World Wide Rabbit - and what can be done to fix it?
[1/12] from: jason:cunliffe:verizon at: 19-Oct-2002 19:38
Hi After playing with Java Web Start, it is clear that REBOL better get its View/Desktop act together fast or else be run over by a fleet of snarling coffee-cups. http://java.sun.com/features/2002/10/webstart.html I ran View/Desktop -first time for a while as my default is shell mode]. gee: There are now about 130+ rebsites..! A. This is good news. B. This is bad news. Why? Because there is a golden moment for Rebol to really shine in contrast to the Java Web Start, but we still have a semi-implemented View/Desktop interface. One which is not designed to cope with large or growing number of rebsites... Is there some new/better version I missed? WHAT'S WRONG WITH View/DESKTOP [Top 5]? 1. DISPLAY MODES No scrollable list 'mode' to display sites as alternative of big icons - 1st column: name and perhaps mini-icon, - 2nd column: the descriptor - 3rd column: add some other metadata such as popularity, age, last visited, last changed, number of files.. 2. SEARCH 'n SORT There is no built-in 'search' or 'sort' doh..why why why ??? With a list such as the above one should be able to search and sort. - take a look at Limewire for example of a pretty neat interface: http://www.limewire.com/ - It won't be long before Sun copies that for their Java Web Start front end 3. STATUS No indicator of on-line or off-line status. - How hard would it be for to at least ping/poke all the rebsites and then ghost out the off-line ones? - How hard to add some RIM or Jabber-like presence messaging to improve that idea? - How hard to indicate? Maybe use a Red Dot on any sites who wish to let people know they had something new. Even better a way so that when you go there it will be obvious what is new. Once visited the dot becomes very muted in color. But the site owner decides when they no longer want to display it as 'new'. 4. FEEDBACK No feedback indicator of rebsites already visited, or not. - At the very minimum, Desktop needs to do what HTML has been doing for 10+ years:
>>> Give users simple graphic feedback of what they have already visited. <<<
There are lots of cheap fast way to do this which Rebol/View is eminently capable of supporting. - Especially when one is new, visiting 130+ sites is fucking confusing. Most have several levels anyway, and default to the same style. 5. BOOKMARKING No fast and friendly way to bookmark sites our group theme in convenient sets of 'favorites'. - Yes you can make your own rebsite, and then set it up to add links to others. But that's dumb. No doubt ther are others strategies people use with user.r Rebol/View already has a caching scheme. It needs to use it for more than just code. So that people can quickly add and delete named groups of sites or sub-sites, pages or files. IMO These belong on their own new REBSITES MENU and one should be able to select it any time to time: ADD current rebsite to group ..etc What do you think? ./Jason
[2/12] from: rebolinth::nodep::dds::nl at: 20-Oct-2002 14:31
Hi Jason, * just a little addon: I think that revol/view is an example brought to us by Carl/RT to show what can be done with a dialect in Rebol. Build your own /view is the suggestion.. Its not an exident that the Rebol/Base just popped up out of the blue, I think the RT got just too many questions on enhancements on the /view and /core part. Well if you have a /base version, then people can build there own /Etc.. version.. This goes also for the /command and IOS version... those are for the highend market, I cant effort them and im seriously thinking of starting a community on the internet that builds a Free /command or /IOS version. Still when we look at the TimeLine of Programming languages from 1954 upto now, I think Rebol is (besides Tcl/tk and Pascal) the best Language that comes near the normal behavior of Human thinking and expectations, still new programming languages and dialects will follow, I sure do hope so! Anyway, the RT should loosen a Little and give the community the chance to improve Rebol and give Rebol-language a Chance to be IT! But this is not happening the way i hoped they would do it. I undertsand that its a Money and (C) issue from the side of the RT, but a language can only be a Language if it Gets its honor... I have a Lot of programmer around me and they all swear on C/C++, pitty it is, i cant get them to work with rebol, And im wondering why? Rebol is amost there...But they should explode within 1 year!! otherwise it will just become 1 out of 10 other languages...that would be a pitty. Greetings, Norman.
