1. switch design
- Posted by mattlewis (admin) Sep 11, 2009
- 1530 views
Let me see if I understand you ... you believe it was your effort and persuasive argument that caused the current switch design?
No, I believe my post was the trigger to vastly improve the design.
That's fine. I'm not trying to affect your belief system. In fact, Derek (for one) had been agitating for those changes since the beginning. It had been an on and off thing for a while. It's true that we began discussing it again when you brought it up.
The development of switch happened in several stages, one of which changed the way that fall-through was implemented. And, yes, you were a part of the discussion around the time that happened. Thanks for helping out.
I'm not sure why we keep talking about this. Do you want to be in the release notes? That's easy to accomplish.
Matt
2. Re: switch design
- Posted by Critic Sep 11, 2009
- 1531 views
I'm not sure why we keep talking about this. Do you want to be in the release notes? That's easy to accomplish.
No, I don't want to be in the release notes. In fact this whole discussion is rather pointless, I agree. But now (yes, I admit, it changed) my main point is: Could upcoming language features please be discussed in this forum? I think they'll have more participants then (well, at least me).
3. Re: switch design
- Posted by jeremy (admin) Sep 11, 2009
- 1558 views
my main point is: Could upcoming language features please be discussed in this forum? I think they'll have more participants then (well, at least me).
This has been discussed before. When we first moved to the mailing list it was because legitimate user questions/comments that had nothing to do with the development of Euphoria (the use of, yes) were getting totally swamped in the development chatter. We then created the mailing list. During key planning, the developer list goes crazy. At one point in time there were quite a few users on the development list. Since then they have almost all left because of the traffic and due to the nature of the messages. Much of it is the technical internals of Euphoria which most users don't care about.
Why take that discussion here and drive them off the forum too?
Now, the suggestion of topics may be a good one but who's going to develop that? When there are things that we discuss of major language design changes/additions, we do bring it up on the forum. Many times not until we at least have an idea how to do it, some ideas about how it could be done, etc... For instance, we wouldn't say on the forum "We need a new way to make decisions" and then come up with the switch statement from that. We would at least hash out, see what's possible, different design techniques and then present options and let the discussion go from there.
What I'm almost sure of, though, with the topic discussion is that things will be flying, the majority of the messages will be of no interest to the users and we will have just as much user participation on the developer forum as we do on the developer mailing list. Those who really want to take part in the daily nitty gritty of Euphoria development can/should join the developers mailing list. Items that the developers think the user will be interested in comes to the forum, just as said before.
That seems to be the best balance and make the most sense.
BTW... I am curious. You want to be part of deciding what Euphoria is and becomes? The language stinks and shouldn't even exist according to you. It's a total failure on any count and for the sake of man kind should be destroyed, including any mention of it... At least, that's the impression I get of how you feel towards it.
Jeremy
4. Re: switch design
- Posted by Critic Sep 11, 2009
- 1483 views
When there are things that we discuss of major language design changes/additions, we do bring it up on the forum.
I am fine with this solution. It's just that I have never seen it happen. Never mind.
BTW... I am curious. You want to be part of deciding what Euphoria is and becomes? The language stinks and shouldn't even exist according to you. It's a total failure on any count and for the sake of man kind should be destroyed, including any mention of it... At least, that's the impression I get of how you feel towards it.
I for myself think that there are way too many crappy programming languages. So, yes the world would be a better place without TCL, PHP, Perl, Objective C and Euphoria (just a few examples; the list of sucking programming languages is long...). However, I am a realist. (Who is laughing here? 
) It won't happen. One reason is that it is too costly to switch the implementation language of existing software. So the second best thing is to try to improve the languages until they do not suck that much anymore. Getting rid of implicit fallthrough in switch statements is a TINY step to achieve this goal.
I will never ever choose to use Euphoria. But my job is programming. I often work with several scripts written in different, but similar scripting languages. Maybe I (or someone I care about) will get in touch with and have to maintain some Eu code? It's not likely, but then Perl's success was not likely either given how much crap Perl is.
Of course, I already told you the other (real) reasons why I keep posting here.
