1. task msging

New upgrade to taskmsging.e is at

http://designerthinking.com/temp/taskmsgs.e

 
procedure task_send_var(object toid, atom fromid, object what)  
function task_get_var(atom toid = "*", object fromid = "*")  
function task_var_list(object toid = "*", object fromid = "*") 
 
procedure task_send_que(object toid, atom fromid, object what) 
function task_get_que(atom toid = "*", object fromid = "*")  
function task_que_list(object toid = "*", object fromid = "*") 
 

do i need anything else?

useless

new topic     » topic index » view message » categorize

2. Re: task msging

I think there is an error in the code? You have atom="*".

Further, what is suppose to happen if you simply call task_get_var(), i.e. w/o any parameters? task_var_list, task_get_que and task_que_list also have all parameters defaulted to something, thus you can call any of them w/o any parameters.

Jeremy

new topic     » goto parent     » topic index » view message » categorize

3. Re: task msging

jeremy said...

I think there is an error in the code? You have atom="*".

Picture me cussing at find/replace and then uploading without retesting. But heck, i didn't know anyone cared to look, let alone run the code!

jeremy said...

Further, what is suppose to happen if you simply call task_get_var(), i.e. w/o any parameters? task_var_list, task_get_que and task_que_list also have all parameters defaulted to something, thus you can call any of them w/o any parameters.

Jeremy

If the programmer wants to do that, well, ok. You wold like it more restrictive? I see no reason to not be able to call task_var_list and task_que_list with no parms to get a complete list, for housekeeping duties mostly. Makes it a shorter bit of code to make sure you aren't eating thru memory as you launch/kill a million tasks every few months.

It makes little sense for any task other than task0 to do that, but should i restrict all tasks from using defaults, or all but task0, or remove all the defaults? After all, i can't tell which task called the msg handler unless it tells me truthfully who it is.

useless

new topic     » goto parent     » topic index » view message » categorize

4. Re: task msging

useless said...

After all, i can't tell which task called the msg handler unless it tells me truthfully who it is.

Actually, I think it can. Any time a task calls a routine, that routine is running as that task, and a call to task_self() should tell you what task called that routine. I don't know why i didn't think to say this before.

Try running this test to confirm it. Does this look correct to everyone? This proves that a subroutine doesn't need the task_id passed to it, it can get it with task_self().

Example:

 
include std/text.e 
 
puts(1, "Test to see if task_self() works in subroutines...\n") 
 
procedure print_task_id(atom id) 
    atom testid 
    testid = task_self() 
     
    puts(1, "id=" & sprint(id) & ", testid=" & sprint(testid)& "\n") 
end procedure 
 
procedure task_test() 
	for i = 1 to 10 do 
		print_task_id(task_self()) 
		task_yield() 
	end for 
	puts(1, "Press any key to continue...") 
	while 1 do 
        if get_key() then 
    		exit 
    	end if 
    	task_yield() 
	end while 
	abort(0) 
end procedure 
 
atom t1 = task_create(routine_id("task_test"), {}) 
atom t2 = task_create(routine_id("task_test"), {}) 
atom t3 = task_create(routine_id("task_test"), {}) 
atom t4 = task_create(routine_id("task_test"), {}) 
atom t5 = task_create(routine_id("task_test"), {}) 
atom t6 = task_create(routine_id("task_test"), {}) 
 
task_schedule(t1, {1, 1}) 
task_schedule(t2, {1, 1}) 
task_schedule(t3, {1, 1}) 
task_schedule(t4, {1, 1}) 
task_schedule(t5, {1, 1}) 
task_schedule(t6, {1, 1}) 
 
while 1 do 
    task_yield() 
end while 
 
new topic     » goto parent     » topic index » view message » categorize

5. Re: task msging

ryanj said...
useless said...

After all, i can't tell which task called the msg handler unless it tells me truthfully who it is.

Actually, I think it can. Any time a task calls a routine, that routine is running as that task, and a call to task_self() should tell you what task called that routine. I don't know why i didn't think to say this before.

<snip>

Yes, it does work that way. Is this preferable to passing the task_id one says one is? If so, then task0 cannot do cleanup of msgs sent to a task_id that no longer exists. Or i need a task_msg_cleanup routine?

useless

new topic     » goto parent     » topic index » view message » categorize

6. Re: task msging

Ok, task msg handler no longer accepts fromid in functions or procedures, it identifies who the task is, and tasks cannot spoof being another task.

