[Fwd: Re: [dyalogusers] Re: November meeting of BAA-London]
Phil Last
Posted from DyalogUsers in case anyone here doesn't look there.
-------- Original Message --------
Subject: Re: [dyalogusers] Re: November meeting of BAA-London
Date: Mon, 23 Nov 2009 00:48:36 -0800 (PST)
From: Dan Baronet <dbar...@yahoo.com>
Reply-To: dyalo...@yahoogroups.com
To: dyalo...@yahoogroups.com
References: <hbujt...@eGroups.com> <4AF7D1DF...@ntlworld.com>
<4B0A47C2...@ntlworld.com>
Phil, someone asked me last time I was there if I would talk about the
new features of 12.1.
Donno if it is still interesting but I could do that too.
/D
Phil Last
Posted from DyalogUsers in case anyone here doesn't look there.
-------- Original Message --------
Subject: RE: [dyalogusers] Re: November meeting of BAA-London
Date: Mon, 23 Nov 2009 08:54:03 -0000
From: Michael Hughes <Mic...@Hughes.uk.com>
Reply-To: dyalo...@yahoogroups.com
To: <dyalo...@yahoogroups.com>
References: <hbujt...@eGroups.com> <4AF7D1DF...@ntlworld.com>
<4B0A47C2...@ntlworld.com>
I will be there and so yes please J
Phil Last
Attending were:
Chris Hogan
Dan Baronet
Dave
Dick Bowman
Ellis Morgan
Jake Jacob
Kai Jaeger
Mike Hughes
Nicolas Delcros
Stephen Taylor
Phil Last
Dan, supported by Nicolas, gave us a nice intro to Dyalog's version 12.1
enhancements and some hints at future developments.
Chris reported on developments in the phrasebook project.
Kai showed us into some of the more advanced and arcane features of the
APL wiki and, I suppose, moin moin wikis in general.
There was no enthusiasm for a December meeting so Stephen suggested we
join the IPSA crowd again at their Christmas Reunion at the Plumbers
Arms in Pimlico on 18 December.
Phil
Dick Bowman
rubbing their noses in the bacchanalian pleasures they have missed? And
making them want to come next time?
>
> You received this message because you are subscribed to the Google Groups "BAA London" group.
> To post to this group, send email to baa-l...@googlegroups.com.
> To unsubscribe from this group, send email to baa-london+...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/baa-london?hl=en.
>
>
>
>
Phil Last
> Should we ... ?
Will you?
Shall I?
Dick Bowman
Phil Last wrote:
Will you? Shall I? -- You received this message because you are subscribed to the Google Groups "BAA London" group. To post to this group, send email to baa-l...@googlegroups.com. To unsubscribe from this group, send email to baa-london+...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/baa-london?hl=en.
John Jacob
right!
Phil Last
topic, at least insofar as it has wandered, and not wanting
unnecessarily to multiply threads.
Nevertheless.
The discussion at the meeting at one point moved to talking about
unicode encoding schemes and Nicolas' answer that there is no
deterministic way to ascertain whether an input file is encoded in UTF-8
or ascii "plain text" assuming, unfoundedly, that it must be one or other.
I have this decision to make in reading an "ini" file. So here's my
probabilistic(*) solution. (*) see below
Here's what I did:
⎕IO
0
⍝ largest integer in mapping for each char?
{⌈/'UTF-8'⎕UCS ⍵}¨⎕UCS⍳129
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74
75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98
99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
117 118 119 120 121 122 123 124 125 126 127 194
⍝ are any less than predecessor?
∨/2>/{⌈/'UTF-8'⎕UCS ⍵}¨⎕UCS⍳10000
0
These two statements show, and prove beyond doubt if it were in
question, at least for the first 10000 code-points, that:
1. below 128, UTF-8 is a straight through mapping identical to ascii.
2. each single character maps to integers at least one of which is at
least as large as the largest of its immediate predecessor.
3. all characters beyond 127 map to at least one integer above 193.
4. so all UTF-8 encodings will either be identical to untranslated ascii
or they will contain at least one item above 193
5. so if there are no integers more than 193 we can assume it's ascii.
