121 replies to this topic

[MedEvil]

Because, one can tell Peter, that there is a problem, that needs fixing and he'll work like a horse to fix it.
But never, ever, under any circumstances, tell him how! He will kill himself to do it another way, any other way.

[Brito]

We never get an answer to our question from Nuno. (Does Nuno approve crpytic syntax ?)

My answer is: "It depends"

Many of us would like a precise answer to the question: "What's the correct staffing ratio for developers to testers in my product development organization?" Usually though, the only answer is "It depends". Your answer depends on your situation: the kind of project you're working on, your schedule constraints, the culture you work in, and the quality expectations for the product.

http://www.jrothman..../ItDepends.html

If you read carefully the paper that I referred on the above link,
you will certainly find a lot of similarities to the case that we are
now debating here.

---------

First I'll write the test cases and document the correct syntax
that is expected. Only then I'll have the conditions to tell which
decision is more appropriate than another for a given way of
doing things.

I dislike cryptic syntax as much as the next developer but I will base
my opinion in facts and try to understand where and why this is needed.

With some luck I might find some free time to start tonight, will keep you
guys updated and then we'll have something to actually talk about with
facts and working code.

[sbaeder]

Critical Troubles begins on variable handling (and check example given by Nuno about macro and Galapo) which cause unpredictability.

Galapo open the Right topic with Right example trying to show the upcoming storm. If it was replied honestly there, we would not spend so many days writing same things over and over again.

YES, this is the crux of the issue...and we need to get back to that, and quit pointing fingers at each other.

Are we going to allow complex expressions to be contained inside a variable, and if so, how far does it go? A variable inside a variable? To me, if we keep it SIMPLE - i.e. a comma (for better or worse) is used to separate
the parameters being passed in as arguments. If there is a variable, and it is going to have "commas" inside it, then they have to be either enclosed in quotes (i.e. have a way to get the quote into the variable) or it has to be the "escaped" comma.
I would certainly hope that 99+% of things could be properly expressed without the crazy #c.

My "vote" certainly is for simple, clear and consistent rules on the syntax. And that we always have full discussion, and write the manual (and get approval on the syntax, with a lot of examples) before we write the code. Speed of processing certainly 2nd place to clarity and consistency

[Xiaopang]

Not trying to butt in on your discussion or anything, but what I don't understand is what the problem is. People want to use quotes as in every major scripting and programming environment. That's totally understandable and I would also have a hard time adapting to the escape character way. PSC on the other side wants to speed up the script processing. That's also an understandably important goal. What I don't understand is why he doesn't write a preprocessor that translates the slower syntax into the faster one upon script execution. Shouldn't be much of a problem if the syntax doesn't change in general. This way everyone would be happy. But what do I know. That's just how I would do it

[Lancelot]

My answer is: "It depends"

In the past it was not needed, now it depends......
your final decision does not represent the votes of current vote topic,
besides final decision is yours.
An answer is an answer.

As a result why not these "dependable" syntaxes have explanations at the end of syntax_page with a warning.

example:

Dependable Warning: Set, can not be used properly when includes comma, you must use escape #$c
Ex:
This is wrong: Set,MacroTest,"Exec,%scriptfile%,test",PERMANENT
This is right: Set,MacroTest,Exec#$c%scriptfile%#$ctest,PERMANENT

My only concern is having these dependable warnings increasing in time which as a result finishing with cryptic codes.
I hope you can make decisions for "dependecablities" for today (hoping low in number) and strick to these decisions after.

I dislike cryptic syntax as much as the next developer but I will base
my opinion in facts and try to understand where and why this is needed.

I can give more details either in this topic or in another topic. As far as i understand from your macro example, on wb code not using quote as flag handled by checking other things per syntax.
(in your example setmacro and i guess there is a missing check of existing permanent which can hopefully be fixed in wb codes).
Well you are programmer, check syntax codes and workarounds used to tolerate quotes on other syntaxes. (in wb source code)

btw, also check wb074 source codes to see how things changed after.

----------------
EDIT: I just remember now, psc already wrote this set thingy is working now

Your code works in the current alpha

Since your answer to the question is depends, it cause a natural question, What will depend? which is waiting your investigations on codes. good luck
----------------


But most importantly, please ask your questions on a topic so we can answer them before you come to a wrong conclusion. Best way to avoid false accusations which I hope you understand.





@sbaeder , @Xiaopang
I think the same with you and my requests have same goals of what you've write. (and I am also sure Galapo's and others too)


EDIT:

[paraglider]

6 pages and probably only 2 examples showing the use of , . I vote for readability. So I much prefer:

echo,"Hello, I'm here!"

Using " as a means to avoid escape sequences is common in all languages I have used.

Unless " are being totally removed from winbuilder syntax then I can't see how allowing the above example could possibly have a large impact on performance of the parser.

Winbuilder is probably unique in using 3 characters for escape sequences. Most languages only use 2 character escape sequences.

