AltME groups: search

sure, gracias for curl kaj. very usefull component and already working 
;-)

That's curious. If I try to unpack    2.4 MB glass-r003.zip, I get 
a folder with just 1.3 MB and only two files in glass/libs. I tried 
with 7zip and the WinExplorer.

My distro is OK ... using Total Commander and its internal zip tool 
to "unpack" (it treats archives as regular dirs)

strange... its ok with some tools and not in others (winrar also 
ok)... the debug dir is an empty dir.

And the constrast between black background and white box hurt my 
poor eyes too.
I wold prefer code block in dark gray bg and white texte !

DideC, yes, we know. Already working to fix it. This comes from using 
MDP format internally and just hard-copy the generated HTML code, 
without style adaption.

I somehow agree with Ashley, though I can understand that you have 
some structure set. The top level menu simply does not show - Docs 
directly. So intuitively I went to Code, and then - huh, what's Level-1? 
If I would not read more about your infrastructure, I would feel 
a bit lost probably :-)

But it's not a big deal to change the structure... if it adds value 
and doesn't bloat the menus.

Thanks to both Kaj and Andreas for taking up the 0mq task. Since 
Andreas made it so easy to try, with prebuilt binaries, I already 
played a bit, and it works fine here. Very exciting stuff guys.

Sweet. Thanks Kaj and Andreas.

Yes, it works with the RMA build, but my tests are to check the level 
of compatibility with the standard r3 build, and, to that respect, 
I need to replace the load-gui by a do %r3-gui.r3. (At least, I suppose, 
I have not checked)

I understand, but, basically, it should be out of the box : the r3dll.dll 
is the standard one, and as I said, the only thing I added is the 
line #define TO_WIN32, which in any case, should be declared.

About disreb: Sorry to be unclear originally. At the moment, it's 
R2 only (I'm currently sticking with the latest version of R2 because 
I understand that it has a more stable GUI), however I am very willing 
to incorporate changes for R3. I'm not yet sure what the best fashion 
of collaboration with the Google code site is, but hopefully I'll 
have some time to investigate tomorrow. Sorry for the late reply, 
but I have imperfect access to AltMe (it's firewalled at work and 
I forgot my password (ha!) so I don't have it working on the computer 
I'm programming disreb with). @Oldes, I hope you find it useful should 
you chance to use it! Thanks!

Robert, it can do both sync and async. But no callbacks, async is 
done using a WAIT-style (or select/epoll-style, if you prefer) approach, 
using zmq-poll.

Just to be sure I understand: 0MQ makes async call however it does 
it, and between 0MQ and Rebol the async is that I have to poll from 
Rebol side to check if the call is already done?

And big kudos for making the Redis library!

Robert: you have two primitives send & recv. Both _block_ per default: 
send until there it manages to enqueue the message in an internal 
buffer, recv until it fully read a message. You call both functions 
with a NOBLOCK flag, in which case they won't block but return immediately, 
with an EAGAIN status code in case they did not manage to write/read 
anything. Finally, you also have a poll primitive, to which you pass 
a list of sockets you are interested in and an (optional) timeout. 
Poll returns you a list of sockets which are "active", i.e. which 
can be safely read from (or written to) without blocking.

Why not leave panel and hpanel as synonyms, and group/hgroup ?

Thanks to the RMA team for their progress posts and updates.

I clicked on the "Draw" button for the plotterdemo and nothing happens 
??

Odd, it doesn't warn me if the skew is within a few minutes or so. 
Anyway, it seem to be just a warning, and I wouldn't worry about 
13 seconds

Sure, and Fossil will nicely warn me when it gets significant, and 
then I'll adjust it

Strange, not for me, and the time of my workstation is routinely 
off

So, I can't replicate your problem, and it doesn't prevent Fossil 
from functioning, does it?

It does not, but it'll give everyone with a synced time a warning 
on each and every clone/pull.

Or actually more likely, it doesn't care about absolute time, but 
only about the relative time of the server and the local machine

And it will try to prevent you from creating a check-ins that are 
"earlier" than their parents.

It doesn't depend on time stamps. Each file has an SHA1 hash, like 
Monotone, Git and Arch 2

No Cron, and I don't plan on adding it

that estimates clock drift and adapts accordingly

Fossil serves distributed versioned databases and it only has 500 
KB resident right now

I think for daytime all you need to do is open a port on 13, read 
what is returned and close the port.

Show them HELP (?) and SOURCE, which are strong tools to get you 
going as a programmer.

And how to build sentences from words and values without irritating 
syntax to get in the way.

The benefit from having so many datatypes, which make functions work 
in many ways depending on argument datatype. One good example, I 
think, is READ, but you can probably find others.

probably not, but why not outline the planned future? R3 is alpha, 
and it should be treated as such. R3 adds nice stuff - unicode, modules, 
control of boot sequence, devices, codecs, etc. Robert might face 
the questions of recent REBOL status anyway ....

