Alpha Video Training
Results 1 to 20 of 20

Thread: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

  1. #1
    "Certified" Alphaholic kkfin's Avatar
    Real Name
    Kenneth
    Join Date
    Dec 2006
    Location
    EU
    Posts
    1,520

    Default Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    I did little test of the performance of Alpha Anywhere interpreter compared to PHP and Adobe ColdFusion 2018.

    The code I did use was very simple.

    Alpha
    Code:
    dim i as N
    i = 0
    while (i < 100000000)
    i= i + 1
    end while
    showvar(i)
    PHP(5.6)
    Code:
    $i = 0;
    while($i < 100000000) {
        $i++;
    }
    echo $i;
    Adobe ColdFusion 2018
    Code:
    <cfset i = 0> 
    <cfloop condition = "i LESS THAN 100000000"> 
    <cfset i = i + 1> 
    </cfloop>
    <cfoutput>#i#</cfoutput>

    It may be good to note that php 5.6 which I did use is much slower that php 7.


    The result in same machine was

    PHP 9 sec
    ColdFusion 2018 25 sec
    Alpha Anywhere unfortunately did not manage test at all.

    Maybe you have better success with AA.

    Ken

  2. #2
    "Certified" Alphaholic mikeallenbrown's Avatar
    Real Name
    Mike Brown
    Join Date
    Nov 2009
    Location
    United States
    Posts
    1,749

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    I don't see a question in this post.

    *shrug*
    Mike Brown - Contact Me
    Programmatic Technologies, LLC
    Programmatic-Technologies.com
    Independent Developer & Consultant

  3. #3
    "Certified" Alphaholic
    Real Name
    David Kates
    Join Date
    Apr 2008
    Location
    Unionville, ON
    Posts
    7,561

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    I think it's a time comparison of running a loop 100 million times among disparate systems. Although I'm never really sure.

    I do think it's really good to know that php 7 is faster than php 5.6. I would be worried if a version 2 years older was faster.

  4. #4
    "Certified" Alphaholic mikeallenbrown's Avatar
    Real Name
    Mike Brown
    Join Date
    Nov 2009
    Location
    United States
    Posts
    1,749

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    Yeah, I understand the time comparison. Just wasn't sure what to do make of it. How does this help Alpha developers? What does this mean for me? I guess what I'm asking is: what's the point here? I don't mean to sound like a jerk, because that's not my intention. I just don't get the post.

    EDIT: The alpha loop does finish after 6-7 minutes ...lol
    Mike Brown - Contact Me
    Programmatic Technologies, LLC
    Programmatic-Technologies.com
    Independent Developer & Consultant

  5. #5
    "Certified" Alphaholic
    Real Name
    David Kates
    Join Date
    Apr 2008
    Location
    Unionville, ON
    Posts
    7,561

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    How did I know you were going to run that Alpha loop... too funny. I ran it as well. Not sure how it did not manage in the first post... but mine ran in a lightning 4 minutes 17 seconds.

  6. #6
    "Certified" Alphaholic iRadiate's Avatar
    Real Name
    Stephen
    Join Date
    Jan 2011
    Location
    London, ON
    Posts
    1,568

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    I got 9 minutes and 4 seconds
    Alpha Anywhere v12.4.6.0.2 Build 5943-5324 IIS v8.5 on Windows Server 2012 in Hyper-V

  7. #7
    "Certified" Alphaholic
    Real Name
    David Kates
    Join Date
    Apr 2008
    Location
    Unionville, ON
    Posts
    7,561

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    Well... it's not always about how fast you get there...

  8. #8
    "Certified" Alphaholic kkfin's Avatar
    Real Name
    Kenneth
    Join Date
    Dec 2006
    Location
    EU
    Posts
    1,520

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    Obviously you do not get speed with Alpha you get time:)

    Alphas interpreter has clearly a big performance problem.
    Unfortunately this lack of performance gets worse/has more impact in multi user environment like mobile or web.