[3/12] from: edanaii:cox at: 20-Oct-2002 8:43
>Hi Jason, >
>I have a Lot of programmer around me and they all swear on C/C++, pitty it is, >i cant get them to work with rebol, And im wondering why? >
Their are several reasons for this, IMHO. First, because computer geeks are limited in their thinking. They are some of the smarter people in this world, but they achieve their ability by limiting the things they are willing to try. Second, as a computer geek, I've always been a fan of faster, better, easier. This is the normal trend for almost everything but computers. Your average computer nerd, OTOH, seems to prefer faster, better, more complicated. This gives them a distinct advantage ; the more incomprehensible the task is, the more genius they appear. Third, is the fear of anything new. For your average nerd, the uncertainty of anything new means he runs the risk of not looking as smart as he would like to. It's sad, really. Over the years I've watched a number of new technologies come and go. And not because they weren't better, but because no one was willing to try them. Amiga, anyone? :) As a fan of that old machine, I loved its ease of use, its intuitiveness, and of course, its power. And what computer wins the computer wars, M$, of course. But more to my point, what computer won the hearts of Computer Geeks everywhere? Linux. Both are systems that are powerful and usable, but just plain incomprehensible to your average Joe Another case in point, I've watched a number of languages come and go over the years, we've had Fortran, COBOL, BASIC, COMAL, Forth, and many more come and go. Some of which attempted to make programming easier for all to understand (COBOL, BASIC), but what languages have actually reigned supreme? C, C++ and Java. All languages that are admittedly powerful, but are just as equally incomprehensible to J. Q. Public. Computers and languages belong to the Mentacrocy, if I may coin a phrase, where the smart people are clearly in charge and intend to stay there by keeping everyone else as confused as possible. And I doubt that it will change. At least it won't until the computers become smart enough to know what we are asking for, not what we are telling them. :) I applaud Carl for trying to make REBOL easier for all to comprehend, but as long as nerds everywhere refuse to try anything new, as long as they prefer to be geniuses, and limited in there thinking, Carl, and all others who would try and make the computer world and easier place, have a tough battle ahead of them. -- Sincerely, | We're Human Beings, with the blood of a million Ed Dana | savage years on our hands! But we can stop it! We Software Developer | can admit we're killers, but we're not going to 1Ghz Athlon Amiga | kill today. That's all it takes! Knowing that we're | not going to kill... Today! -- Star Trek.
[4/12] from: steve:shireman:semaxwireless at: 20-Oct-2002 12:05
Ed Dana wrot
> I applaud Carl for trying to make REBOL easier for all to comprehend, > but as long as nerds everywhere refuse to try anything new, as long as > they prefer to be geniuses, and limited in there thinking, Carl, and > all others who would try and make the computer world and easier place, > have a tough battle ahead of them. >
Pogo said, "We have met the enemy, and they is us". As developers, we get modern messaging, dialect and style (distributed + soft architecture + reuse) technology, but we ask for old fashioned libraries, objects, and SDK. It is old bad habits we all have to break. Anyway, there is no need for any "battle". Just do it right this time. Let's not be so concerned what outsiders think of what we are doing. We have advantages that put us ahead. As Dale Carnegie said, the only way to win an argument is to not have one. You don't change the current world, you make a new world. Steve Shireman
[5/12] from: greggirwin:mindspring at: 20-Oct-2002 12:47
Ed, Jason et al I have a different view of REBOL, and what it means for REBOL to "succeed". I don't expect REBOL to displace C++, Java, or the WWW anytime soon, but that doesn't matter to me. What matters to me is that RT can stay alive, no matter how small they are, and continue provide the tools and toys we've come to know and love. If they can make a good living, and I can use my tool of choice to help people solve their problems, I'm happy. Jason, you said that you got a new DL of Java from Sun because yours got mangled (which apparently happens a lot). How often has your REBOL installation become unstable? Don't forget the size difference either. To me REBOL's design is based on general principles applied in many ways, which tends to mean that any particular feature, no matter how well it's implemented, will not be optimized for every possible use. Another thing that is often, if not always, in my mind is that a lot of the features and examples in REBOL are just "road signs" for us to follow, so we have some familiar language "landmarks" as we come to REBOL from other langugages. When we see language, GUI, or runtime features in other languages, we tend (myself included) to think "Hey! They could add that to REBOL and it would be even better!", but this is not always true. Maybe it never is. :) REBOL is different, and it allows us to solve problems in different ways. <<WHAT'S WRONG WITH View/DESKTOP [Top 5]? 