Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Re: Ramdisk

107 views
Skip to first unread message

Shadow

unread,
Dec 20, 2019, 6:22:30 PM12/20/19
to
On Fri, 20 Dec 2019 17:54:40 -0500, "Mayayana"
<maya...@invalid.nospam> wrote:

>I'm doing this on XP but that group seems to be rather quiet
>these days.
>
> Trying to set up ramdisk swap file on 32 bit. Have
>16 GB RAM. Gavotte version didn't seem to work. ImDisk
>worked fine. But there's a catch. Virtual memory settings
>show that I set up 3 GB virtual memory on the ramdisk,
>drive R, but there's now a pagefile.sys on C drive that's
>taking up 3 GB. Settings show no swap file for C drive.

Have you tried measuring speed? CrystalDiskMark shows quite
impressive numbers for RAMDisks. If it's a faux RAMDisk, numbers will
be much lower.
>
> What does this mean? Did it fail? Or is C:\pagefile.sys
>just a quirk that can't be avoided when setting up a
>ramdisk swap file?

I have the same problem. Can't use anything above the 4GB
limit.
Currently using the last free version of SoftPerfect RAM
Disk(v 3.4.8) for my browser profile folder. It's very fast, saves the
image at shutdown, pretty stable (1 crash in 6 months, but I keep
backups).
I don't use it for a swapfile because I rarely need more than
1GB swap on XP.
If you ever find something that works above the "limit",
please post.
[]'s

PS X-posted to the XP group. Just in case someone there has a
working setup.
--
Don't be evil - Google 2004
We have a new policy - Google 2012

Paul

unread,
Dec 20, 2019, 8:00:11 PM12/20/19
to
Dataram RAMDisk Driver
V4.0.4.2
Dataram RAMDiskVE Driver
Dataram, Inc.

RAMDiskXP.sys 62,464 bytes

Dataram_RAMDisk_v4_0_5_RC0.msi 5,567,488 bytes
SHA1: 95F123AF724206B9C8DD61DD76BF388098FE9B54

AFAIK, 4GB max size, for free. Even if the file versions
don't match, it's still a fine addition to a WinXP system
that has too much RAM.

*******

I have a single reference to that file in my Sent folder.
It goes as follows.

Dataram_RAMDisk_v4_0_5_RC0.msi

5,567,488 bytes
MD5 = 1c9c709c647d979b1277f6d756dce265
SHA1 = 95f123af724206b9c8dd61dd76bf388098fe9b54

Try here.

http://filehippo.com/download_ramdisk/tech/14475/

[Actually, try here. Took me a while to find this! Immediate download...]
https://filehippo.com/download_ramdisk/14475

The one "RAMDisk 4.0.5 RC0 (Windows) http://ramdisk.en.uptodown.com"
is corrupted. Looks like RAID array damage, 256KB
chunks of zeros in the file. So don't use that, until
someone reports it and they fix it.

Always check the checksums. The SHA1 is safer, if one is
available. If I start using SHA256, there would be more
whining from people attempting to check. You can always upload
a file to virustotal and they can calculate a SHA256 for you,
but the upload on virustotal is "pure flaky pastry".

So what you do with that, is install it, use the Configuration
Utility, then go to the Advanced tab to see your "freebie"
license limit and how much RAM the machine has.

https://i.postimg.cc/W1ZbJKjV/Advanced-tab.gif

As far as I know, if you purchase a license, it applies
to specific versions. My purchased license is for a later
version that doesn't do PAE. Which is fine, because that
machine has a better setup anyway. The above version
is specifically to get "more than your moneys worth"
from WinXP :-)

HTH,
Paul

Mayayana

unread,
Dec 20, 2019, 9:46:17 PM12/20/19
to
"Paul" <nos...@needed.invalid> wrote

|
| http://filehippo.com/download_ramdisk/tech/14475/
|
| [Actually, try here. Took me a while to find this! Immediate
download...]
| https://filehippo.com/download_ramdisk/14475
|

Thanks. That seems to have worked. What a lot
of confusion! That's the first option I've found that
actually has instructions. ImDisk and Gavotte provided
no instructions at all. Dataram provides full instruction,
recognized the installed RAM, and specifically provided
an option to put the ramdisk above 4 GB.
Gavotte also didn't work but did bluescreen a couple of
times before I got it installed, only to see no ramdisk
and no explanation. ImDisk worked, showed a ramdisk,
but didn't know enough to put the swap file above 4 GB
in RAM.