[paraglider]

One good reason for allowing a script to set variables that are visible to other scripts is for things like the WAIK directory. That is a variable that needs to be set by the user and needs to be visible to many scripts. I agree that it does not need to be permanent though. Global should be good enough.

[pscEx]

One good reason for allowing a script to set variables that are visible to other scripts is for things like the WAIK directory. That is a variable that needs to be set by the user and needs to be visible to many scripts. I agree that it does not need to be permanent though. Global should be good enough.

What you mean, is implemented by

Set.%Variable%,value,GLOBAL

For all scripts processed AFTER this command, %Variable% contains value.

In the next run, at beginning, %Variable% is undefined.

EDIT: I recommend to use PERMANENT:

The internal processes are not slower whether the variable is in the "GLOBAL" rather than in the "PERMANENT" list.
Th only difference is that "PERMANENT" is ONCE writtenj to scriot.Project.

The advantage is, that e.g. in a CodeBox snippet, the PERMANENT variable is known, while the GLOBAL is not (build has been finished)
Question about Paraglider's WAIK location sample: May it be used in a CodeBox snippet?

If some thinks it is necessary, at end of build he / she can say:
Set,%Variable%,NIL,PERMANENT
Then it disappears from script.project.



Peter

[pscEx]

What I don't understand is why he doesn't write a preprocessor that translates the slower syntax into the faster one upon script execution. Shouldn't be much of a problem if the syntax doesn't change in general. This way everyone would be happy. But what do I know. That's just how I would do it

Since some days I'm working on it. And it looks rather well.

But since yesterday that will be ONLY an enhancement on my personal WinBuilder.

I stopped publishing because I was sad to waste time in answering to "Many Many Many Many Many Many times" posts and similar.

Fact: When I do not publish, people have no reason to complain, agress me or tell that I'm lying.
And I personally feel happy with my personal WB version.

@Nuno: No harm that I'll publish my version anywhere. That's just my personal joy.

Peter

[amalux]

OMG, I expect better from my 4 year old!

[amalux]

Since some days I'm working on it. And it looks rather well.

But since yesterday that will be ONLY an enhancement on my personal WinBuilder.

I stopped publishing because I was sad to waste time in answering to "Many Many Many Many Many Many times" posts and similar.

I personally feel happy with my personal WB version.

@Nuno, so when can we get an adult who can actually continue development for the rest of us?

[Lancelot]

Hi amalux,

@Nuno, so when can we get an adult who can actually continue development for the rest of us?

I guess Nuno start again

.....With some luck I might find some free time to start tonight, will keep you guys updated and then we'll have something to actually talk about with
facts and working code.

Personally I am in "waiting updates" mode and happly (without need to reply to false things on topics) working on scripts.

[Xiaopang]

Since some days I'm working on it. And it looks rather well.

Doesn't that mean that all this discussion is moot and the syntax basically can stay the same?

I stopped publishing because I was sad to waste time in answering to "Many Many Many Many Many Many times" posts and similar.

I understand where you're coming from. In fact, my viewpoint regarding my personal releases is not much different from yours. However, why don't you just release the version with the transcription layer you've been working on and be done with all the problems? This version would certainly be superior to all others so far and would end the debate, wouldn't it? I wonder why you didn't use this solution in the first place instead of revamping the syntax? Isn't it a better concept to keep the workload on the program instead of piling it onto the user? Ah, I'm already part of the discussion and I wanted to keep quiet. Don't see this as an attempt to question your skills.

No more questions from me <<< Hint for Nuno: Would you believe that the board lacks a proper smiley for this expression

[pscEx]

OMG, I expect better from my 4 year old!

I'm a bit older, and I currently made for the third or fourth time this experience of beeing agressed rather than beeing asked to discuss questions.

I think that your child already at the third time would react with ignorance. Childs learn quickly.

Try to tell your child: "As I already told you many many many many times" or "As I told you loooooong time ago"
At least all of my 4 childs would have react with "Forget it", like me now.

Btw: Nobody besides me is allowed to control my feelings.

Has been a nice time!

Peter

[amalux]

All the issues of late stem from a horrible lack of communication, people get frustrated and eventually assume the worst when they can't get a strait answer. Along with others seeking answers that only you can provide, I tried my best yesterday to bring some focus and clarity and respectfulness to this discussion and this is your response? I expected better.

[Brito]

@Nuno, so when can we get an adult who can actually continue development for the rest of us?

I actually understand Peter's perspective.

We are looking on the wrong direction. Projects like LiveXP are overly complex and winbuilder.exe cannot be expected to continue feeding this problem.

Instead of blaming a wb developer for trying to find a solution to .script developers, try adopting a simpler scripting because it is indeed possible when you put your mind to it.

If you look on projects like UXP or WinRoot:
http://winbuilder.ne...oad.php?view.24
http://winbuilder.ne...load.php?view.6

You'll see how scripting should be simple and painless to achieve good results.

We don't need crazy macros nor strict encoded characters to achieve these results, but it will depend on .script developer's effort to keep it simple.

