AltME groups: search

If by fine you mean that you first have to go to the target paste 
area (for example a browser address bar), clear it, then to AltME 
and copy the text, and then back to the paste area and use the middle 
mouse button to paste. Any deviation from this procedure will break 
it in weird ways

I tried to create a zip archive, but it is about the same size, so 
no help. IIRC there was some AltME version, which allowed to use 
bigger sizes? I think it would be good to have it as a command-line 
option for the server startup. It is my private world, and I would 
allow 20MB easily. It is not first time I got file bigger than 10MB 
I needed to post ....

Petr, it's a big security hole, because the AltME server uses a multitude 
of the file size in memory during the upload, which can easily crash 
the server and at least overloads the swap space

Quite a while back I was confused about the clipboard in Kubuntu. 
I investigated heavily until I understood what was happening, that 
there were two clipboards and how they are used. I think I did a 
small bit of tweaking to Kubuntu settings, but basically it was just 
learning how it worked and practising a bit that cured my confusion. 
I got used to it and don't have much trouble with it in other apps. 
AltME is the exception - it just doesn't copy and paste consistently 
for some reason.

AltME works, as long as you remember it goes to the X clipboard. 
 Gtk/etc applications access it with highlight/middle click, while 
AltME uses Ctrl+C/Ctrl+V.

The other linux clipboard (can't remember what supplies it) is accessed 
by Ctrl+C/Ctrl+V in Gtk/etc applications.  This one, AltME and Rebol 
can't access.

though in qtask, I discovered, that if you switch window and come 
back to altme, the copy may work the first time.


the focal support in R2 sucks.  its just weirdly implemented and 
it creates many bugs, especially since most faces share their para 
object by default.


I've added this to my style init  in many new styles:  [para: make 
para []]

Welcome to the freedom of being reborn. Forget your past and start 
fresh. I'm just a few weeks old now myself.

The link is working...

But it is written correctly in blue with hyperlink only http://www.digicamsoft.com/cgi-bin/rebelBB.cgi?thread=%3C26Nov20101203072
and to the end of url it is in black...

for example, the host-kit and extensions group are the best reference 
for someone wanting to get up to speed.  it has discussions and a 
few web links on further on-line resources which some of us have 
posted.

I'm just amazed the program and data all fit in 52.5mb

ah just need to drag a shortcut, should be easy enough.  This system 
is being formatted thats all and I didn't want to redownload all 
content again

I am using altme from time to time from my USB stick (a bit slow, 
due probably to slow flash disk plus the way how altme relies on 
heavy file access, but still usable), and the only thing I need is 
to just copy it ...

me too, I am so used to having to try and export stuff and then import 
in a new install

It's the X11 clipboard hell. There are two of them, and AltME uses 
the one where you have to use middle mouse click to paste in X apps

Also, the order is critical. You have to empty the target area in 
the X app, then cut in AltME, then go back to the X app and use middle 
click to paste

Steeve, I'm a little confused, you say it "lost" something.  That 
should not be the case.

Just unstall AltME, and then reinstall it, it "should" grab a fresh 
copy of everything.


There is a bug that sometimes means you have to post something in 
a group to get it to kick start.  But in general, this should work.

It will take a few minutes to get everything, but you should be all 
fresh and happy once it does.

I'm connected on a restricted wifi spot. The connections on ports 
other than 80 are slow like Hell and indue lot of reco/deco (timeout 
frequently reached in Altme).

Due to that, reloading everything in altme is the worst experience.

I wouldn't be "all fresh and happy" - I'd be a little bit grumpy 
about it.
But I'm interested in how the old messages were lost.
Did the power outage occur when AltME was writing files?

Perhaps we could place all the world files in a Git repo and run 
a cron job to update the repo regularly.

I missed that horror.

With a slow connection, when datasets are large and take too much 
time to reload, Altme performs an automatic deco/reco and all the 
current loading is lost.
Probably because it fails to ping the server during the load.