To remove msgs to tasks, i wrote

-- This is to remove any vars and queues sent to dead tasks. 
-- "*" will clear all terminated task's vars, unscheduled msgs will remain. 
-- Specifying the task_id will remove those/that's task_id's vars, running or not. 
--  _clean("*") , _clean(1) , _clean({1,5}) 
-- This does not remove msgs sent BY the deceased task, only msgs sent TO the deceased task. 
export procedure task_msg_clean(object who = "*") 

Those interested, refresh:

http://designerthinking.com/temp/taskmsgs.e

useless

new topic     » goto parent     » topic index » view message » categorize

7. Re: task msging

"Que" is a pronoun in French or Spanish.

"Queue" is a line or waiting list.

"Cue" is a signal or sign for somebody to do something, or a straight tapering rod used to impel the balls in various games (e.g. billiards).

-Greg

new topic     » goto parent     » topic index » view message » categorize

8. Re: task msging

Umm, I think she's just abbreviating. At least, that was kind of apparent to me.

new topic     » goto parent     » topic index » view message » categorize

9. Re: task msging

jaygade said...

Umm, I think she's just abbreviating. At least, that was kind of apparent to me.

That may be fine for demonstrative purposes, but I hope the production code uses the proper word. Right now I read that as "task_send_what", "task_get_what", and "task_what_list", which makes little sense to me.

-Greg

new topic     » goto parent     » topic index » view message » categorize

10. Re: task msging

ghaberek said...
jaygade said...

Umm, I think she's just abbreviating. At least, that was kind of apparent to me.

That may be fine for demonstrative purposes, but I hope the production code uses the proper word. Right now I read that as "task_send_what", "task_get_what", and "task_what_list", which makes little sense to me.

LOL, glad I wasn't the only one, though I parsed 'que' as 'that' (no accent mark, right?), which makes a little more sense. I agree that the full word should be spelled out.

Matt

new topic     » goto parent     » topic index » view message » categorize

11. Re: task msging

mattlewis said...

LOL, glad I wasn't the only one, though I parsed 'que' as 'that' (no accent mark, right?), which makes a little more sense. I agree that the full word should be spelled out.

I am no linguist and my years of French and Spanish are gradually slipping. But I believe the accent applies when asking a question versus making a statement.

-Greg

new topic     » goto parent     » topic index » view message » categorize

12. Re: task msging

Ok, then, i updated

http://designerthinking.com/temp/taskmsgs.e

to reflect the most important part of this, so everyone in every language can understand, and Matt might stop laughing at me.

useless

new topic     » goto parent     » topic index » view message » categorize

13. Re: task msging

useless said...

Matt might stop laughing at me.

Geez...give it a rest already.

Matt

new topic     » goto parent     » topic index » view message » categorize

14. Re: task msging

mattlewis said...
useless said...

Matt might stop laughing at me.

Geez...give it a rest already.

Matt

Gosh, i'm sorry, but i used "queue" properly, and abbreviated it to "que" for shorter function and variable names, and that's what you have to post here about? And you get upset when i don't like your

mattlewis said...

LOL, glad I wasn't the only one, though I parsed 'que' as 'that' (no accent mark, right?), which makes a little more sense. I agree that the full word should be spelled out.

Matt

When will YOU give it a rest? I thought i was contributing to Euphoria with needed code, when you proove yet again i am only a joke and a side show. You gave zero constructive feedback. If i did spell out all the words, and switched back and forth from "queue" (the msg queue) to "cue" (the individual msg, which i think is wrong), you'd LOL at that too!

useless

new topic     » goto parent     » topic index » view message » categorize

15. Re: task msging

useless said...
mattlewis said...
useless said...

Matt might stop laughing at me.

Geez...give it a rest already.

Gosh, i'm sorry, but i used "queue" properly, and abbreviated it to "que" for shorter function and variable names, and that's what you have to post here about? And you get upset when i don't like your

Ok, please let me know the topics upon which I may post. I'm sure if you look hard enough, you'll see that I do not stalk you, and that I am not at all inactive on this site.

useless said...
mattlewis said...

LOL, glad I wasn't the only one, though I parsed 'que' as 'that' (no accent mark, right?), which makes a little more sense. I agree that the full word should be spelled out.

When will YOU give it a rest? I thought i was contributing to Euphoria with needed code, when you proove yet again i am only a joke and a side show. You gave zero constructive feedback. If i did spell out all the words, and switched back and forth from "queue" (the msg queue) to "cue" (the individual msg, which i think is wrong), you'd LOL at that too!