[amalux]

This has nothing to do with anything you mentioned so I understand what Galapo and Lancelot have been saying, you don't get it. This is only about asking simple, respectful questions e.g. 'will removing commas and quotes increase processing speed enough to justify more cryptic syntax?' etc. and getting 'I'll do it for myself but you can't have it, na na na na na ..' responses.

[Lancelot]

I actually understand Peter's perspective.

We are looking on the wrong direction. Projects like LiveXP are overly complex and winbuilder.exe cannot be expected to continue feeding this problem.

LiveXP always followed psc's advices with its evolving period. If it does NOT work for anything than there is problem with wb development handling syntax rules, not livexp or any other project.

Instead of blaming a wb developer for trying to find a solution to .script developers, try adopting a simpler scripting because it is indeed possible when you put your mind to it.

AGAIN WRONG. we do not blame a wb developer for trying to find a solution to script developer. WE ASK the syntaxes worked as written. INSTEAD we get a Silent syntax update which is opposite to what've been written all these 2 years.

Even today, we can not get clear answer from "a developer" to the question "What Depends"

Above statement I guess (since not clearly written) says Nuno approves "cryptic syntax"

"It is not a problem for us to write "cryptic syntax", If This is what Development decides."
==>
It#$sis#$snot#$sa#$sproblem#$sfor#$sus#$sto#$swrite#$s#$qcryptic syntax#$q#$c#$sIf#$sThis#$sis#$swhat#$sDevelopment#$sdecides.

It only proves after 2 years (and maybe more) of lies saying quotes (normal) and none-quotes (cryptic) are optional, since yesterday

But to use the optional "Strict" parser

Decision is now cryptic is a must. Hence the working wb lines of today are in fact false and avoids syntax rule 3.


Thanks Nuno, for your decision, for your questions (where no exists), for unclear answer and for other things you made us live through these days.

Happy#$swinbuilder#$sscripting#$sto#$sAll.

[patsch]

This has nothing to do with anything you mentioned so I understand what Galapo and Lancelot have been saying, you don't get it. This is only about asking simple, respectful questions e.g. 'will removing commas and quotes increase processing speed enough to justify more cryptic syntax?' etc. and getting 'I'll do it for myself but you can't have it, na na na na na ..' responses.

imho psc would not have spent time in this if this syntax correction would not have been showing much improvements ... this is my personal estimation because he spent a lot work in improving win7rescuepe to make it faster and more comfortable ( = his project nativeEx Win7)

[amalux]

imho psc would not have spent time in this if this syntax correction would not have been showing much improvements ... this is my personal estimation because he spent a lot work in improving win7rescuepe to make it faster and more comfortable ( = his project nativeEx Win7)

Good, I'll ask you from now on if I have a question for Peter

[patsch]

Good, I'll ask you from now on if I have a question for Peter

what is your problem man?
what is so difficult about "imho" and "my estimation"?
it is wrong to write my personal opinion?
you have my great respect

[amalux]

what is your problem man?
what is so difficult about "imho" and "my estimation"?
it is wrong to write my personal opinion?
you have my great respect

Just a joke, man, take it easy

[Lancelot]

win7rescuepe to make it faster

it is only a cache mechanism. Simply caching core files with initial build by packing than with 7z and extracting with 2nd 3rd big blues.

It was already (before psc reinvented) written many times by me and by others that same can be done even for application scripts (which i experimented having 10x and more speeds on 2nd 3rd runs) but since we are dealing with medusa bugs (which means working things again not working) a lot time and with having not constant script syntax, never publiced and finished.

Nothing miracle, cache is a well known, very old way of speed up. Besides not related to "cryptic syntax" and "none cryptic syntax".

Welcome#$sto#$snew#$swinbuilder#$sera#$c#$sHappy#$scryptic#$ssyntaxing

[patsch]

it is only a cache mechanism. Simply caching core files with initial build by packing than with 7z and extracting with 2nd 3rd big blues

It was already (before psc reinvented) written many times by me and by others that same can be done even for application scripts (which i experimented having 10x and more speeds on 2nd 3rd runs) but since we are dealing with medusa bugs (which means working things again not working) a lot time and with having not constant script syntax, never publiced and finished.

might be that you also had the idea but Peter uses it massively with sucess ... that is the difference

and back to topic: I do not have any problems with these escape sequences (or strict syntax or in your words "cryptic syntax"). Since I'm not a good programmer I have to find my way to simplify my life, for example in using replacements (like in latex) for certain commands ... for me autohotkey is a very good tool to make my life easier in creating my simple working scripts

[amalux]

and back to topic: I do not have any problems with these escape sequences (or strict syntax or in your words "cryptic syntax").

So, you changed your mind?

when an apostrophe is allowed, why not a comma?
how is a string stored within winbuilder that it could causes problems?
are there parser that have problems in distinguish between comma as a string and coma as a command?
In my opinion, in a string every literal should be allowed and not be represented by its escape sequence or other placeholders