1. DISPLAY MODES No scrollable list 'mode' to display sites as alternative of big icons >> OK, this is a good example I think. I agree that a list mode might be a nice feature, but how much does it buy you? I.e. how many items can you navigate easily in icon mode, versus list mode? 100 vs. 500 maybe? Not a bad gain, to be sure, but is it enough? How many sites are going to be out there? 10,000? 100,000? Are mini-icons really that helpful? Unless they're a well designed, scaled down version of a larger icon, can they really provide a lot of meaning? Searching and sorting, as you said, are very important. How easy will it be to search Java WAR (web archive) files versus REBOL data? << - take a look at Limewire for example of a pretty neat interface: http://www.limewire.com/ >> I didn't download it (I'm on slow dial-up) but it does look pretty neat all right. Now, consider that the LimeWire team is developing an application that runs on three platforms and their development team is probably as large as RTs (if not larger) while RT is developing a sub-platform that runs on *lots* of platforms. What would it take to do something like LimeWire in REBOL? Is that something that we should do, or RT should do? RT isn't really in the "applications" business at this point. << 5. BOOKMARKING No fast and friendly way to bookmark sites our group theme in convenient sets of 'favorites'. >> What about bookmarks.r? OK, there's no menu option that says "Add Bookmark", but you can tweak the desktop quite readily, though I certainly haven't excercised this to its full potential myself (e.g. adding new services). Since the Desktop is part of View, everything that is added to it adds to the basic View package. Maybe someone who knows will jump in and tell us how we can call code snippets from custom Service or Bookmark entries and we can configure the Desktop ourselves. That doesn't solve all our problems, but it would be pretty handy. Some of the features you mentioned (highlighting visited links) might not be hard to do, especially considering that each face can have its own effect, or have its icon composited with another image. Maybe they could just give us a hook. Send it to feedback and see what they say. In looking, briefly, at the Java Web Start stuff, something else struck me. The JNLP (Java Network Loading Protocol) has to be set up for each app you want to load, and you have to configure the web server for a new MIME type, correct? Contrast that with setting up a reb site. On to some thoughts on languages and such... << Another case in point, I've watched a number of languages come and go over the years, we've had Fortran, COBOL, BASIC, COMAL, Forth, and many more come and go. Some of which attempted to make programming easier for all to understand (COBOL, BASIC), but what languages have actually reigned supreme? C, C++ and Java. All languages that are admittedly powerful, but are just as equally incomprehensible to J. Q. Public. >> I'll differ with Ed's view a bit here. COBOL and BASIC are far from gone, and Forth was never really a big time language, but it's not gone either (you can program Lego Mindstorms with pbForth :). I have to admit that I've never heard of COMAL (oCAML yes, COMAL no). Now, I don't know of anybody who learns COBOL today because it's hip or cool, and VB has grown ever further from its BASIC language roots, but they're still out there serving a need. Whether a language is powerful, to me, is determined by how well it lets you express yourself and solve problems efficiently. This means different things in different contexts. Assembly is powerful in an entirely different way than SQL, but both are powerful and each limits your thinking in certain ways. You can't solve a problem that you can't express. <<...as long as they prefer to be geniuses, and limited in there thinking, Carl, and all others who would try and make the computer world and easier place, have a tough battle ahead of them. >> Until the time is right which, unfortunately, sometimes doesn't happen in time. :) VB overcame its "BASIC" stigma because it provided something that people needed. Now lots of tools and languages have emulated it in many ways. I think the same may happen with REBOL. The "smart people" will see that it provides something we need (something we're not really exploiting very well yet: messaging) and, at the same time, is accessible. Lisp and Scheme, along with many other languages, are very powerful, but they are not accessible (IMHO). If you've seen "Field of Dreams", think of those great lines James Earl Jones has when he's saying "people will come...they won't know why...", etc. Maybe they won't come to REBOL, but REBOL represents something they want. --Gregg