6. this leaves those other 60 odd bytes in question.
7. fortunately the UTF-8 mappings fall into ranges such that certain
combinations of bytes including those 60 odd are impossible.
So here we go:
UTF8orNot←{
{
⍵↓⍨(1⍴⍵)=⎕UCS 65279 ⍝ remove byte order mark if present
}⍵{ ⍝ input text
∧/⍵<194:⍺ ⍝ Return text if can't be UTF-8
11::⍺ ⍝ Return text if invalid UTF-8
'UTF-8'⎕UCS ⍵ ⍝ decode integers to text
}⎕UCS ⍵ ⍝ convert to integers
⍝ assume either ascii or UTF-8
}
Unfortunately most of the values from 194 onwards (see bottom line) are
the "European" characters with diacritics. Any use of them in an
otherwise plain text file will cause the function to attempt the
decoding although if the file were not actually UTF-8 encoded it is
almost(*) certain to trigger the error guard and return the untranslated
text.
(*) probabilistic; almost. I haven't come across a counter example yet.
There are better(?) ways to do this by examining character frequencies.
⎕UCS 4 64⍴⍳256
!"#$%&'()*+,-./0123456789:;<=>?
@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~
¡¢£¤¥¦§¨©ª«¬®¯°±²³´µ¶·¸¹º»¼½¾¿
ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ
Phil Last
at that meeting regarding December, the postings here following the
meeting or any combination of the foregoing - in case you don't know:
There was no enthusiasm for a December meeting so Stephen suggested we
join the IPSA crowd again at their Christmas Reunion at the Plumbers
Arms in Pimlico on 18 December.
That's tomorrow.
Not yet confirmed at the venue but our next meeting will almost
certainly be on Friday 30 January 2010 at the Templar.
Agenda to follow. November was one of our best attended yet. Let's hope
we can do even better in January.
Happy Solstice & New Year to all.
Phil
Phil Last
of January; not as stated then, the 30th but the real last Friday
which is the 29th.
The venue is booked: The Knights Templar in Chancery Lane 14.00
downstairs for 14.30 up.
AAsk
"101 reasons why APL is not on life support"
i.e. 101 'snippet' applications/idioms illustrating how APL fit in the
current technology framework.
Could make a great Vector article aiming to market APL.
Dick Bowman
which contains only positive words - as in my contributions to the APL
Wiki "Ten Good Things About APL" and "A Positive Outlook For APL".
Dick Bowman
Ajay Askoolum has proposed his "why APL is not on life support" - aside
from two less-than-enthusiastic reactions to the implied negativity I've
seen no further response or other suggestions.
Before the November meeting I sent out a missive regarding APL2010 - but
we didn't discuss it. What I've done in the interim is send the
following to the APL2010 organisers. Whether there's anything in it
that could trigger our discussion, or anyone deciding to take any of it
further, I don't know. But here's what I sent...
> I apologise in advance if this does not seem well though-out, but I have
> a suggestion.
>
> I think that it's a very positive thing to see the return of a
> conference which is centred on APL rather than specific vendor products.
>
> For a number of years I was an enthusiastic conference-goer, but I
> stopped for a number of reasons. I've been to a couple of Dyalog's in
> recent years, but my old frustrations are emerging.
>
> Specifically, that the format doesn't really live up to what I'd like to
> get from a "conference" - it seems to centre too much on formal
> presentations and doesn't encourage participation. I come to think more
> and more that "presentations" aren't the right format for the state of
> the APL world as it sits today.
>
> We do see a certain amount of to-and-fro discussions on places like
> comp.lang.apl, but that stuff always seems prone to dissolve in the air
> - almost as it's written.
>
> What I'm wondering about is having the conference include a serious
> discussion forum - something that can make people say they were glad to
> make the effort to attend. I'd like to see something where every
> attendee is encouraged to contribute.
>
> A specific topic I'd propose is "The Future for APL" - specifically
> aimed at getting people who use APL to talk about the direction they'd
> like to see APL take over the next 5 or 10 years (maybe more). To offer
> a counterpart to having us consume whatever the vendors decide to
> offer us.
>
> Some of the stuff I'd like to see...
>
> A pre-conference submission from attendees (and maybe people who won't
> attend, but still feel they have something to say.
> Moderator-led discussions on significant sub-topics.
> Recording of discussions.
> Post-conference publication of key topics.
>
> As for what some of the sub-topics might be...
>
> What's next after object-orientation?
> Does APL need data types and structures beyond what we have now?
> Should APL consciously mimic other programming languages and their
> development methodologies/environments?
> What are the prospects for vendor-independant language extensions?
> Can we ever see an end to archaisms like bracket-indexing?
> Why have we never been able to build software libraries?
>
> That's enough for now.
I'd happily see this replaced my another more focussed topic.
Michael
Brilliant idea.
__________ Information from ESET Smart Security, version of virus signature database 4775 (20100115) __________
The message was checked by ESET Smart Security.
__________ Information from ESET Smart Security, version of virus signature database 4775 (20100115) __________
The message was checked by ESET Smart Security.
__________ Information from ESET Smart Security, version of virus signature database 4775 (20100115) __________
The message was checked by ESET Smart Security.
Michael
-----Original Message-----
From: baa-l...@googlegroups.com [mailto:baa-l...@googlegroups.com] On Behalf Of Dick Bowman
Sent: 15 January 2010 17:27
To: baa-l...@googlegroups.com
Subject: Re: January meeting of BAA-London
__________ Information from ESET Smart Security, version of virus signature database 4775 (20100115) __________
Phil Last
> Unless of course I can Skype into someone's portable - I just can't get to London and back.
We discussed this early on but haven't done anything about it.
There's a Skype scuppering setting somewhere in my machine and I'm
waiting for my fairy godmother to come and fix it.
But with 7.5Mbps mobile broadband and mains electricity it should be
possible for someone to have a go.
Michael Hughes (Blackberry)
From: Phil Last <phil...@ntlworld.com>
Date: Fri, 15 Jan 2010 19:24:43
To: <baa-l...@googlegroups.com>
Subject: Re: January meeting of BAA-London
John Jacob
a view to doing this.
I also tried video capture at one session but the results were iffy. I
need a wide angle camera for that and my external mike wasn't playing.
Michael Hughes (Blackberry)
Dick Bowman
Is anyone still interested in the Wikipedia pages? I ask because I look
from time to time. Sometimes you learn something new - today I found
out that ↑ is a monadic operator that returns its argument in ascending
order.
Or shall we just leave CyberWally and his chums to maintain their drivel
as they see fit?
Not sure whether I'll be there - I do hope any Skyping can be worked out
in advance.
John Jacob
> Not sure whether I'll be there - I do hope any Skyping can be worked
> out in advance.
the day and Mike Hughes is aiming to join us in spirit if not body.
You're welcome to give it a go as well. No guarantee it'll work of course.
I plan to use the skype ID baa.london to host a conference session (max
5) so anyone wishing to set-up a contact with that ID in advance ought
to be able to join in as well.
Again no promises but until we try it for real we won't know.
PGilbert
You can find my original solution to detect the encoding of a file
(with or without BOM) using .Net in the method 'EncodingDetector' of
the class 'DIO' that I have posted here:
http://microapl.freeforums.org/disk-and-file-input-output-utility-class-written-using-net-t119.html
Regards,
Pierre Gilbert
Phil Last
It seems that the on-line option to edit the Subject (which I've used a
couple of times lately) only puts the new subject above that actual
posting rather than the entire thread, which I suppose is reasonable in
the circumstances.
What it means is that in the Discussions you see the posting under its
old subject; the new subject just doesn't appear at all.
Whereas if you're signed up for email postings you get the mail under
the new.
The lesson for me is not to post new notices as continuations of the
previous month's. In fact not to post on-line at all.
Anyway.
The meeting is on Friday as announced under "January meeting ..." if you
are receiving individual emails or under "November meeting ..." if you
are on-line.
Phil
Jake
request a contact with the skype ID baa.london.