What are you talking about? I was replying to Greg's post, and the 'LOL' was directed at him. It was a slightly amusing multi-language portmanteauI'll put LOL in lowercase next time to emphasize the slightness of the amusement. The only one saying you're a joke and a side show is you.

I'd agree that 'cue' would be wrong, since the whole point of the code was to create a queue. My constructive feedback was that I thought that the word queue should be spelled out.

Matt

new topic     » goto parent     » topic index » view message » categorize

16. Re: task msging

Nice work Kat.

new topic     » goto parent     » topic index » view message » categorize

17. Re: task msging

useless said...
mattlewis said...
useless said...

Matt might stop laughing at me.

Geez...give it a rest already.

Gosh, i'm sorry, but i used "queue" properly, and abbreviated it to "que" for shorter function and variable names, and that's what you have to post here about? And you get upset when i don't like your

mattlewis said...

LOL, glad I wasn't the only one, though I parsed 'que' as 'that' (no accent mark, right?), which makes a little more sense. I agree that the full word should be spelled out.

When will YOU give it a rest? I thought i was contributing to Euphoria with needed code, when you proove yet again i am only a joke and a side show. You gave zero constructive feedback. If i did spell out all the words, and switched back and forth from "queue" (the msg queue) to "cue" (the individual msg, which i think is wrong), you'd LOL at that too!

Let's give this all a rest already. Neither Matt nor I meant any disrespect to you, Kat. We were merely discussing proper usage and definition of the words "que" and "queue". We are a community of developers, and when someone posts code that is incorrect, we correct if for them. And the fact of the matter is, the word that you used was not correct.

I'm sure we all appreciate your contributions very much. I know I do. But you can't keep starting drama and reverting into this self-defamatory attitude every time someone says something about something you do. No one's trying to attack you or bring you down, you just seem to take people's criticisms way too personally.

-Greg

new topic     » goto parent     » topic index » view message » categorize

18. Re: task msging

ghaberek said...

Let's give this all a rest already. Neither Matt nor I meant any disrespect to you, Kat. We were merely discussing proper usage and definition of the words "que" and "queue". We are a community of developers, and when someone posts code that is incorrect, we correct if for them.

"we correct if for them"? I am merely questioning your use of the words, please do not take this personally, even though i am quoting you and replying to your post.

ghaberek said...

And the fact of the matter is, the word that you used was not correct.

I abbreviated "queue" to "que". I abbreviate "euphoria" to "eu". I know "que" is a spanish word, i took several years of spanish in school, and it's an important enough language that i wrote code in Tiggr to do translations. At least getting a task msg is a lot like asking the task msg handler "what did taskx send me?". Never the less, i was not writing in spanish. When i write in php or mirc, i do not adhere to eu syntax, keywords, or style.

I object to kinz's icon suggestion because it is "eu" only in russian.

ghaberek said...

I'm sure we all appreciate your contributions very much. I know I do. But you can't keep starting drama and reverting into this self-defamatory attitude every time someone says something about something you do. No one's trying to attack you or bring you down, you just seem to take people's criticisms way too personally.

-Greg

Yeas, i hear the same thing when the neighbor commanded his dogs to attack me on my property, day after day for years. It wasn't personal. And my threshold for Matt is zero after his behavior on the devlist, so i have little incentive to explain my use of words when he cannot understand the word "to".

useless

new topic     » goto parent     » topic index » view message » categorize

19. Re: task msging

useless said...
ghaberek said...

I'm sure we all appreciate your contributions very much. I know I do. But you can't keep starting drama and reverting into this self-defamatory attitude every time someone says something about something you do. No one's trying to attack you or bring you down, you just seem to take people's criticisms way too personally.

Yeas, i hear the same thing when the neighbor commanded his dogs to attack me on my property, day after day for years. It wasn't personal. And my threshold for Matt is zero after his behavior on the devlist, so i have little incentive to explain my use of words when he cannot understand the word "to".

Again, what are you talking about? I wasn't questioning the meaning of the word 'to'. I was saying you were using it incorrectly.

No one here is attacking you, with dogs or words or anything else. I second Greg's sentiment about your contributions, and, unfortunately, about your reactions seemingly any time anyone engages you about something you've coded.

Matt

new topic     » goto parent     » topic index » view message » categorize

20. Re: task msging

useless said...

"we correct if for them"? I am merely questioning your use of the words, please do not take this personally, even though i am quoting you and replying to your post.