  9. #9
    "Certified" Alphaholic
    Real Name
    Gregg Schmidt
    Join Date
    Mar 2001
    Location
    Milwaukee
    Posts
    1,269

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    86 ms using javascript, a bit longer using xbasic.
    2019-02-15_2-01-34.png

  10. #10
    "Certified" Alphaholic kkfin's Avatar
    Real Name
    Kenneth
    Join Date
    Dec 2006
    Location
    EU
    Posts
    1,520

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    I did also same test using Adobe ColdFusion 2016. It was about 27% slower than 2018 version and confirms so Adobes claims about better performance in version 2018.

    I tested Asp.Net Web Forms(4.6.1 C#) and it also did basically fail like Alpha. It did somehow surprise me but confirms why the traffic is from Asp.Net Web Forms to other platforms. On the other hand MS has new platform Asp.Net Core which I did not test but seems to be the fastest framework available currently.

    I also did test Rad Studio 10.3 Delphi and the result was immediate ( under 1 sec).


    Ken

  11. #11
    Member
    Real Name
    mumfie
    Join Date
    Dec 2008
    Location
    UK
    Posts
    197

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    Quote Originally Posted by kkfin View Post
    I tested Asp.Net Web Forms(4.6.1 C#) and it also did basically fail like Alpha. It did somehow surprise me but confirms why the traffic is from Asp.Net Web Forms to other platforms. On the other hand MS has new platform Asp.Net Core which I did not test but seems to be the fastest framework available currently.

    Ken
    FYI
    I tested Net core 2.1 with minimal console app which takes approx 0.247 seconds using following code

    Code:
    static void Main(string[] args)
            {
                Console.WriteLine("NET Core 2.1");
                Stopwatch stopwatch = new Stopwatch();
                stopwatch.Start();
                int i = 0;
                while (i < 100000000)
                {
                    i = i + 1;
                }
                stopwatch.Stop();
                Console.WriteLine("i: {0}", i);
                Console.WriteLine("Time elapsed seconds: {0}", stopwatch.Elapsed.TotalSeconds);
                Console.ReadKey();
    
            }

  12. #12
    Member
    Real Name
    Richard James
    Join Date
    Jun 2004
    Location
    Atlanta, Ga.
    Posts
    224

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    I've been using Alpha since 2005, when I did comparisons between VB, Access and Filemaker. I didn't choose Alpha because it could finish a loop in the shortest time. I chose Alpha because it could deliver a finished application in the shortest time.

  13. #13
    Member
    Real Name
    Rich Fulham
    Join Date
    May 2011
    Location
    Portland, OR
    Posts
    999

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    Quote Originally Posted by kkfin View Post
    … Alphas interpreter has clearly a big performance problem. …
    But in all fairness, there are at least two other things to consider here: 1) Is there a speed different between running in the Interactive window Control in Alpha vs. running under the actual Alpha Web server, and 2) how does the CPU utilization compare between the various methods? It could be that Alpha is yielding a lot more CPU time to other processes where as PHP and Cold Fusion might be hogging more CPU time and making other things sluggish.

  14. #14
    "Certified" Alphaholic kkfin's Avatar
    Real Name
    Kenneth
    Join Date
    Dec 2006
    Location
    EU
    Posts
    1,520

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    The loop test tests performance. It is a standard test.

  15. #15
    "Certified" Alphaholic
    Real Name
    David Kates
    Join Date
    Apr 2008
    Location
    Unionville, ON
    Posts
    7,561

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    Is it not a standard test. It is an isolated test which has nothing to do with performance of software or a web server or a cpu. Where do you get this stuff? Yikes.
    Last edited by Davidk; 02-16-2019 at 01:35 PM.

  16. #16
    "Certified" Alphaholic kkfin's Avatar
    Real Name
    Kenneth
    Join Date
    Dec 2006
    Location
    EU
    Posts
    1,520

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    I did have time to test also other Basic language interpreters I have. This is interesting:)

    BBC Basic for Windows 6.12a (repeat-until) 39 sec!!! Bravo
    Purebasic 5.70 LTS(x64) 12 sec
    Spiderbasic 2.2.1 under 1 sec

    All tested in same machine as before others.

    Who say that Basic can not be fast.

    Ken

  17. #17
    "Certified" Alphaholic kkfin's Avatar
    Real Name
    Kenneth
    Join Date
    Dec 2006
    Location
    EU
    Posts
    1,520

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    I tested this code in Alpha Application Server in test.a5w page

    Code:
    <!--A5Language:Javascript-->
    <%a5 
    i = 0;
     while (i < 100000000){
    	i++;
    	}
    print(i);
    %>
    The result was immediate under 1 sec.