Entering in endless attemps to reload and reload the same dataset 
(not sure though).

It's an OLD issue, I already notified that in the past. several times.

but I always had trouble making myself trust by people who do have 
not that problem.

I missed that horror.

With a slow connection, when datasets are large and take too much 
time to reload, Altme performs an automatic deco/reco and all the 
current loading is lost.
Probably because it fails to ping the server during the load.

Entering in endless attemps to reload and reload the same dataset.

It's an OLD issue, I already notified that in the past. several times.

but I always had trouble making myself trust by people who do have 
not that problem.

What would be useful isd for RT (or whoever currently holds the development 
rights for AltME) to publish the API that syncs the local copy with 
the server.

Then the better programmers among us could write a resync/repair 
tool. Run that routinely before starting AltME, and a number of issues 
(like the . problem) may be solved.

flush out the altme is great altme has the right too forget things 
and start blank :)

I would make the case that anyone wanting to write a book about the 
emergence of R3 from an engineering standpoint might want  access 
to an altme archive ... or has that ship also sailed ?  Also quite 
open/transparent:  IO language IRC exchanges and the evolution of 
the Falcon language at falconpl.org with their maillist.  But Carl 
is talking about getting rid of the maillist for Rebol - when archived 
maillists are often so useful in hindsight. ...

footnote: the Griswold papers on SNOBOL and ICON ended up in the 
library at UMN - his collaborators may never have thought of posterity 
taking an interest.

Only when they're used. Many of the people in the community don't 
really use email anymore (including Carl and me).

Why are special chars like: ä ö ü ß totally screwed up when posted 
from OSX and read from a Windows version an vice versa? Is there 
a solution to this?

I think they're each interpreting text according to their own native 
character set. Windows in UTF-16, OS X and Linux probably in UTF-8. 
AltME doesn't compensate

Yes, AltME simply ignores character encoding.It simply regurgitates 
the text it receives from the client. I believe that REBOL/View uses 
the default codepage under Windows(not utf-16) and MacRoman under 
OS/X. I suspect it uses ISO-8859-1 under Linux but am not sure.

That's probably because you're using Windows and the people who wrote 
the messages are also using Windows.

in the last year, I've been noticing that the message copying worked 
somehow and there was a pattern... I just coudn't figure it out.

then, today I did 11 copy/pastes in a row and was typing at the same 
time (and they all worked!)...

that's when I realized that going from typing to copying was the 
secret!

Symptoms: if you've recently re-installed AltME, then the program 
runs continuously, endlessly resyncing that group.


This happened before, and the workaround was to copy over a good 
copy of the 439.set file (thanks to Peter Wood for the good copy). 
But that no longer seems to work.


I am going to delete the group and set up a new one with the same 
name. The archive  of past postings will remain available here:
    http://www.rebol.org/aga-display-posts.r?post=r3wp439x1

Is anyone using AltMe from linux (like a recent Ubuntu)? I moved 
to linux for 90% of things but I can't run altme there, and rebol 
view crashes also. I didn't look into depth of why..

altme 1.2.25 works ok with debian 6 and ubuntu 10.10 for me. fonts 
are so-so, but at least usable

Been using it on Ubuntu 8.04 and now 10.04 (AltMe versions 1.1.29 
and currently 1.2.21)

What drives me nuts with AltME is, that when it's reconnecting, that 
it switches display and don't just stay at the group and message 
where I am. It resets the list etc.

Possible bug in Altme: I connect to world rebol3 then click on world 
name on the top-left corner that opens a combobox to select other 
worlds which I connected previously, but non-existing worlds anymore. 
When I choose one of them Altme tries to connect and fails, then 
when I click on Return to World button altme get closed immediately 
without showing the message "Disconnecting".

hmm... I just experienced AltME in the Rebol 3 world with a memory 
leak and eating CPU. It was upward of 150 MB RAM usage, and then 
I killed the process.

