On Wednesday, October 27, 2021 at 3:34:29 PM UTC+13, alanfe...@gmail.com wrote:
> So Emacs lets you put in key definitions in a readable file and you don't need to
> compile it into a huge section file and thereafter not be able to list your commands?
> If it does, that's cool!

It’s all just a bunch of ELISP definitions. Your custom functions and commands can have docstrings as well, to integrate with the built-in help system.

Here is my current collection of public customizations: <https://github.com/ldo/emacs-prefs>.

> I've used emacs on occasion. A lot of two-keys-at-the-same-time commands. And often
> two of _them_ in a row. And the 4,5 bit bothers you on EDT? Here's an emacs key
> definition with three in a row!: C-x 4 C-o

Less common commands take more keystrokes, more common ones (like the movement ones I mentioned) typically take fewer.

> And another problem I always had with emacs: Once you start a session for the first time
> in a long time, you can't figure out how to exit. Well, I'd always forget it after a while and
> have to look it up. You'd think it would be a single key, like Ctrl-X or something, but
> NOOOOOOO. It's C-x C-c ...

I tried for a while, binding it to a single custom keystroke. I removed that after I hit it by accident once too often.

The thing with Emacs is, just keep it running all the time. And also turn on its server mode. I have the following command defined in Bash:

    alias e='emacsclient -n'

So in any terminal session, I can just type “e «filename»” to have that file open immediately in the always-available Emacs window.

> Hey, if emacs works for you, that's great. Go for it.

There are depths to it I haven’t explored yet. For example, the text-attribute and overlay features (there seems to be some overlap between the two). One of the things I have done with overlays is being able to collapse (hide) selected regions of a file.

On Wednesday, October 27, 2021 at 12:36:33 AM UTC-4, Phillip Helbig (undress to reply) wrote:
> In article <ba4e4aeb-b8c0-44b7...@googlegroups.com>,
> "alanfe...@gmail.com" <alanfe...@gmail.com> writes:
>
> > I don't have any fields, but I _do_ have electric lights!
> As Maxwell said, I have an electric field.

He had two -- zzzzt -- two --zzzzt -- two fields in one!

Some of you younger folk might not get the reference.

On Monday, 25 October 2021 at 08:59:32 UTC+1, Jan-Erik Söderholm wrote:
> Den 2021-10-25 kl. 06:28, skrev alanfe...@gmail.com:
> > On Sunday, October 24, 2021 at 12:22:44 PM UTC-4, Stephen Hoffman wrote:
> >> On 2021-10-22 06:45:02 +0000, alanfe...@gmail.com said:
> >>
> >>> This used to be broken. Strange things would happen under certain
> >>> conditions, like only the directory being changed or the directory
> >>> being hidden in sys$disk. It appears to be fixed on EISNER. Is there
> >>> documentation of this fix somewhere? I've been searching on the web,
> >>> only to come up empty. TIA! (^_^(
> >> *You rang?*
> >
> > Yep! That was the best part of the show, no? When Lurch would suddenly turn into the scene seemingly out of nowhere and say, "You rang?"
> > [...]
> >>
> >> Logical names are volatile key-value store design and straight out of
> >> ~1984, and long since and widely used and misused.
> >>
> >> For device redirection, they're adequate.
> >>
> >> Logical names have had "surprising" edge cases ~forever, too.
> >>
> >> The whole of the default directory stuff cited is just utterly
> >> hilarious, but that design is just too entrenched to ever be fixed and
> >> the existing stuff deprecated and removed.
> >
> > But Hoff, it _was_ fixed! Looks like no one knows when or what version of VMS it was fixed in.
> >
> > Here's a bonus question. When was EDT updated to allow more than 22 lines? IDK, but it's made me happy!
> There as an EDT patch just a year ago or so from VSI.
> Had something with numnber of lines to do, I think...
> >
> > Bonus No. 2: In what version of VMS was "Proactive Reclamation of Memory from Idle Processes" introduced? This one I found on my own. Turns out it was v5.4-3. Made a big difference in my main system at the time!
> >>
> > [...]
> >> --
> >> Pure Personal Opinion | HoffmanLabs LLC
> > AEF
> >
Damn! What a pity I missed out on the EDT update! (I retired 18 months ago)
EDT was my favourite editor by a very long way.