Yes, and I meant exactly that. You will see examples all over this forum of a person posting code, then someone else correcting it, often saying "you did something wrong here, this is how it should be done". It's helpful; I see nothing wrong with it. By entering into a public forum, one should understand that content can and will be criticized and/or corrected by other members.

useless said...

I abbreviated "queue" to "que". I abbreviate "euphoria" to "eu". I know "que" is a spanish word, i took several years of spanish in school, and it's an important enough language that i wrote code in Tiggr to do translations. At least getting a task msg is a lot like asking the task msg handler "what did taskx send me?". Never the less, i was not writing in spanish. When i write in php or mirc, i do not adhere to eu syntax, keywords, or style.

I think the general consensus is that "que" is not an abbreviation of "queue"; though I suppose that's debatable. I believe "eu" on the other hand, is inherently understood to be an abbreviation of "euphoria". I did not know you abbreviated it on purpose, I simply thought you made a mistake, and I was just trying to help.

useless said...

Yeas, i hear the same thing when the neighbor commanded his dogs to attack me on my property, day after day for years. It wasn't personal. And my threshold for Matt is zero after his behavior on the devlist, so i have little incentive to explain my use of words when he cannot understand the word "to".

What happened to you is a downright shame. And it's too bad you and Matt can't get along; I could only imaging what the two of you could come up with if you ever collaborated on something.

Regardless, we need to start keeping this personal banter to a minimum. Lately threads have been running longer than required just because one person didn't like what the other had to say. Perhaps we need some sort of PM system, or personal talk pages...

-Greg

new topic     » goto parent     » topic index » view message » categorize

21. Re: task msging

mattlewis said...
useless said...

<snip> And my threshold for Matt is zero after his behavior on the devlist, so i have little incentive to explain my use of words when he cannot understand the word "to".

Again, what are you talking about? I wasn't questioning the meaning of the word 'to'. I was saying you were using it incorrectly.

Wow, here we go again! PLEASE note i said "on the devlist". You did not understand the difference between "yield the task" and "yield to the task", even after repeated posts. So AGAIN, i have little reason to attempt to explain "abbreviation" to you.

useless

new topic     » goto parent     » topic index » view message » categorize

22. Re: task msging

Kat said...

PLEASE note i said "on the devlist". You did not understand the difference between "yield the task" and "yield to the task", even after repeated posts.

By the way, I understood the words that you used but I did not understand the meaning you give the phrases. However I do now. After some discussion it turns out that ...

"yield the task" = The subtask calls task_yield()
"yield to the task" = The parent calls task_yield()


Is that right, Kat?

new topic     » goto parent     » topic index » view message » categorize

23. Re: task msging

DerekParnell said...
Kat said...

PLEASE note i said "on the devlist". You did not understand the difference between "yield the task" and "yield to the task", even after repeated posts.

By the way, I understood the words that you used but I did not understand the meaning you give the phrases. However I do now. After some discussion it turns out that ...

"yield the task" = The subtask calls task_yield()
"yield to the task" = The parent calls task_yield()


Is that right, Kat?

Not exactly. In the context of that fight, it was about the subtask yielding the parent or root task, not itself.

Consider:
I'll throw you. you are displaced
I'll throw TO you. --
i pass something, you recieve it

task1 yields the parent parent is displaced
task1 yields TO the parent --
task1 passes control, parent recieves it

useless,
and bewildered

new topic     » goto parent     » topic index » view message » categorize

24. Re: task msging

useless said...

task1 yields the parent parent is displaced
task1 yields TO the parent --
task1 passes control, parent recieves it

The whole confusion (as I recall) wasn't about TO or not, it was the word parent and somehow that became to mean the parent task. Through everyone being totally confused about what anyone else was talking about the conversation went south in debating little words like TO because everyone had their own little idea of what was really happening behind the scenes. It was in the context of an application loading a DLL. So the DLL would call yield in the context of the main application's scheduler instead of it's own internal scheduler (which doesn't/never did exist), which in turn is the problem of calling task_yield() in the DLL (prior to NOP change)).

Jeremy

new topic     » goto parent     » topic index » view message » categorize

25. Re: task msging

So basically its like I use someone else's pen but when somebody asks me for one I can't have one to lend? Is that what you mean with the task_yeild() thing?

new topic     » goto parent     » topic index » view message » categorize

26. Re: task msging

Since you didn't click "reply with quote", i don't know who you are asking. Are you not understanding how task_yield() works? Or not understanding how the DLL task_yield() was (mis)understood to work/(not)proposed?

useless

new topic     » goto parent     » topic index » view message » categorize

27. Re: task msging

useless said...