Got a few error message in the terminal session that AltME always 
opens (and leaves open after quitting.):


altme(6254,0xa038c620) malloc: *** mmap(size=3021340672) failed (error 
code=12)
*** error: can't allocate region
*** set a breakpoint in 
malloc_error_break to debug

Thanks Doc -- I had not noticed that Red is flagged [web-public].

Theoretically, [web-public] gets you:

....automatically published on REBOL.net (though that appears to 
no longer be updated):
       http://www.rebol.net/altweb/rebol3/index.html

....and published on REBOL.org at the discretion of the Library Team. 
       http://www.rebol.org/aga-groups-index.r?world=r3wp


As Graham suggests there may be some issues, especially as REBOL.org 
is funded by RT. I'll ask the Library team to discuss.

In another AltME world I'm participating in, AltME file sharing is 
used to share 103 files, all of which I have downloaded. Every time 
I (re)connect, AltME offers me to re-download 63 of them. Irrespective 
of whether I re-download or not, I'm again offered this each and 
every time I connect.

I faintly recall having heard of this problem before, anyone has 
a clue what's going on (and more importantly: how to fix it?). I'm 
using AltME 1.2.25 on Linux.

Once a year, go through and redownload every file in the world. And 
be careful to not repost those files, which is hard to do because 
of UI problems.

I deleted and re-downloaded all files several times already, so I 
guess it's a different problem.

and also - there might be one bug I noticed and reported. When you 
start first world on 5400, and the other one on 5401, altme.com registers 
the second one on 5400 too. I overcame it by disallowing port 5400 
on my firewall for the time of registration of the 5401 world, or 
it would connect users to my first world.

Alan, does your firewall allow the other port to be reached? I've 
never had much problems switching ports. Altme.com should detect 
the port change when the world is restarted, but it needs to be able 
to contact altme.com and the world shouldn't be run double on the 
old port

Kaj, I'm pretty sure I opened the whole range for altme 5400-5409 
and I tried a port that I know is open which I had another world 
running on.

amacleod - when starting second world, e.g. on port 5401, during 
the time of setting-up the world, please prohibit altme.com from 
accessing another port on already opened 5400 port. That is imo a 
bug I found, and it did not get fixed yet. So disable 5400 port temporarily 
on port 5400. Maybe it will be enough if you turn-off the world running 
on 5400 port, because I got into situation, when Altme.com pointed 
two different worlds to the same 5400 port ...

anyone knows whats this? i had altme open on my intel mac for a few 
weeks and this is what it was spitting onto the terminal. there was 
interesting glicth happening at the same time: my clicks were interpreted 
with some offset.

Ah yes, Syllable has a similar log format, and the program name mentioned 
in front is often not the place the message is originating from

@  Reichart : We were discussing in the OPEN GL group about the default 
way AltMe permits to enter text - and we were 3 that think the default 
way to enter text would be the EDIT mode (the one that accepts Line 
feeds without submitting the post automatically. 


Would you be considering this option in a future release - even if 
I now realize that this is a global switch since when I submits new 
posts this mode is kept alive.


I didn't ask for others to vote for this change but you can do it 
if you want - I find it  really annopying - more than the obligation 
to commit messages by having to use the mouse each time. I know it's 
a productivity consideration but as I can't reedit my sent posts, 
sometimes they didn't look good - until I asked someone how to enter 
LFs.

Thanks a lot for considering this request. 
Regards, Gerard

Gerard, I'm not invovled in AltME anymore.... 

I like your idea, and I wish AltME was open source (which I have 
pushed hard for).

well, the alternative is to build a functionally identical clone 
and post all sources on github. I'm sure there are a couple in here, 
who could take a stab at it.

its just a pity that all the server stuff is sorted out and works 
pretty much

And we can build it on top of 0MQ.

What data exchange format are you talking about Max?

The thing is, it doesn't scale. Way to slow for big groups. IMO it 
must handle 100.000 posts with instant realtime search and filterting.