REBOL has fostered ideas in many people, and I'm sure, whose ideas 
will evolve into a future with better programming tools, than the 
world has today. So yes, I'll say there is a future.

They are there, just not finished and optimal ...

There are reasons that you don't see anybody writing extra devices 
and codecs

Kaj - I think  that you are not accurate, and that you even know 
it?. Codecs are useless, and even Carl later introduced the plan,of 
how to reimplement them. But devices? There is list of devices out 
there, no? http://www.rebol.net/wiki/DevicesThe other fact is, that 
those are not usable via extensions, but it does not mean the abstraction 
is not there?

Robert, code, including gui's,  created by other code and then executed, 
is a great feature of Rebol.

Robert: here is a proposition for presenting View + VID + REBOL coding:


1) Show this code on a slide, explain briefly what it does and run 
it: 
REBOL []

view layout [
	code: area 300x200
	btn "Run" [view/new layout load code/text]
]


2) Enter some VID code in the textarea, click Run and listen to the 
"wow" in the audience ;-)

You can type, for example:

across
text "Name:" f: field
btn "Say" [alert join "Hello " f/text]

BTW, you might want to patch the 'button VID style to point to 'btn 
for this demo. "button" would be much better for the code demo, but 
the look of "btn" is way nicer (and more modern).

but one thing you need to highlight, is that for all of your needs 
(code, style, layout, server, client, templates, controls, localization, 
etc) you are using a single highly natural language.    This is the 
power of Dialects, (DSL). 


datatypes are also a great differentiator of REBOL... just run code 
and leave an HTML tag in the code, as-is... that is cool.  most languages 
require literal data to be assigned, but in rebol data IS an expression 
so as long as there are no syntax issues, you can just embed data 
within the code (for whatever reason you need it in source managing 
tools).

I have Dialects using VID and Doc's example to show: data is code, 
code is data.

show that you also use the same language for creating style sheets, 
new controls, sending a VID layout over tcp, etc.  just showing how 
small the code snippets are is enough I guess.


Make a little parallel between the web nonsense, and the simplicity 
of doing the same thing in rebol. (5-7 different languages, vs one)

I think I will save the last 40s to even include a Q&A section into 
my pecha kucha :-) because  Rebol is just that compact and simple, 
that I don't need 20 slides ;-)

That too, since 'lib and 'sys are predefined in the lib and sys contexts 
for convenience.

Wow, there are three languages less popular
 [than REBOL, on Github].


I wouldn't worry about those stats too much at the moment. It seems 
that projects are re-indexed when there is commit activity within 
the project. So some REBOL projects on Github are not yet correctly 
classified and counted.


On the other hand, the same probably holds true for other recently 
added languages.

that is the exact reason I am coding Liquid in C and not in C++. 
 my project will build on anything out there, or can be easily converted 
to any other language.   furthermore, C allows me to do more advanced 
OOP tricks than C++ allows... things like run-time class mutation 
(required for liquid)

Sorry, @maxvessi : works great under Windows 7 Opera and Firefox 
!

@Steeve: me too, I have to create it and it works !

it works under Windows XP, Opera and Firefox
Can the DLLs be installed automaticly ?

@oldes: Facebook started to freeze groups, during tha next week "worthy 
groups" will be updated to the new style/feature, "unworthy groups" 
will be archived. In both cases all users will be deleted and users 
wil have to join again to the group!
P.S. I sent you a message trough Facebook

It's like that, I'm an administrator, but not the creator, so I ndot' 
see any button! Is a creator push the button, you loose all users, 
BUT you maintain postos, phots and videos... It's stupid! How can 
you invite again thousands of users?

I pushed the button and it's updated probably... FB is stupid as 
you can have many groups with the same name. I visit it just a few 
times pre year.

Oldes: It seems to have worked, I got an email saying that it's been 
upgraded, and I'm still a member of it

Simply, I put trust  on google to keep my personal data (and not 
to sell them) because I know they make their money without this urge.

Facebook on the contrary never stoped to ly about the privacy topic. 
Year after year, they experimented new way to exploit your credulity 
to sell your personal data.

From the beginning I felt this trap will arrive , tha"t why I never 
entered in that Facebook mania.

my biggest complaints are the lack of groups and also some silly 
UI bugs. If I choose to follow a popular person and he posts something 
with 500 responses, those 500 responses cannot be collapsed. I have 
to scroll through them.

for sure their interface will improve, But I think Google has a much 
better record for making things better in time... 


IMHO, Facebook hasn't really changed at all in years.  Its all cosmetics, 
and even then I can't really tell you what its improved.  I've still 
got a useless stream of hundreds of posts a day which I can't *easily* 
manage.  configuring **anything** in FB is tedious to say the least. 
  


And the idea of "annoy everyone you know" and "give your personal 
data to unknown companies, without my conscent" is not something 
I readily enjoy. 

even friend lists are limited in length!   basically, the whole UI 
is a disaster.