On Wednesday, October 27, 2021 at 6:38:37 AM UTC-4, dhay...@gmail.com wrote:
> EDT was my favourite editor by a very long way.

30+ years ago, I wrote ~200 lines of TPU to customize the EVE global section to be
a blend of EVE and EDT interfaces. I had to tweak it in 1989 for the V5 TPU, but
since have used the exact same TPU$SECTION file ever since.

On 10/26/2021 6:46 PM, Lawrence D’Oliveiro wrote:
> On Wednesday, October 27, 2021 at 2:41:29 AM UTC+13, Arne Vajhøj wrote:
>> The last Windows server version that required a unique drive letter per
>> disk was NT 4.0 Server (released 1996, end of support 2002/2004). All
>> newer Windows versions (2003, 2008, 2008R2, 2012, 2012R2, 2016, 2019
>> and 2022) supports volume mount points.
>
> Which are too fiddly for ordinary people to use,

Likely.

But then I do not expect "ordinary people" to be responsible for
"entrust mission-critical business functions to" systems.

And if they are then I would not blame MS for the consequences.

> can cause incompatibilities with some software,

Invisible to software.

> are tricky to set up for things like USB sticks

It may be a problem to stuff 30 USB sticks in a Windows
server.

But believe me that is not a problem for "entrust mission-critical
business functions to" systems.

Arne

On 10/27/2021 8:25 AM, David Jones wrote:
> On Wednesday, October 27, 2021 at 6:38:37 AM UTC-4, dhay...@gmail.com wrote:
>> EDT was my favourite editor by a very long way.
>
> 30+ years ago, I wrote ~200 lines of TPU to customize the EVE global section to be
> a blend of EVE and EDT interfaces. I had to tweak it in 1989 for the V5 TPU, but
> since have used the exact same TPU$SECTION file ever since.

Very similar experience here.

Hacked the entire EVE code in 4.4 in 87. Rewrote to build
on top of EVE code in 5.5 in 89.

A bit more lines as I wanted it to be more CDC NOS FSE like.

Arne

On 2021-10-27, Arne Vajhøj <arne@vajhoej.dk> wrote:
> On 10/27/2021 8:25 AM, David Jones wrote:
>> On Wednesday, October 27, 2021 at 6:38:37 AM UTC-4, dhay...@gmail.com wrote:
>>> EDT was my favourite editor by a very long way.
>>
>> 30+ years ago, I wrote ~200 lines of TPU to customize the EVE global section to be
>> a blend of EVE and EDT interfaces. I had to tweak it in 1989 for the V5 TPU, but
>> since have used the exact same TPU$SECTION file ever since.
>
> Very similar experience here.
>
> Hacked the entire EVE code in 4.4 in 87. Rewrote to build
> on top of EVE code in 5.5 in 89.
>
> A bit more lines as I wanted it to be more CDC NOS FSE like.
>

Ok, at this point, does anyone want to confess to writing an Edlin
clone in TPU ? :-)

Simon.

--
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.

On 10/27/2021 9:44 AM, Arne Vajhøj wrote:
> Very similar experience here.
>
> Hacked the entire EVE code in 4.4 in 87. Rewrote to build
> on top of EVE code in 5.5 in 89.

5.0 of course - not 5.5.

Arne

On Thursday, October 28, 2021 at 2:30:00 AM UTC+13, Arne Vajhøj wrote:
>
> On 10/26/2021 6:46 PM, Lawrence D’Oliveiro wrote:
>>
>> On Wednesday, October 27, 2021 at 2:41:29 AM UTC+13, Arne Vajhøj wrote:
>>>
>>> The last Windows server version that required a unique drive letter per
>>> disk was NT 4.0 Server (released 1996, end of support 2002/2004). All
>>> newer Windows versions (2003, 2008, 2008R2, 2012, 2012R2, 2016, 2019
>>> and 2022) supports volume mount points.
>>
>> Which are too fiddly for ordinary people to use,
>
> Likely.
>
> But then I do not expect "ordinary people" to be responsible for
> "entrust mission-critical business functions to" systems.