Not exactly. In the context of that fight, it was about the subtask yielding the parent or root task, not itself.

Consider:
I'll throw you. you are displaced
I'll throw TO you. --
i pass something, you recieve it

task1 yields the parent parent is displaced
task1 yields TO the parent --
task1 passes control, parent recieves it

useless,
and bewildered

Yes, and you still haven't understood that when you yield, there is no choice in the act of yielding, who will be the next to execute. You simply yield, and the task scheduler decides who goes next. Execution could return to you immediately.

The way that the word "to" is being used here indicates some measure of choice being applied to select the task that will execute next when calling task_yield(). There is no such thing in euphoria multitasking. A task yields, and the scheduler decides, based on the state of all of the existing tasks (how often they want to run, when they last ran, etc) which one should be run.

So, to use the throwing analogy:

I'll throw the ball in the air, whoever is closest to it will pick it up.

From the manual:

Euphoria's task scheduler decides which task should be active at any given time.

Matt

new topic     » goto parent     » topic index » view message » categorize

28. Re: task msging

mattlewis said...
useless said...

Not exactly. In the context of that fight, it was about the subtask yielding the parent or root task, not itself.

Consider:
I'll throw you. you are displaced
I'll throw TO you. --
i pass something, you recieve it

task1 yields the parent parent is displaced
task1 yields TO the parent --
task1 passes control, parent recieves it

useless,
and bewildered

Yes, and you still haven't understood that when you yield, there is no choice in the act of yielding, who will be the next to execute. You simply yield, and the task scheduler decides who goes next. Execution could return to you immediately.

And you still have not learned i do understand that. I have a number of apps setting up tasks, and until one of them hits a blocking OS call, they all run 0,1,2,3,4,0,1,2,3,4,0,1,2,3,4,etc. When running just one task, it has ALWAYS run 0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,etc, do you understand? It's the way i task_scheduled them. Until they cannot complete in alloted time, and don't yield, the sequence of executions aren't mixed up. And yes, for the 400th time, i understand they can get mixed up, and i don't care, but that's how they have been running.

mattlewis said...

<and he prattles on some more, like he did for days on the devlist, telling me it ain't as i describe above.>

And this is what i call abuse, he tells me i have no clue, over and over and over, yet my code runs the way i want it to, and the way i describe it. And he'll tell me yet again, this is the way it is, it doesn't happen like that. He will attempt to keep pounding into my head that my code doesn't run like he says it should, no matter that he has never seen my applications here. No matter what i do, it isn't happening correctly, the Matt way, unless it's not like i want it.

useless

new topic     » goto parent     » topic index » view message » categorize

29. Re: task msging

useless said...
mattlewis said...
useless said...

Not exactly. In the context of that fight, it was about the subtask yielding the parent or root task, not itself.

Consider:
I'll throw you. you are displaced
I'll throw TO you. --
i pass something, you recieve it

task1 yields the parent parent is displaced
task1 yields TO the parent --
task1 passes control, parent recieves it

useless,
and bewildered

Yes, and you still haven't understood that when you yield, there is no choice in the act of yielding, who will be the next to execute. You simply yield, and the task scheduler decides who goes next. Execution could return to you immediately.

And you still have not learned i do understand that. I have a number of apps setting up tasks, and until one of them hits a blocking OS call, they all run 0,1,2,3,4,0,1,2,3,4,0,1,2,3,4,etc. When running just one task, it has ALWAYS run 0,1,0 do you understand? It's the way i task_scheduled them. Until they cannot complete in alloted time, and don't yield, the sequence of executions aren't mixed up. And yes, for the 400th time, i understand they can get mixed up, and i don't care, but that's how they have been running.

Ok, if you understand that, which you have never acknowledged until just now, then why do you insist on describing things incorrectly? Especially on the dev list, where we're discussing how it should work. Precision is important in this sort of discussion.

useless said...
mattlewis said...

<and he prattles on some more, like he did for days on the devlist, telling me it ain't as i describe above.>

And this is what i call abuse, he tells me i have no clue, over and over and over, yet my code runs the way i want it to, and the way i describe it. And he'll tell me yet again, this is the way it is, it doesn't happen like that. He will attempt to keep pounding into my head that my code doesn't run like he says it should, no matter that he has never seen my applications here. No matter what i do, it isn't happening correctly, the Matt way, unless it's not like i want it.