I had to clear all swap and reboot in order to set
it up, but once I did that I was able to delete
C:\pagefil.sys.

It's funny how often the best programs are not
even on anyone's top ten list. Dataram seems to
say that they only allow 1 GB on the free version,
but I didn't find that. I had to go back to v. 3.5 to
get XP support *and* avoid the mess of installing
.Net 4. (And to skip the version you warned about.)
So maybe that's why there was no limit. It tried to
go lonline once during install, but aside from that
it now seems to be fine. I'll have to edit some giant
images and see how it works.


Mayayana

unread,
Dec 20, 2019, 10:00:30 PM12/20/19
to
-
"Shadow" <S...@dow.br> wrote

| Have you tried measuring speed? CrystalDiskMark shows quite
| impressive numbers for RAMDisks. If it's a faux RAMDisk, numbers will
| be much lower.

See my post to Paul. I got it set up. Then I tried
Crystal, which I'd never seen before. I don't know
how to interpret it. The ramdisk had only 94 MB free
so I tested a run of 64 MB. The result was 10 times
higher than read/write to a normal partition. I don't
know what that means. Maybe I'd need to set up a
folder on the ramdisk that's not swap in order to test
it.


Paul

unread,
Dec 21, 2019, 12:12:07 AM12/21/19
to
You can use HDTune free version to test the
"whole surface" of your RAMDisk in read mode.
You don't need to test write mode particularly,
as there's no reason for read versus write to
vary. There's no "physical process" involved here.

My HDTune read is a flat line at around 4GB/sec, which
isn't really all that good of a result. The best
results I can get in the room, are around 7GB/sec.
And this is nowhere near what a Streams benchmark
should be reporting, especially on the other machine.

When you work out the bandwidth claimed based on
the memory numbers, the results on the RAMDisk won't
even be close. You get what you can from it.

I discovered one other anomaly, which is that performance
is more consistent on OSes like Windows 10, if you set
the Power schema to "High Performance" and jam the CPU
to the nominal clock. That'll help if your graph line isn't
flat, when HDTune testing. You'll get stairsteps
in the graph, if the machine is running the "Balanced"
schema. I reported a previous result for Windows 10,
where HDTune didn't seem to be returning the right
numbers on Windows 10, and perhaps this latest
discovery of the schema setting, might help with
that too.

Paul

Mayayana

unread,
Dec 21, 2019, 8:45:31 AM12/21/19
to
"Paul" <nos...@needed.invalid> wrote

| You can use HDTune free version to test the
| "whole surface" of your RAMDisk in read mode.
| You don't need to test write mode particularly,
| as there's no reason for read versus write to
| vary. There's no "physical process" involved here.
|
| My HDTune read is a flat line at around 4GB/sec, which
| isn't really all that good of a result. The best
| results I can get in the room, are around 7GB/sec.
| And this is nowhere near what a Streams benchmark
| should be reporting, especially on the other machine.
|
I suppose the real test is with RAM use. As it is,
things on my XP box are pretty much instant. I like
the idea of ramdisk to reduce writes to the SSDs.
I also see value for big operations. There's not really
any other reason it would matter.

Maybe I'll try to set up some tests, like applying
filters in Paint Shop Pro to gigantic images. If the
ramdisk is working as expected then it should be
able to do those operation nearly instantly, on
operations where my other XP box will show a progress
bar and take some number of seconds to complete
the operation.



Shadow

unread,
Dec 21, 2019, 8:50:44 AM12/21/19
to
On Fri, 20 Dec 2019 21:59:50 -0500, "Mayayana"
<maya...@invalid.nospam> wrote:

>-
>"Shadow" <S...@dow.br> wrote
>
>| Have you tried measuring speed? CrystalDiskMark shows quite
>| impressive numbers for RAMDisks. If it's a faux RAMDisk, numbers will
>| be much lower.
>
> See my post to Paul. I got it set up. Then I tried
>Crystal, which I'd never seen before. I don't know
>how to interpret it. The ramdisk had only 94 MB free
>so I tested a run of 64 MB. The result was 10 times
>higher than read/write to a normal partition. I don't
>know what that means.

It means it's a real RAMDisk.


> Maybe I'd need to set up a
>folder on the ramdisk that's not swap in order to test
>it.