How about a system where the hot-plug mount points are set up automatically as part of system defaults?

>> can cause incompatibilities with some software,
>
> Invisible to software.

You might think so. But then consider proprietary software installers that check for available space at the top level of the C: drive, and refuse to proceed if there’s no room there, even if they are actually going into another mount point.

>> are tricky to set up for things like USB sticks
>
> It may be a problem to stuff 30 USB sticks in a Windows
> server.

It may be a problem anything business-critical on a Windows server going forward, at least the non-cloud version, given that Microsoft is winding back its development efforts on that.

Another thing is that the mount-point feature is specific to NTFS. Unlike Linux, which has a generic VFS layer which handles things like that, for some reason Microsoft tied up important features like this with one particular filesystem, so it won’t work with anything else.

On Monday, October 25, 2021 at 9:25:10 PM UTC-4, alanfe...@gmail.com wrote:
> On Monday, October 25, 2021 at 3:47:23 PM UTC-4, Dave Froble wrote:
> > On 10/25/2021 2:28 PM, alanfe...@gmail.com wrote:
> >
> > > But it does a lot more than that. SET DEFAULT had 2 or 3 silly bugs that I figured wouldn't take long to fix.
> > What bugs? I've never seen any.
> I posted on earlier in the thread. I think it's even mentioned in the docs once, but I don't recall where and doubt I can find it.
>
> Here is my summary of them:
> The DCL program SET DEFAULT has two problems with nested logical
> names:
>
>
> 1.) If the first translation has a trailing colon and there is no
> explicit directory-spec in the second translation, SYS$DISK is changed
> to the 1st translation and the directory portion is not changed. The
> "actual" current directory remains hidden in SYS$DISK.
> [ I assume this mean if you specify <LNM>:[directory] ]
> 2.) If the 1st translation has no trailing colon, then only the
> directory portion of the default is changed! This can leave SYS$DISK
> incorrect thereby leaving the default incorrect.
> [This is the example posted by someone long ago which I put in one of my first posts in this thread.]
> Additionally, in certain circumstances, SET DEFAULT returns an error
> but still changes your process to a bad default.
> [Unfortunately, I don't recall any examples of this.]
> > --
> > David Froble Tel: 724-529-0450
> > Dave Froble Enterprises, Inc. E-Mail: da...@tsoft-inc.com
> > DFE Ultralights, Inc.
> > 170 Grimplin Road
> > Vanderbilt, PA 15486
> AEF

The main problem happens when your equivalence name is a logical name and you don't include the colon. For example:

$ DEFINE HOME SYS$LOGIN

is potentially troublesome. The better way is

$ DEFINE HOME SYS$LOGIN:

I found that if you always include the colon as above, you don't get into trouble.
Still, there was some rare case where you get a fatal error but it set your default anyway. Unfortunately, I don't recall what that was. If I stumble upon it later I will post it here.

On Monday, October 25, 2021 at 12:28:47 AM UTC-4, alanfe...@gmail.com wrote:
[...]
>
> Here's a bonus question. When was EDT updated to allow more than 22 lines? IDK, but it's made me happy!

Found it! It's

vms842l2a_edt-v0100.release_notes

The info about the SET DEFAULT bugs and word that it would be fixed in a later version of VMS is in

vmsu2055_release_notes

Still looking for the release notes where SET DEFAULT and SHOW DEFAULT were fixed.

>
[...]
> AEF

On Thursday, October 28, 2021 at 2:30:00 AM UTC+13, Arne Vajhøj wrote:
> But believe me that is not a problem for "entrust mission-critical
> business functions to" systems.

There are many mission-critical business functions running on Linux servers making use of storage technologies like SANs and LVM, which can lead to a proliferation of mounted volumes. Also remember that, on Linux, the visibility of mounted filesystems can be segregated into “filesystem namespaces”, such that a process can only see those filesystems within its assigned filesystem namespace.

This is a key part of container technologies, like Docker. Microsoft even tried to support Docker under Windows for a while, but it seems to have given up. Wonder why ...