I'm sorry you can't see beyond your own code, especially on the dev list, or when describing euphoria multitasking in general. I'm glad your code runs the way you want it to run. I can't comprehend how this is abuse. It's not about my way, but you're right that it wasn't correct. You've just admitted it above. Thank you for that, by the way.

On the dev list, we get into a lot of discussions that I suppose would seem, like this thread, to be splitting hairs. Yet, it's very important, because seemingly minute differences in understanding can result in drastically different outcomes. It's often not immediately apparent, either, but requires a lot of discussion and reflection to figure it out. And the result of the discussion is going to affect how euphoria changes, so it's important to be precise, and in a community project like this one, we aim for consensus, which often takes a lot of talking, sometimes mistaken for 'abuse', apparently. I'm certainly not trying to abuse anyone here.

Matt

new topic     » goto parent     » topic index » view message » categorize

30. Re: task msging

mattlewis said...

Ok, if you understand that, which you have never acknowledged until just now, then why do you insist on describing things incorrectly? Especially on the dev list, where we're discussing how it should work. Precision is important in this sort of discussion.

What did i describe incorrectly?

mattlewis said...
useless said...

And this is what i call abuse, he tells me i have no clue, over and over and over, yet my code runs the way i want it to, and the way i describe it. And he'll tell me yet again, this is the way it is, it doesn't happen like that. He will attempt to keep pounding into my head that my code doesn't run like he says it should, no matter that he has never seen my applications here. No matter what i do, it isn't happening correctly, the Matt way, unless it's not like i want it.

I'm sorry you can't see beyond your own code, especially on the dev list, or when describing euphoria multitasking in general. I'm glad your code runs the way you want it to run. I can't comprehend how this is abuse. It's not about my way, but you're right that it wasn't correct. You've just admitted it above. Thank you for that, by the way.

Again, you didn't read what i wrote, you baiting sob. I said "No matter what i do, it isn't happening correctly, the Matt way, unless it's not like i want it" meaning you will call anything i do incorrect unless i do it your way, in which case it's not doing what i want. You must be the one who wrote the first news.ex, which i sped up 10 times by doing it my way.

useless

new topic     » goto parent     » topic index » view message » categorize

31. Re: task msging

useless said...
mattlewis said...

Ok, if you understand that, which you have never acknowledged until just now, then why do you insist on describing things incorrectly? Especially on the dev list, where we're discussing how it should work. Precision is important in this sort of discussion.

What did i describe incorrectly?

The way that task yielding works. Go back and re-read what I wrote. Especially the ball analogy. You admitted that your description, while it describes how multitasking works under a specific set of circumstances (which you only specified when you gave the example with the sequence of task numbers executed), does not accurately describe how multitasking works.

useless said...
mattlewis said...
useless said...

And this is what i call abuse, he tells me i have no clue, over and over and over, yet my code runs the way i want it to, and the way i describe it. And he'll tell me yet again, this is the way it is, it doesn't happen like that. He will attempt to keep pounding into my head that my code doesn't run like he says it should, no matter that he has never seen my applications here. No matter what i do, it isn't happening correctly, the Matt way, unless it's not like i want it.

I'm sorry you can't see beyond your own code, especially on the dev list, or when describing euphoria multitasking in general. I'm glad your code runs the way you want it to run. I can't comprehend how this is abuse. It's not about my way, but you're right that it wasn't correct. You've just admitted it above. Thank you for that, by the way.

Again, you didn't read what i wrote, you baiting sob. I said "No matter what i do, it isn't happening correctly, the Matt way, unless it's not like i want it" meaning you will call anything i do incorrect unless i do it your way, in which case it's not doing what i want.

I read every word of what you wrote. I'm not sure why you say this. It has nothing to do with "my" way. Please explain how "my way" is different from the way that euphoria actually behaves. Seriously, please explain why what I said is incorrect. If I'm wrong, I'd like to know it.

useless said...

You must be the one who wrote the first news.ex, which i sped up 10 times by doing it my way.

I certainly didn't write news.ex. I wasn't even aware of it until there was a bug report filed about it crashing on linux (reported a week or two before you started working on it, IIRC), and then you enhanced the http library to be non-blocking. That was good work. However, that doesn't have any real bearing on the issue at hand here, which is how multitasking operates.

Matt

new topic     » goto parent     » topic index » view message » categorize

32. Re: task msging

mattlewis said...
useless said...
mattlewis said...

Ok, if you understand that, which you have never acknowledged until just now, then why do you insist on describing things incorrectly? Especially on the dev list, where we're discussing how it should work. Precision is important in this sort of discussion.

