an loop instruction, and also contains several push to wrong data pr code (I mean, the pointers are not being dislayed)...they are pudh 24564, when they should be push Code24564, or push Data24564

I´m using boith files, because i´m tryiong to build a pe header viewer that will display all the infromatinos about the header.

Ag night i send to you the file, to you take a look. (Or i i´ll post the probelm on he board. (I guess, i already did something similar);

New testing from 28/04/04

Timings.zip

Hi René, I posted thios on the email I sent to you last night. here goes the final test i did.

I did n´t post there because Randal may use it as a reason to attack.

I did with this type of file to check the behaviour i have at my office. The results wer the same.

RosAsm compile it in 800 ms, while msm take about 500.

The main problem is the parsing time..It is taking much time to be perfoemed. For example, when i use n version with macros, the diferences between the parsing time, from an app that was disassembled and another that is not, may reach 10x in some cases.

Also tehre is a variance on the timings...Assemble the file NoSpaces.exe, then...keep pressing Compile button to you see the results.

If you press 10 times....the 10 will be different (The difference is not that big....but it seems to be more significative in the parsing time) If the difference is always (or mostly) at the parsing time, that means that RosAsm is perfoming different checks to compile the same file every time you press the compile button..

Since, it is the same file, same sources, etc...nothing is changed..there should not be any reason for the parsing time be different, and RosAsm perform different checks for this.

The compile time per se...is always good....but the parsing time is the most that is variying and taking so long to execute.

Best Regards,

Guga

Hi René here goes the file Grid.exe…It also contains the source.

Download