So google stole the altme idea put it on steroids and called it google 
plus!

I uderstand it, that Reichard wanna know, what is able to "handle 
100.000 posts with instant realtime search and filterting."

179,000+ posts and real time search:
   http://www.rebol.org/aga-index.r

Yes, should go into AltMe. And much, much faster display of large 
groups, without message limit. Virtual tables is the trick...

altme 1.2.25 is invisible in the application circling in gnome3 (alt+ 
tab) and font are very small ...

yeah even with it my eyes are on fire because it's too small ... 
and when my eyes hurts I'm not in a good mood for anything less reading 
tons of pages of those tiny fonts texts

main problem is that  the fonts in R2 scale isn't accurate ... we 
discussed it or at least try to when we  writed area-tc and all we 
had as reply is don´t worry r3 is so much better and that was 2 years 
ago ...

yeah and who will implement you dictator wish henrik to hide your 
lack of result and your lack of balls to talk to me ?

And I don't have the time to write one myself :(

Hmm, I guess, that some of the administrators here enforce a convention, 
some may not, and that is why there is a "mix".

That's not our decision. Users choose which method is best for communication 
and so far, AltME holds up pretty well. Users are free to use open 
forums like rebolforum.com or the Synapse forum.

I started adding the ! and ~ to help denote products that were REBOL, 
and offtopic stuff "~", etc.  

Sort on Alpha, anf it becomes very clear.

I added links to Boron and World to "Links" under checklists.

Changed Boron to the website home, and added an extra link for this 
source code repository

Official 0.4.2 available on AltME site works OK on gNewSense and 
auto-upgrades to 1.2.25

The beta links on the site are dangerous and should have been removed 
years ago

Sorry not sure if this is the right place to ask. I'm overwhelmed 
with some many forum/place to post, and visit. I want to find a copy 
of Excel Dialect, but it's no longer available at http://www.robertmuench.de/download/, 
could you help me to locate a copy of it? Thanks a lot!

To whom it may concern: I guess that it might make sense to define 
a new REBOL-SYNTAX group discussing REBOL syntax, and its description 
at

    https://github.com/rebolsource/rebol-syntax

Not wanting to be particularly picky, but could someone please rename 
to "REBOL syntax" group to "REBOL Syntax". Trying to be consistent 
with "title case" used in other group titles .... Thanks in advance 
and thanks a lot :)

So, we do have a single point of failure here...and no recovery plan

We've been bitten by closed source Rebol and we rely on closed source 
altme :(

and threads

Giuseppe - yes, we have moved, as this world went down, and was down 
fore few days, with no sign, when it will be back online ...

In fact - he took last 100 ppl logging here or so, and sent them 
an invite.

Giuseppe - I can see you have an account there, and that you never 
logged in. So - check your email  ...

Thanks and sorry for the recall - I didn't see it  when it passed 
...

Per a friendly suggestion, I have cleaned up documentation on my 
Twitter client:

	http://www.ross-gill.com/page/Twitter_API_and_REBOL


Please, if you have any difficulties/suggestions/issues (or heaven 
forbid, bugs) let me know.

And we should collect specifications for what we need for http/s 
web protocols for R3 as well.

I think this is good for generic discussion, and I'll make it web 
public.

Graham, I agree. We need all the sticks, pokers, and examples we 
can get to make sure we don't hit roadblocks with protocols.

hi Chris (RG). I tried your twitter example, but I can't make it 
work. It waits for a long time and then throws an timeout error.

is there a way to remove re-tweets from showing up in our timeline 
view?   I don't give a damn about 99% of all the re-tweets I'm getting 
and it completely fills up the timeline  so that I actually don't 
see any posts from the people I'm following... its totally stupid... 


to me twitter is the worst conversation design model I've ever seen. 
 in just about all aspects.  you just can't understand any thread 
and its pretty useless as I see it so far.