What did i describe incorrectly?

The way that task yielding works. Go back and re-read what I wrote. Especially the ball analogy. You admitted that your description, while it describes how multitasking works under a specific set of circumstances (which you only specified when you gave the example with the sequence of task numbers executed), does not accurately describe how multitasking works.

Well, the specific set of circumstances i optomise to result in greater speed, distributed over the tasks in such a way that it doesn't always work like you say it does. Why are you fighting this?

mattlewis said...
useless said...
mattlewis said...
useless said...

And this is what i call abuse, he tells me i have no clue, over and over and over, yet my code runs the way i want it to, and the way i describe it. And he'll tell me yet again, this is the way it is, it doesn't happen like that. He will attempt to keep pounding into my head that my code doesn't run like he says it should, no matter that he has never seen my applications here. No matter what i do, it isn't happening correctly, the Matt way, unless it's not like i want it.

I'm sorry you can't see beyond your own code, especially on the dev list, or when describing euphoria multitasking in general. I'm glad your code runs the way you want it to run. I can't comprehend how this is abuse. It's not about my way, but you're right that it wasn't correct. You've just admitted it above. Thank you for that, by the way.

Again, you didn't read what i wrote, you baiting sob. I said "No matter what i do, it isn't happening correctly, the Matt way, unless it's not like i want it" meaning you will call anything i do incorrect unless i do it your way, in which case it's not doing what i want.

I read every word of what you wrote. I'm not sure why you say this. It has nothing to do with "my" way. Please explain how "my way" is different from the way that euphoria actually behaves. Seriously, please explain why what I said is incorrect. If I'm wrong, I'd like to know it.

Read up where i gave you the sequence of 0,1,0,1,0,1 etc. Task1 yielded to task0 repeatedly, and that is what i wanted to happen. To you, this is "wrong" because the taskmanager is supposed to do that however it feels like, not how i want it to happen. If the code runs well, then according to you: i've screwed it up, it's not reality, you must deny it.

useless said...

You must be the one who wrote the first news.ex, which i sped up 10 times by doing it my way.

I certainly didn't write news.ex. I wasn't even aware of it until there was a bug report filed about it crashing on linux (reported a week or two before you started working on it, IIRC), and then you enhanced the http library to be non-blocking. That was good work. However, that doesn't have any real bearing on the issue at hand here, which is how multitasking operates.

Matt

[/quote]

I do not think that is the issue at hand. I think the issue is that you cannot leave me alone. You must dispute me at every turn, and find ways to "prove" you are correct in the face of a different reality.

useless

new topic     » goto parent     » topic index » view message » categorize

33. Re: task msging

useless said...
mattlewis said...

The way that task yielding works. Go back and re-read what I wrote. Especially the ball analogy. You admitted that your description, while it describes how multitasking works under a specific set of circumstances (which you only specified when you gave the example with the sequence of task numbers executed), does not accurately describe how multitasking works.

Well, the specific set of circumstances i optomise to result in greater speed, distributed over the tasks in such a way that it doesn't always work like you say it does. Why are you fighting this?

Yes, I agree, that in that circumstance, that's how it behaves. That was not the context of your previous statements. Why is this difficult to understand? We can't read your mind. You were making general statements, while apparently only meaning them to apply to some set of circumstances that only you knew about at the time. Like I said before, when engaging in this sort of discussion, you need to be able to see beyond your code, or any particular example. At a minimum, you need to make sure that the participants are aware of the context.

useless said...
mattlewis said...

I certainly didn't write news.ex. I wasn't even aware of it until there was a bug report filed about it crashing on linux (reported a week or two before you started working on it, IIRC), and then you enhanced the http library to be non-blocking. That was good work. However, that doesn't have any real bearing on the issue at hand here, which is how multitasking operates.

I do not think that is the issue at hand. I think the issue is that you cannot leave me alone. You must dispute me at every turn, and find ways to "prove" you are correct in the face of a different reality.

No, I dispute you where I believe that you are in error. I'm happy to be shown where I'm wrong. It's happened many times. This has nothing to do with you personally, as far as I'm concerned. It's simply about correcting statements that were made on a public forum--not to mention your continuing references to me and to the past discussion.

Matt

new topic     » goto parent     » topic index » view message » categorize

34. Re: task msging

mattlewis said...
useless said...
mattlewis said...