On 10/31/2021 5:05 PM, Lawrence D’Oliveiro wrote:
> This is a key part of container technologies, like Docker. Microsoft
> even tried to support Docker under Windows for a while, but it seems
> to have given up. Wonder why ...

MS still support docker. On Windows Server 2016, 2019 and 2022.

Arne

On Monday, November 1, 2021 at 10:34:07 AM UTC+13, Arne Vajhøj wrote:
>
> On 10/31/2021 5:05 PM, Lawrence D’Oliveiro wrote:
>>
>> This is a key part of container technologies, like Docker. Microsoft
>> even tried to support Docker under Windows for a while, but it seems
>> to have given up. Wonder why ...
>
> MS still support docker. On Windows Server 2016, 2019 and 2022.

For certain values of “support” <https://www.theregister.com/2019/06/18/the_future_of_docker_on_windows_is_linux_says_docker/>:

    At Microsoft's Build conference last month, Gabe Monroy, lead program
    manager for the Azure Container Compute team, was asked whether
    Windows Containers are for legacy and Linux Containers for new
    projects. "I think that is a fair description," he said ...

As for Windows Server 2022 <https://www.theregister.com/2021/07/28/windows_server_2022_sac/>:

    At a high level, it seems that Microsoft is investing in Azure and
    AKS, and that on-premises Windows Server users are in some respects
    being left behind. It is notable, for example, that Active Directory's
    functional level has not changed since Windows Server 2016, and this
    remains the case in Server 2022.

...

    SMB over QUIC, which promises secure access to Windows file shares
    from anywhere, is another key feature which will be available only in
    Windows Server Azure Edition.

On 10/31/2021 6:22 PM, Lawrence D’Oliveiro wrote:
> On Monday, November 1, 2021 at 10:34:07 AM UTC+13, Arne Vajhøj wrote:
>>
>> On 10/31/2021 5:05 PM, Lawrence D’Oliveiro wrote:
>>>
>>> This is a key part of container technologies, like Docker. Microsoft
>>> even tried to support Docker under Windows for a while, but it seems
>>> to have given up. Wonder why ...
>>
>> MS still support docker. On Windows Server 2016, 2019 and 2022.
>
> For certain values of “support” <https://www.theregister.com/2019/06/18/the_future_of_docker_on_windows_is_linux_says_docker/>:
>
>     At Microsoft's Build conference last month, Gabe Monroy, lead program
>     manager for the Azure Container Compute team, was asked whether
>     Windows Containers are for legacy and Linux Containers for new
>     projects. "I think that is a fair description," he said ...

Here is free mini course in IT terminology:
* supported means that the vendor accept bug reports and create fixes
* legacy means that it works but is not seen as a good choice
for new stuff

Being legacy does not imply not supported.

Arne

On 10/31/2021 6:22 PM, Lawrence D’Oliveiro wrote:
> As for Windows Server 2022 <https://www.theregister.com/2021/07/28/windows_server_2022_sac/>:
>
>     At a high level, it seems that Microsoft is investing in Azure and
>     AKS, and that on-premises Windows Server users are in some respects
>     being left behind. It is notable, for example, that Active Directory's
>     functional level has not changed since Windows Server 2016, and this
>     remains the case in Server 2022.
>
> ...
>
>     SMB over QUIC, which promises secure access to Windows file shares
>     from anywhere, is another key feature which will be available only in
>     Windows Server Azure Edition.

I suspect that both TheRegister writer and you did not really
understand the content of that article.

MS offers 3 different options:
* Azure software hosted by MS
* Azure software hosted on-premise
* traditional stuff hosted on-premise

Thew new features goes into the Azure software. But that
does not mean no support for on-premise. It just means that
on-premise will have to function cloud-like.

Arne

On Monday, November 1, 2021 at 12:15:55 PM UTC+13, Arne Vajhøj wrote:
> Here is free mini course in IT terminology:
> * supported means that the vendor accept bug reports and create fixes

When was the last time you got Microsoft to accept a bug report and issue a fix?

On Monday, November 1, 2021 at 12:24:29 PM UTC+13, Arne Vajhøj wrote:
> MS offers 3 different options:
> * Azure software hosted by MS
> * Azure software hosted on-premise
> * traditional stuff hosted on-premise
>
> Thew new features goes into the Azure software. But that
> does not mean no support for on-premise. It just means that
> on-premise will have to function cloud-like.

Those extra features are simply not available in the third option.

On 11/1/2021 1:55 AM, Lawrence D’Oliveiro wrote:
> On Monday, November 1, 2021 at 12:15:55 PM UTC+13, Arne Vajhøj wrote:
>> Here is free mini course in IT terminology:
>> * supported means that the vendor accept bug reports and create fixes
>
> When was the last time you got Microsoft to accept a bug report and issue a fix?

I have not.

But other people have.

Some people experience a problem and report it and MS fix it.

Some white hats find a vulnerability and send it in and MS fix it.

Some black hats find and exploit a vulnerability and MS fix it.

That is what MS does. What IBM does. What Redhat does. What Oracle
does. What various open source projects do.

Arne

On Tuesday, November 2, 2021 at 4:31:53 AM UTC+13, Arne Vajhøj wrote:
>
> On 11/1/2021 1:55 AM, Lawrence D’Oliveiro wrote:
>>
>> On Monday, November 1, 2021 at 12:15:55 PM UTC+13, Arne Vajhøj wrote:
>>>
>>> Here is free mini course in IT terminology:
>>> * supported means that the vendor accept bug reports and create fixes
>>
>> When was the last time you got Microsoft to accept a bug report and issue a fix?
>
> I have not.
>
> But other people have.

Any of them with Docker for Windows?

Because if not, then by your own definition of “support”, you have not demonstrated that Microsoft is offering “support” for Docker for Windows.

On Tuesday, November 2, 2021 at 9:20:54 AM UTC+13, John Dallman wrote:
> I work for a large company ... which pays a lot for Microsoft support.

Is that on top of the actual licence fees for Windows itself?

Just to make clear what ordinary folks are (not) getting for the money they pay to Microsoft ...

On Wednesday, October 27, 2021 at 12:32:09 AM UTC-4, Phillip Helbig (undress to reply) wrote:
> In article <4016c182-65dc-421c...@googlegroups.com>,
> =?UTF-8?Q?Lawrence_D'Oliveiro?= <lawren...@gmail.com> writes:
>
> > What drove me mad about EDT was its insistence on a \_current
> > direction_/ mode.
>
> It's not a bug, it's a feature. (By the way, your 8-bit characters have
> been substituted with 7-bit replacements to be more usenet-friendly, via
> an EDT macro.)
>
> But you can have what you want! Say you want to SEARCH for a string:
>
> PF1 PF3 <string> ENTER
>
> is probably what you're doing. But
>
> PF1 PF3 <string> KP4
>
> is "find in forward direction"
>
> and
>
> PF1 PF3 <string> KP5
>
> is "find in reverse direction.
>
> ENTER is, as usual, "FIND NEXT" in the current direction.

Why Press Enter? Why not press PF3. That way your finger is already in place to find the next one!

>
> So you can change (or confirm) the current direction at the same time as
> you tell it to search, all with one keystroke. That's what I call
> efficiency.

AEF

On Monday, October 25, 2021 at 3:47:23 PM UTC-4, Dave Froble wrote:
> On 10/25/2021 2:28 PM, alanfe...@gmail.com wrote:
>
> > But it does a lot more than that. SET DEFAULT had 2 or 3 silly bugs that I figured wouldn't take long to fix.

> What bugs? I've never seen any.

Here are some SET DEFAULT bugs:

From vmsu2055_release_notes.txt

1.1.5 SET DEFAULT and SHOW DEFAULT Commands-Behavior Changed

V5.5-2

In VMS Version 5.5-2, the DCL command SET DEFAULT does not
set the default correctly when you use a logical name that
has multiple translations. For example:

$ SET DEFAULT DISK1:[DIRECTORY1]
$ DEFINE myfile myfile1
$ DEFINE myfile1 DISK2:[DIRECTORY2]
$ SET DEFAULT myfile
$ SHOW DEFAULT
DISK1:[DIRECTORY2]

In this example, the default should read:

DISK2:[DIRECTORY2]

This problem will be corrected in a future release of VMS.