I don't use if for conversation, but I do track quie a number of 
interests through it - I use TweetDeck and have set up many columns 
with various persistent searches. I really don't see why you need 
to "follow" anyone. As a stream of pointers to blog posts, artices, 
videos, etc on what interests me, it's been quite good - esp since 
TweetDeck (and Twitter, I suppose) supports complex searches so that 
you can have one search that will find most relevant tweets.

Twitter is cool in the simple case where ONE wants to broadcase a 
message to many.  Any "warning" system is a perfect match.

Things that update people of something, esp. like servers telling 
a team (think Nagios), this is all pretty cool.


But, it would be better to have one system, tags, and then let the 
UI decide what to do with it all.


So Email, Chat, Twitter, even RSS are one "system", and something 
in it tells the UI what to do with it.

I follow a lot of RSS feeds and I find that Twitter is a system that 
is pretty good at leading me to find things to follow in more depth, 
RSS being one such mechanism. I guess it's the noise approach that 
actually brings value for me - if something should be looked at, 
others that share your interests will likely pick up on it too and 
the "signal" strength will increase for that subject. 


Gabriele, I'm just curious how you, in particular, determine which 
information out there is "good". Twitter isn't a medium that qualifies 
the information it publishes - you investigate the links just as 
you would with Google search results - only Twitter is more up-to-date. 
Reichart: I'd like a unified system too, but it's not there yet.

A friend of mine built a very cool system that weighed RSS feeds, 
sort of like how Netflix weighs movie ratings.  It was real time. 
 But no one understood how cool it was.

So for example, you might want all feeds that mention REBOL.  You 
set it to grab 100%, in so doing, and since I know you, if I even 
ask for 5% of my news to be REBOL, it would give me a little more 
since others that I know really like "REBOL" related stuff.  

http://jaanix.com/
 
Play with it.  He works for Skype now.

Adrian, the same "others will share it" works with RSS, it's not 
necessary to have a single company "owning" the messages.


But, if all you do is share links and look at the most popular ones, 
then isn't reddit a much better choice?

The fact is that there are quite a few ways to stay on top of what 
others consider worthy of interest, but, in the end, what brings 
the most value to these systems is how widespread their use is. Statistics 
become more valid when the sample size is increased and ia as diverse 
as possible. Right now, Twitter is the most ubiquitous such system, 
so it's where I look.

Twitter module for REBOL by Chris RG works great !
http://www.ross-gill.com/page/Twitter_API_and_REBOL

I'm a little confused about the reference to a secret key and the 
like:  "Before you start, you’ll need an Consumer API Key and Secret. 
 If you don’t already have these, you can register an application 
at Twitter’s Developers Site"  I went to the site, but didn't have 
a clue as to what I should do there.  Any suggestions or links for 
me to investigate?  Thanks.

Follow the tutorial : http://www.ross-gill.com/page/Twitter_API_and_REBOL

For me, I've create RebTweet application. Maybe you can create P61Twitter 
for exemple. No matter. You just need to declare your application 
in order to get Consumer API Key and Secrect.

Can someone help me.  I'm having trouble using the code:

settings:	[

        ; Source from:	http://www.ross-gill.com/page/Twitter_API_and_REBOL

        ; prerequisite:	https://dev.twitter.com/apps/new(one time for below)
        c-key:		#VPC....MLyQ		; C-Key
        c-secret:		#jYK4....POjI		; C-Secret
        u-store:		%/MyApps/TwitterUsers.r
        twitter:		http://reb4me/r/twitter.r
    ]

do settings			; assigns parms

do/args twitter context [
	Consumer-Key:	c-key
	Consumer-Secret:	c-secret
	User-Store:		u-store
	]

I get the following:

connecting to: reb4me
** Access Error: Cannot connect to reb4me
** Where: open-proto
** Near: do/args twitter context [
    Consumer-Key: c-key
    Consumer-Secret: c-secret
    User-Store: u-store
]

I must be missing a step or something.