I'm sure google will do a lot of things better by integrating with 
other google sites like google docs, maps, gmail,  android-things, 
youtube, news etc.

facebook cannot integrate with them easily and they have a loooong 
way to build similar stuff.

Here is a poll result from techtree.com for the question: Will you 
leave Facebook to join Google+?
http://www.techtree.com/techtree/jsp/pollResult.jsp?poll_id=633

I downloaded the archive before the retraction and it does have very 
recent source files in it ...

Thx. This testing part is absolut fundamental for us. We are constantly 
extending it and will add reports so we see what tests are there, 
what has been coverered etc.

Robert and company. Thank you so much.

you are welcome... I just added rectangle-pack script and included 
ImageMagick dependencies which would be needed in some other scripts 
which I use.

Will this language be an open or closed source project?


Question has been noted. I'll collect questions and make a QA here 
in AltME around release time. The countdown might answer questions 
and/or inspire people for new questions.

And of course, very typical question - what was the main montivation, 
aka - why yet another language? :-)

Well, OTOH it is cool to have some options. RED and Topaz are nice 
projects, we will see, what World is all about. Maybe those projects 
can be somehow complementary, who knows ...

It's good to have more options, especially if they have different 
approaches. Topaz is completely different and usable. Red will have 
compile-to-native, fast, run on ARM / Pic etc., R2 is the only one 
stable-enough to use in production in general.

Only Geomol knows the main goal of his project though I woud guess 
it is not disimilar to that of Boron, Red and Topaz.

Interesting anyway, because Boron is a more classical REBOL (architecture 
wise, and well, just my wild guess), Topaz is very different aproach, 
and RED takes different aproach too. But maybe it was in a making 
even before Topaz and Red were announced, who knows. And maybe it 
is just a hobby project :-)

I don't think that World is a hobby project, it is like Boron, Red 
and Topaz which I believe are being developed to meet the authors 
want for a tool that meets their specific needs.

And was REBOL itself so different?

Steeve - and where's your REBOL clone? Nowadays it seems all REBOL 
gurus are coming with one. Should I expect Ladislav and Cyphre enter 
the game? :-)

Sure, but I would like first to have some of you guys confirm, that 
the release can be downloaded and works.

no binaries in therer? Should I download sources, and build?

The .dat files are the binaries. Rename one to "world" and make it 
executable.

It would save a step and some thought.

Pity Carl resumes his hobby, and still leaves whole community without 
a single word of explanation for 13 months already:


http://mendoradio.wordpress.com/2011/11/20/resuming-the-mendoradio-blog/

?? prints test 3 and now is returned evaluated after so its printed 
by the console

this is true of all lit-word arguments and is probably why we don't 
use them often, they brake the visible chain of command.

yep.. usually a better alternative... and put a prin "test3: " on 
the previous line.

I generally only use ? and ?? for words when I want to know the type 
as well, and because of that very issue I started using this:
probe: func [D][
    print join join type? D "!  " mold :D
    D
]

It's in my primary include file.  Helps with 'none versus none!, 
etc...

can never remember which copy and which inline

maybe I do want exclude and then unqiue

I tried yours and it didn't seem to work... but this definitely does:

c: 0
blk:  [ 1 2 3 1 2 3 1 4 ]

unique remove-each i copy blk [ c: c + 1 not find (at blk c + 1) 
i]

since set2 is a unique version of set1, it will always find and always 
remove

I've long considered + and - on blocks to be useful... but, would 
others agree?

though I don't mind if we use 'subtract and 'add

next you'll be asking for the ability to create combinations and 
permutations!

but its so usefull in simplifying many other loops and algorithms.

when managing data sets, you often want to know what is unique "within" 
that set.   

duplicates and singletons are often-required things you want to know 
about a data set.

Henrik, I don't think that's nice, because you can't really translate 
"File" and "cannot be found" separately. Don't expect all languages 
to have the same grammar structure etc.

Anyone got a suggestion on how to flatten the results of a SQL query.
Got [ [ "xxx" ] [ "yyy" ] [ "zzz" ]]
and want 
[ "xxx" "yyy" "zzzY ]

sure I could iterate over the whole lot and create a new block ...

Ah... I'm using 2.7.6 and they're not there

well,  I will use the to-block form and if that errors out drop back 
to flatten

cause its going to be A LOT faster and for things like database queries, 
this can become quite a bottleneck.

I want pretty much *everythihng* to be a mezzanine until we know 
the gains are worthwhile and will benefit enough people. That still 
gives us room to be subjective. ;-)

On the balance of things, I think it would make things better; code 
more readable and more consistent with R3 code, so I'm for it.
A better name for CC is definitely needed, though.

Well... I think we should still aim high and make a wish.

If MAKE object! specs catch RETURN and EXIT, that's an error.