    Big big difference compared to xbasic.

    Latest roadmap did has a mention (again) of server side javascript in AA. Lets hope it is an alternative choice one day.

  18. #18
    Member
    Real Name
    Rich Fulham
    Join Date
    May 2011
    Location
    Portland, OR
    Posts
    999

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    Ken,
    So, how was the first timing taken? Was it from the Control Panel Interactive window?

    Was first timing from one of these places:
    1) Server-Side Grid Embed (like OnGridInitialize)
    2) Server-Side Grid callback function
    3) Server-Side UX Embed (like OnDialogInitialize)
    4) Server-Side UX callback function

    If your timings were taken from one of the above then you have a good case to take back to Alpha and ask, "What is going on with the WAS? The time taken to run a tight loop in xbasic on an a5w page vs from inside a grid or UX should be vastly closer. Is there a problem in how it is yielding processor time?"

    If your initial timings were taken from the Control Panel Interactive window and the above four places can run the loop in about one second then I think this case can be dismissed.

    However, don't give up in finding ways to make AA run faster. I've been trying to get a couple of components to run faster under AA and so far I have had pretty good luck removing conditional expressions from server-side and client-side expressions and reducing the amount of data / controls shown in Grids and lists.

  19. #19
    "Certified" Alphaholic kkfin's Avatar
    Real Name
    Kenneth
    Join Date
    Dec 2006
    Location
    EU
    Posts
    1,520

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    Quote Originally Posted by RichCPT View Post
    Ken,
    So, how was the first timing taken? Was it from the Control Panel Interactive window?

    Was first timing from one of these places:
    1) Server-Side Grid Embed (like OnGridInitialize)
    2) Server-Side Grid callback function
    3) Server-Side UX Embed (like OnDialogInitialize)
    4) Server-Side UX callback function

    If your timings were taken from one of the above then you have a good case to take back to Alpha and ask, "What is going on with the WAS? The time taken to run a tight loop in xbasic on an a5w page vs from inside a grid or UX should be vastly closer. Is there a problem in how it is yielding processor time?"

    If your initial timings were taken from the Control Panel Interactive window and the above four places can run the loop in about one second then I think this case can be dismissed.

    However, don't give up in finding ways to make AA run faster. I've been trying to get a couple of components to run faster under AA and so far I have had pretty good luck removing conditional expressions from server-side and client-side expressions and reducing the amount of data / controls shown in Grids and lists.

    I do not think performance in a top value in Alpha and this is also true with xbasic. It is in my opinion a collection of wrapper functions currently. I think the point is in Richard post #12.

    There is many ways to make performance better. And some that you can not change.

    Ken

  20. #20
    Member jgrannis's Avatar
    Real Name
    Jeff Grannis
    Join Date
    Apr 2016
    Location
    New Iberia, LA
    Posts
    339

    Default Re: Alpha performance compared PHP and Adobe Coldfusion 2018 interpreter

    Quote Originally Posted by kkfin View Post
    Latest roadmap did has a mention (again) of server side javascript in AA. Lets hope it is an alternative choice one day.
    I believe that Alpha's approach to server-side javascript is already to hand it over to Node.js via calling a Node service from Xbasic.
    So if something gets too hard to do with Xbasic, is for some reason much slower when using Xbasic, or maybe in the case where a nice Node module already exists to get the job done, we can do server-side javascript today.

    I do remember the convenience, in the 90's, of doing a lot of server-side javascript (SSJS) with Netscape's iPlanet web servers. Never thought I would still be doing javascript on the server 20 years later.

Similar Threads

  1. Important Info on the latest version of Alpha Anywhere Nov 2018
    By Richard Rabins in forum Announcements
    Replies: 0
    Last Post: 12-13-2018, 11:55 PM
  2. Replies: 0
    Last Post: 10-02-2013, 12:31 AM
  3. Alpha Five compared to Access
    By Richard Rabins in forum Alpha Five Version 5
    Replies: 8
    Last Post: 10-23-2002, 10:39 AM

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •