Discussion of E and other capability languages ()

headers Sean W. | 9 Nov 03:32 · /bin/

_______________________________________________
e-lang mailing list
e-lang@...
http://www.eros-os.org/mailman/listinfo/e-lang

_______________________________________________
e-lang mailing list
e-lang@...
http://www.eros-os.org/mailman/listinfo/e-lang

Hello Folks,

Our office is moving, so as part of the move, the MediaWiki server
wiki.erights.org will be offline for 1 to 2 days later this week.

The current plan is to shut down everything Thursday at noon, CDT.

The new IP address is planned to be 50.77.162.161, but this won't take
effect until Friday at the earliest.

I'll send out a reminder and more details as they develop.

As usual, please discuss any issues by sending an email to this list.
Now would be a good time to grab a copy of the current backup file...
just in case.

James

Continuing with our experimental syntax changes to improve the 
reliability of E, here's a patch that adds a "for-must-match" feature.

http://gitorious.org/~tal-itinnov/repo-roscidus/it-innovation/commit/3ad4f5c3072447647c617fdb56738bf929cb8a1a

(see also: http://wiki.erights.org/wiki/Surprise_list)

Normally in E, this silently skips the third item:

   for a:int in [1, 2, "hi"] {
     println(a)
   }

Now, it throws an exception:

   pragma.enable("for-must-match")

   for a:int in [1, 2, "hi"] {
     println(a)
   }

To make a loop that skips non-matching items as before, use the "match" 
keyword. e.g.

   pragma.enable("for-must-match")

   for match a:int in [1, 2, "hi"] {
     println(a)
   }

In our code-base of > 20,000 lines of E, I only found two places making 
use of the old behaviour (or at least, all the tests passed after I 
updated those two .

In contrast, there have been many occasions where the old behaviour 
caused problems. Most commonly, someone is looping over some tuples:

   for [foo, bar] in tuples { ... }

and someone changes the format to add an extra field. Since this doesn't 
cause any errors, it's hard to detect that anything is wrong.

--

-- 
Dr Thomas Leonard
IT Innovation Centre
Gamma House, Enterprise Road,
Southampton SO16 7NS, UK

tel: +44 23 8059 8866

mailto:tal@...
http://www.it-innovation.soton.ac.uk/

People often try to use return inside when blocks. Currently, this just 
produces confusing "Ejector must be enabled" errors.

How about requiring the use of "return" to return a value from a when block?

There are a few places where this would be useful:

- Often you will move a whole block in to or out of a when() block. It's 
useful if you don't have to change all return lines when you do this.

- People sometimes don't notice that the value of the last expression is 
used, and put something after it.

e.g. they change

return when (foo) -> {
	foo.invoke()
}

to

return when (foo) -> {
	foo.invoke()
	traceln("here")
}

and then wonder why they're getting null pointer exceptions elsewhere.

- You can accidentally return a value, when all you intended was to tell 
the caller when you were done.

Patch to add #pragma.enable("easy-when-binds-return"):

http://gitorious.org/~tal-itinnov/repo-roscidus/it-innovation/commit/c10e7bc99bbfbabbbb28ea714e19855fbca8c389

--

-- 
Dr Thomas Leonard
IT Innovation Centre
Gamma House, Enterprise Road,
Southampton SO16 7NS, UK

tel: +44 23 8059 8866

mailto:tal@...
http://www.it-innovation.soton.ac.uk/

We've got quite a few methods that take or return maps. e.g.

to purchase(customerDetails :Map[String,any]) {
   ... customerDetails["name"] ...
   ...
   if (problem) {
     notify(customerDetails["email"])
   }
   ...
}

The problem is that it's not obvious from the method signature what keys 
need to be present. Often someone fails to include a mapping, and the 
system only fails later in some case where it needed that value.

What I think I'd like to do is something like:

def CustomerDetails := makeNamedTuple([
	"name" => String,
	"email" => String,
])

...

to purchase(customerDetails :CustomerDetails]) {
   ... customerDetails["name"] ...
}

So that:

- if the caller fails to provide "name" and "email" then it fails the guard
- if I try to access a field I didn't define (e.g. "fax") then I get an 
error, whether or not the client provided it, so I'm forced to keep the 
interface up-to-date

Ideally it should handle fields being added later (e.g. if two parties 
are running different versions of the software it should ignore fields 
it doesn't understand).

Any ideas on the best way to do this? Maps? TermTrees?

Some of the other systems using this are written in Java, so ideally it 
should be easy to use from there too.

--

-- 
Dr Thomas Leonard
IT Innovation Centre
Gamma House, Enterprise Road,
Southampton SO16 7NS, UK

tel: +44 23 8059 8866

mailto:tal@...
http://www.it-innovation.soton.ac.uk/

JSON allows "1e+1" as well as "1e1":

http://www.json.org/

However, E's JSON parser rejects it:

? def surgeon := <elib:serial.deJSONKit>.makeSurgeon()
? surgeon.unserialize("1e+1")
# value: 10.0

new syntax error: Missing exponent
#   1e+1
#   ^^

This patch fixes it:

http://gitorious.org/~tal-itinnov/repo-roscidus/it-innovation/commit/a2bf52beff24defc63bd8fe5cf1f8cfe85547c23

Note: a side-effect is that it also allows this syntax in regular E 
code. I assume that's not a problem?

--

-- 
Dr Thomas Leonard
IT Innovation Centre
Gamma House, Enterprise Road,
Southampton SO16 7NS, UK

tel: +44 23 8059 8866

mailto:tal@...
http://www.it-innovation.soton.ac.uk/

There's quite a lot of complexity (and slowness) in the E build for 
producing platform-specific releases (e.g. E-win32-x86-0.9.3a.zip). Is 
anybody using these?

The build process seems to be:

- src/bin/* contains various SWT binaries (presumably out-of-date by 
now; last update was Feb 2008)
- just the binaries for the building platform get copied to dist/bin/$osdir
- dist gets copied to export/dist
- export/dist gets copied to export/puredist
- export/puredist/bin/$osdir is deleted

How would people feel about having a single E release (just puredist), 
without any jars? This means that people writing SWT applications would 
be responsible for distributing swt.jar with their application somehow 
(e.g. using a Maven pom.xml, a 0install feed, or bundling a copy).

However, you already have to do that anyway if:

- You're producing non-Windows releases
- You're producing 64-bit Windows releases
- You're producing 32-bit Windows releases using an up-to-date SWT
- You want to use other libraries

We could make the rune "run" command depend on SWT in the 0install feed 
so that the SWT example scripts continue to work without extra effort 
(for people using 0install). That would avoid making every E program 
depend on SWT (e.g. headless servers). People not using 0install or a 
distribution package manager would need to install SWT for their 
platform manually.

--

-- 
Dr Thomas Leonard
IT Innovation Centre
Gamma House, Enterprise Road,
Southampton SO16 7NS, UK

tel: +44 23 8059 8866

mailto:tal@...
http://www.it-innovation.soton.ac.uk/

One problem I frequently hit with updoc is that it doesn't report errors 
until the script finishes. But, often, the error prevents the script 
from finishing. Then you have to bisect the script manually until you 
see the error.

As a simple test case, consider:

   ? def a
   ? bind a = 3
   ? interp.waitAtTop(a)
   ? a
   # value: 3

This script (which contains a syntax error on the second line), will 
silently hang:

$ rune demo.updoc

updoc: started

/tmp/demo.updoc:..
[ hangs ]

With this patch, it reports the error:

$ rune demo.updoc

updoc: started

/tmp/demo.updoc:...
at: <file:/tmp/demo.updoc#:span::3:4::3:13>
expr: bind a = 3
missing original syntax error
new syntax error: use ':=' for assignment, or '==' for equality

Another side-effect of this is that if the abortEarly flag is set by a 
test case, the tests do actually abort.

Patch:

http://gitorious.org/~tal-itinnov/repo-roscidus/it-innovation/commit/dca612f2a472451881c00ea02e290a398a92b22e

--

-- 
Dr Thomas Leonard
IT Innovation Centre
Gamma House, Enterprise Road,
Southampton SO16 7NS, UK

tel: +44 23 8059 8866

mailto:tal@...
http://www.it-innovation.soton.ac.uk/

Simple patch to fix E error reporting:

http://gitorious.org/~tal-itinnov/repo-roscidus/it-innovation/commit/c9a58031350e65d0b721f9b865dba821327703ba

Without the patch exceptions get reported twice:

$ cat test.e
foo
$ rune test.e
# problem: Failed: Undefined variable: foo
#
#   @ </home/tal/tmp/test.e#:span::1:0::1:2>
#   - EExpr#evalToPair(Scope)
#   . e`foo`.evalToPair(<a Scope>)
#   @ evalToPair/1: 
<jar:file:/opt/src/e/dist/e.jar!/org/erights/e/elang/cmd/cmdMakerMaker.emaker#:span::128:57::128:66>

# problem: Failed: Undefined variable: foo
#
#   @ </home/tal/tmp/test.e#:span::1:0::1:2>
#   - EExpr#evalToPair(Scope)
#   . e`foo`.evalToPair(<a Scope>)
#   @ evalToPair/1: 
<jar:file:/opt/src/e/dist/e.jar!/org/erights/e/elang/cmd/cmdMakerMaker.emaker#:span::128:57::128:66>
#   - static Ref#fulfillment(Object)
#   . <makeRef>.fulfillment(<ref broken by problem: Failed: Undefined 
variable: foo>)
#   @ fulfillment/1

--

-- 
Dr Thomas Leonard
IT Innovation Centre
Gamma House, Enterprise Road,
Southampton SO16 7NS, UK

tel: +44 23 8059 8866

mailto:tal@...
http://www.it-innovation.soton.ac.uk/

_______________________________________________
e-lang mailing list
e-lang@...
http://www.eros-os.org/mailman/listinfo/e-lang

3	December 2011
8	November 2011
13	October 2011
16	September 2011
7	August 2011
9	July 2011
17	June 2011
10	May 2011
63	April 2011
9	March 2011
33	February 2011
36	January 2011
27	December 2010
15	November 2010
32	October 2010
13	September 2010
6	August 2010
13	July 2010
20	June 2010
36	May 2010
34	April 2010
40	March 2010
43	February 2010
50	January 2010
46	December 2009
7	November 2009
23	October 2009
54	September 2009
30	August 2009
33	July 2009
10	June 2009
39	May 2009
45	April 2009
71	March 2009
39	February 2009
26	January 2009
4	December 2008
16	November 2008
19	October 2008
47	September 2008
58	August 2008
42	July 2008
26	June 2008
44	May 2008
16	April 2008
167	March 2008
68	February 2008
38	January 2008
24	December 2007
52	November 2007
45	October 2007
62	September 2007
55	August 2007
30	July 2007
38	June 2007
49	May 2007
71	April 2007
73	March 2007
60	February 2007
99	January 2007
105	December 2006
81	November 2006
30	October 2006
23	September 2006
58	August 2006
91	July 2006
56	June 2006
121	May 2006
21	April 2006
31	March 2006
70	February 2006
18	January 2006
13	December 2005
13	November 2005
1	October 2005
19	September 2005
90	August 2005
102	July 2005
124	June 2005
93	May 2005
49	April 2005
32	March 2005
44	February 2005
121	January 2005
194	December 2004
30	November 2004
33	October 2004
47	September 2004
101	August 2004
65	July 2004
18	June 2004
24	May 2004
127	April 2004
76	March 2004
114	February 2004
89	January 2004
100	December 2003
93	November 2003
18	October 2003

Invitation to connect on LinkedIn

Lecture series on Web security by Mark Miller (Google)

wiki.erights.org Downtime and IP address change

for-must-match

easy-when-binds-return

Interfaces / data types in E

Patch to the lexer to accept "1e+1"

Remove the binary distributions?

Patch for review: show updoc errors immediately

Patch for review: don't report errors twice

Anybody read Ukrainian?

Mon	Tue	Wed	Thu	Fri	Sat	Sun

		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