5. Re: switch design
- Posted by jeremy (admin) Sep 11, 2009
- 1489 views
I for myself think that there are way too many crappy programming languages.
<snip>
I will never ever choose to use Euphoria.
Why not choose one that is not as crappy as Euphoria and join their dev team or forum? Why waste your time on this junk (according to you)? You said it yourself that you will never choose Euphoria as a language. Seems to me you need to learn that time is of tremendous value. What's the chances of someone you know using Euphoria? Since you think it's such junk, I'm sure you will steer them away from it. Thus, I'm failing to see any benefit for you in wasting one more second of your life here.
I know, personally, if I had your view on the language, I wouldn't be here.
Jeremy
6. Re: switch design
- Posted by Critic Sep 11, 2009
- 1457 views
Why not choose one that is not as crappy as Euphoria and join their dev team or forum?
How come you think I am only posting here?
You said it yourself that you will never choose Euphoria as a language. Seems to me you need to learn that time is of tremendous value.
Thanks for reminding me. Fortunately I do not spend that much time here.
Thus, I'm failing to see any benefit for you in wasting one more second of your life here.
I know, personally, if I had your view on the language, I wouldn't be here.
Do you actually care for me and my valuable time or do you want me to leave? If you want me to leave, why don't you just ban me?
7. Re: switch design
- Posted by DerekParnell (admin) Sep 11, 2009
- 1464 views
I for myself think that there are way too many crappy programming languages. So, yes the world would be a better place without ... Euphoria ... However, I am a realist. ... the second best thing is to try to improve the languages until they do not suck that much anymore.
I will never ever choose to use Euphoria ...
Curious! Ok, let's take you up on your raison d'etre. You participate in the Euphoria forum because you want the help make Euphoria less of a ""sucking programming language". Fine and good ...
Tell us one thing that you feel must change or improve in Euphoria right now. Something that really can't wait any longer. And tell us why it needs to change.
8. Re: switch design
- Posted by Critic Sep 11, 2009
- 1491 views
- Last edited Sep 12, 2009
Curious! Ok, let's take you up on your raison d'etre. You participate in the Euphoria forum because you want the help make Euphoria less of a ""sucking programming language". Fine and good ...
Tell us one thing that you feel must change or improve in Euphoria right now. Something that really can't wait any longer. And tell us why it needs to change.
See our discussion about stacks. Some kind of reference semantics in addition to copying semantics. Because the current workaround of using a global sequence variable + index is everywhere in Eu code. It needs to change because:
- It's awfully inconvenient.
- It encourages memory leaks.
- Copy semantics easily introduces O(n) complexity where O(1) would suffice.
Lack of references is the biggest issue with Euphoria right now, it really can't wait any longer.
Forked into: Pass by Reference
9. Re: switch design
- Posted by kinz Sep 12, 2009
- 1496 views
Hey, gentlemen, where is something about real switch design in this topic?
Just curious...
Regards,
kinz
10. Re: switch design
- Posted by mattlewis (admin) Sep 12, 2009
- 1385 views
Hey, gentlemen, where is something about real switch design in this topic?
Yes, it was really a meta-subject, talking about the process of designing the switch.
Matt
11. Re: switch design
- Posted by Fendaril Sep 12, 2009
- 1363 views
I read the manual on switch design but I don't see much mention of what with faullthrough does. is it like an alternate break?
12. Re: switch design
- Posted by jeremy (admin) Sep 12, 2009
- 1360 views
I read the manual on switch design but I don't see much mention of what with faullthrough does. is it like an alternate break?
Hm, the manual needs to be expanded on this subject then. Can you create a ticket for it?
What fallthrough does is this:
integer a = 2 switch a with fallthtru do case 1 then ? 1 break case 2 then ? 2 -- notice no break case 3 then ? 3 break case 4 then ? 4 end switch
The output of the above program is:
2 3
Once an initial match is made (i.e. case 2 then), fallthru causes all remaining case statements to be ignored. They are actually just jumped right over, no test is done.
P.S. I sent an email to your new email address and have not gotten a response. Did you enter your correct address?
Jeremy
14. Re: switch design
- Posted by jeremy (admin) Sep 12, 2009
- 1381 views
i responded
Thanks.
Jeremy