[6/12] from: gchiu:compkarori at: 21-Oct-2002 13:18
>never heard of COMAL (oCAML yes, COMAL no). Now, I don't
Hi Gregg, I have a COMAL 80 interpreter for my CBM C64 sitting on my shelves :) Sort of a hybrid Basic/Pascal. -- Graham Chiu
[7/12] from: greggirwin:mindspring at: 20-Oct-2002 23:03
Thanks Graham, << I have a COMAL 80 interpreter for my CBM C64 sitting on my shelves :) Sort of a hybrid Basic/Pascal. >> Of course, I had to go find out what it was, since it wasn't anywhere on my radar/ :) --Gregg
[8/12] from: rebol-list2:seznam:cz at: 21-Oct-2002 13:01
Hello Jason, Sunday, October 20, 2002, 1:38:41 AM, you wrote: JC> Hi JC> After playing with Java Web Start, it is clear that REBOL better get its JC> View/Desktop act together fast or else be run over by a fleet of snarling JC> coffee-cups. ... JC> doh..why why why ??? Because Rebol is just a Rebol and Sun is Sun. I'm not sure you can compare these two companies. Sun is the Goliash and Rebol is David. Few years ago I was claning windows (to clean my head) with my friens who works as climbers on one of large Sun's HQ building in Prague. There must be a lot of such a Sun buildings all over the world and in each of such a building there are working a lot of smart heads. How many people is developing Rebol? 5 or less in RT? And then few of us making it more as a hobby then full time job. I think there are no changes in the desktop, because nobody needs them so much. It's not such a difficult to make it. You just have to spend some time to analyse/change the code of 'ctx-viewtop object/block. You can make your own desktop just replacing the ctx-viewtop on startup (in the user.r file)
>> source desktop
desktop: func [ if block? ctx-viewtop [ctx-viewtop: context ctx-viewtop] ctx-viewtop/init-desktop do-events ] so for example place this into %user.r : ctx-viewtop: make block! [ init-desktop: func[ ;place here your own desktop layout view/new layout [text "Hello" button "hmm" [print now]] ] ] Can you do that in the Java Web Start? PS: I don't know like you, but I prefere Carl working on more important parts of the Rebolution then making another View/desktop - for example if Carl is working on radio-line style (improved radio style) for VID he cannot work for example on Rebol integration into browsers:-(
[9/12] from: anton:lexicon at: 26-Oct-2002 1:14
Oldes, Does this radio-line style really exist? What features does it have more than the usual radio style ? Anton.
[10/12] from: greggirwin:mindspring at: 25-Oct-2002 10:32
Hi Anton, << Does this radio-line style really exist? What features does it have more than the usual radio style ? >> The idea of RADIO-LINE is that it combines a RADIO button with a TEXT face, which is how you use them most often. That way you don't have to create separate items yourself. I think it's planned for a future release. --Gregg
[11/12] from: anton:lexicon at: 28-Oct-2002 1:38
Oh, ok, if that's all it is then some of you would know that I have radio-label style that combines the existing VID radio and label styles together. http://www.lexicon.net/anton/rebol/gui/demo-radio-label.r I am not sure how I feel about making it built-in. It would be handy, though is it worth the extra 3kB ? Though if Rebol Tech. would make one, I imagine it would be more efficient, and make use of some general code that could be used to stick other styles together in the same way - or some other use like that. Anton.
[12/12] from: petr:krenzelok:trz:cz at: 27-Oct-2002 16:38
>Oh, ok, if that's all it is then some of you >would know that I have radio-label style that
<<quoted lines omitted: 4>>>It would be handy, though is it worth the extra >3kB ?
it would be good imo, if Rebol community could would rather produce 100KBs of usefull stuff, which proves itself in real-life or not, rather than worrying about additional 3kb of code :-) What is more - when you will use Encap, it will be compressed ... As for me - I have following priority when working with View - easy of usage. If styles as above menthioned - radio-line, help me to more easily produce usefull UI, then I don't care too much of additional 3kbs, as what is much worse for me is - several similar solutions, from which noone is generally accepted to become part of standard package ... So, that's why there is VID 1.3 project ... you will see, that working with auto-resize system, as well as menu and tree styles is like working with other VID styles ... they should be easy to deploy, yet provide flexible set-ups ... -pekr-
- Quoted lines have been omitted from some messages.
View the message alone to see the lines that have been omitted