The way that task yielding works. Go back and re-read what I wrote. Especially the ball analogy. You admitted that your description, while it describes how multitasking works under a specific set of circumstances (which you only specified when you gave the example with the sequence of task numbers executed), does not accurately describe how multitasking works.

Well, the specific set of circumstances i optomise to result in greater speed, distributed over the tasks in such a way that it doesn't always work like you say it does. Why are you fighting this?

Yes, I agree, that in that circumstance, that's how it behaves. That was not the context of your previous statements. Why is this difficult to understand?

It's not at all difficult to understand. I don't know why you think i cannot understand that. What i don't feel like doing is discuss ANY thing with you. You're the one who could not fathom the meaning of "to", why should i bother telling you how http.e can be made to crash, or give you the code to fix it? Why give you any context AT ALL, seeing as how this is how talking to you goes on. This conversation went on far too long on the devlist, and now it's here. You ran me off the devlist, i can only assume you mean to do the same here. There's nothing for us to discuss, yet every time you post, you haveto say that i don't understand something, or i don't work well with people who prefer to pick me apart than get on with the coding.

mattlewis said...

We can't read your mind. You were making general statements, while apparently only meaning them to apply to some set of circumstances that only you knew about at the time.

We never got past what "to" meant, and your incessant pasting from the source code and docs. We never got to the context. We never got to anything useful. The statements were not general to my programming style, they were specific. It was you who took off on tangents and generalised them. To make sure that doesn't happen, you may have noticed the task manager code isn't on my site any more, neither is the http.e fixes, or news4.ex or news5.ex.

mattlewis said...

Like I said before, when engaging in this sort of discussion, you need to be able to see beyond your code, or any particular example. At a minimum, you need to make sure that the participants are aware of the context.

useless said...
mattlewis said...

I certainly didn't write news.ex. I wasn't even aware of it until there was a bug report filed about it crashing on linux (reported a week or two before you started working on it, IIRC), and then you enhanced the http library to be non-blocking. That was good work. However, that doesn't have any real bearing on the issue at hand here, which is how multitasking operates.

I do not think that is the issue at hand. I think the issue is that you cannot leave me alone. You must dispute me at every turn, and find ways to "prove" you are correct in the face of a different reality.

No, I dispute you where I believe that you are in error. I'm happy to be shown where I'm wrong. It's happened many times. This has nothing to do with you personally, as far as I'm concerned. It's simply about correcting statements that were made on a public forum--not to mention your continuing references to me and to the past discussion.

Matt

Hey Matt, you are WRONG. And i never learn from those i have no respect in. So you can give it up. You know how i can get you to post to a thread? I post first.

useless

new topic     » goto parent     » topic index » view message » categorize

35. Re: task msging

useless said...

Hey Matt, you are WRONG. And i never learn from those i have no respect in. So you can give it up. You know how i can get you to post to a thread? I post first.

Thanks for proving my point. I'll stop now.

Matt

new topic     » goto parent     » topic index » view message » categorize

36. Re: task msging

mattlewis said...
useless said...

Hey Matt, you are WRONG. And i never learn from those i have no respect in. So you can give it up. You know how i can get you to post to a thread? I post first.

Thanks for proving my point. I'll stop now.

Matt

I don't believe there was a point. What was it, and why didn't you state it up front? We can't read your mind, you know.

useless

new topic     » goto parent     » topic index » view message » categorize

37. Re: task msging

useless said...

You must be the one who wrote the first news.ex, which i sped up 10 times by doing it my way.

Before anyone else get's blamed for writing news.ex, news.ex was in the system for a long time, in fact, it's in SVN r1, i.e. the day it was open sourced. Thus, none of the current authors wrote news.ex. I would guess Rob did. So, you should probably send him an email explaining how you sped it up 10 times.

Jeremy

new topic     » goto parent     » topic index » view message » categorize

38. Re: task msging

jeremy said...
useless said...

You must be the one who wrote the first news.ex, which i sped up 10 times by doing it my way.

Before anyone else get's blamed for writing news.ex, news.ex was in the system for a long time, in fact, it's in SVN r1, i.e. the day it was open sourced. Thus, none of the current authors wrote news.ex. I would guess Rob did. So, you should probably send him an email explaining how you sped it up 10 times.

Jeremy

Why would i do that? He's on the devlist when i explained it there. Besides, i was only looking for a reason Matt might have to be giving me so much of his attention. It's not doing me any good or Euphoria.

useless

new topic     » goto parent     » topic index » view message » categorize

Search



Quick Links

User menu

Not signed in.

Misc Menu