Currently, DCL does not show the right default device in
the SHOW DEFAULT command when you use a logical name that
has multiple translations. For example:

$ SET DEFAULT DISK1:[DIRECTORY]
$ DEFINE myfile myfile1
$ DEFINE myfile1 DISK2:[DIRECTORY]
$ SET DEFAULT myfile: ! Must be with a colon
$ SHOW DEFAULT
myfile:[DIRECTORY]
$ WRITE SYS$OUTPUT F$ENVIRONMENT("DEFAULT")
myfile:[DIRECTORY]

In this example, "myfile" is used as the default
device. The correct default directory display should be
DISK2:[DIRECTORY].

This SHOW DEFAULT behavior will be changed in a future
release of VMS.
======================================================================
The bug in the first example has been fixed. I don't know when except
that it was probably 7.0 or later.

The second one is still broken on OpenVMS V8.4-2L2.

Regardless, you won't encounter these problems if you always use
TO.COM to change your default. (TO.COM v5.3 will be released soon!)
======================================================================
======================================================================
OLD:

The DCL program SET DEFAULT has two problems with nested logical
names:

1.) If the first translation has a trailing colon and there is no
explicit directory-spec in the second translation, SYS$DISK is changed
to the 1st translation and the directory portion is not changed. The
"actual" current directory remains hidden in SYS$DISK.

The above is still broken as of OpenVMS V8.4-2L2.

Got new details on this. SET DEF keeps translating until it gets to a
logical name with a trailing colon. Then SYS$DISK is changed to that
logical name, leaving the directory unchanged; unless it contains a
device and directory, in which case you get just that: the device and
the directory upon running SHOW DEFAULT. Note that if you include a
trailing colon in the SET DEFAULT command, the logical name will be
assigned to SYS$DISK, just as if it were a search list.

Example:

$ SHOW DEF
DSA3:[DECUSERVE_USER.FELDMAN.TO-V530]
$ SHOW LOG A
"A" = "B:" (LNM$PROCESS_TABLE)
1 "B" = "C:" (LNM$PROCESS_TABLE)
2 "C" = "D:" (LNM$PROCESS_TABLE)
3 "D" = "SYS$LOGIN:" (LNM$PROCESS_TABLE)
4 "SYS$LOGIN" = "DISK_USER:[DECUSERVE_USER.FELDMAN]" (LNM$JOB_831D6780)
$ SET DEF A
$ SHOW DEF
B:[DECUSERVE_USER.FELDMAN.TO-V530]
$ SET DEF D:
$ SH DEF
D:[DECUSERVE_USER.FELDMAN.TO-V530]
$ SET DEF D
$ SH DEF
DISK_USER:[DECUSERVE_USER.FELDMAN]
$

2.) If the 1st translation has no trailing colon, then only the
directory portion of the default is changed! This can leave SYS$DISK
incorrect thereby leaving the default incorrect.

The above, no 2., has been fixed. I don't know when, though.

The above, no 2., has been fixed. I don't know when, though.

Additionally, in certain circumstances, SET DEFAULT returns an error
but still changes your process to a bad default. But I haven't been
able to reproduce this on EISNER (OpenVMS V8.4-2L2). Maybe it's been
fixed.

UPDATE on 2022/10/16

It's still broken:

$ SH DEF
DISK_USER:[DECUSERVE_USER.FELDMAN]
$ SET DEF DSA2:DSA3
%RMS-F-DIR, error in directory name
$ SH DEF
DSA2:[DECUSERVE_USER.FELDMAN]
%DCL-I-INVDEF, DSA2:[DECUSERVE_USER.FELDMAN] does not exist
$

As you can see, SET DEFAULT produced a fatal error, yet still changed
the default. Yeah, it's unlikely you'll ever type something like that.
Maybe there's another type of typo that causes this.

Anyway,

TO.COM fixes all of the above problems and does much more. See the
other .TXT files for additional information.

TO.COM v5.3 to be released soon!

> --
> David Froble Tel: 724-529-0450
> Dave Froble Enterprises, Inc. E-Mail: da...@tsoft-inc.com
> DFE Ultralights, Inc.
> 170 Grimplin Road
> Vanderbilt, PA 15486

Alan Feldman