Yes, just temporally move your swapfile to c:\ (or whatever),
reboot, and format the RAMDisk before you test it.
[]'s

Mayayana

unread,
Dec 21, 2019, 9:24:03 AM12/21/19
to
"Shadow" <S...@dow.br> wrote

| > See my post to Paul. I got it set up. Then I tried
| >Crystal, which I'd never seen before. I don't know
| >how to interpret it. The ramdisk had only 94 MB free
| >so I tested a run of 64 MB. The result was 10 times
| >higher than read/write to a normal partition. I don't
| >know what that means.
|
| It means it's a real RAMDisk.
|

No, I mean it was 10 times higher number. It took
10 times as long.

| Yes, just temporally move your swapfile to c:\ (or whatever),
| reboot, and format the RAMDisk before you test it.

Formatting at boot to FAT-32 is part of the deal.

But I did try Paul's suggestion of HDTune. I'm
getting an average about 120 MB/sec with the SSD,
showing a line that looks like Donald Trump hooked
up to a lie detector. The ramdisk, as Paul noted, is
almost a straight line, at about 2300 MB/sec. It looks
like I'm in business. And it was all very simple, with
clear directions, to use the Dataram option. I'm glad
I asked here. I was wasting a lot of time doing
research online and reading chat groups where no
one seemed to quite know what they were talking
about.


Shadow

unread,
Dec 21, 2019, 2:36:16 PM12/21/19
to
On Sat, 21 Dec 2019 09:23:19 -0500, "Mayayana"
<maya...@invalid.nospam> wrote:

>"Shadow" <S...@dow.br> wrote
>
>| > See my post to Paul. I got it set up. Then I tried
>| >Crystal, which I'd never seen before. I don't know
>| >how to interpret it. The ramdisk had only 94 MB free
>| >so I tested a run of 64 MB. The result was 10 times
>| >higher than read/write to a normal partition. I don't
>| >know what that means.
>|
>| It means it's a real RAMDisk.
>|
>
> No, I mean it was 10 times higher number. It took
>10 times as long.

The result is in MB/s
Try an earlier version, the latest is not "guaranteed" to work
on XP.
Here's the"old" portable version:

http://c3sl.dl.osdn.jp/crystaldiskmark/68624/CrystalDiskMark6_0_2.zip

(that's the official mirror)

>
>| Yes, just temporally move your swapfile to c:\ (or whatever),
>| reboot, and format the RAMDisk before you test it.
>
> Formatting at boot to FAT-32 is part of the deal.
>
> But I did try Paul's suggestion of HDTune. I'm
>getting an average about 120 MB/sec with the SSD,
>showing a line that looks like Donald Trump hooked
>up to a lie detector. The ramdisk, as Paul noted, is
>almost a straight line, at about 2300 MB/sec. It looks
>like I'm in business. And it was all very simple, with
>clear directions, to use the Dataram option. I'm glad
>I asked here. I was wasting a lot of time doing
>research online and reading chat groups where no
>one seemed to quite know what they were talking
>about.

I just spent a couple of hours uninstalling my SoftPerfect
RAMDisk and installing Dataram_RAMDisk_v4_0_5_RC0.

Here's a test result:

https://postimg.cc/HcDC4L1H

Here's some notes I made:
1) Block DataRam from phoning home. It attempts to reach out to
memory.dataram.com and
license.dataram.com

No idea why it needs to check the license, I'm using the free
version, but the configuration window takes up to 4 minutes to appear
while it waits for a reply.
I unplug my cable whenever I install anything new.
So - firewall that.

2) Don't choose "Unformatted" in that first window, or Windows Disk
Manager won't see it. Choose "FAT32". Once it's up, you can format to
NTFS, and change the letter(I used "Z" so it doesn't interfere with my
pendrive-backup scheme).
DataRam will remember the drive letter and format(I save the
image, since I'm using it for my Palemoon profile).

3) I now have a lot more "total" RAM, so Dataram DOES use memory above
4GB, unlike the other offerings I tried.

4) I had tried Dataram previously, but it crashed a lot. Not the
version recommended by Paul though. So I'm going to keep track of the
TBC(time between crashes). Softperfect only corrupted the image file
once in 6 months, but required re-installing the driver and copying my
browser profile from a backup.

PS You can use OSFMount
https://www.osforensics.com/tools/mount-disk-images.html

For XP(not listed)
https://www.osforensics.com/downloads/osfmount_v1.5.1018.exe

to mount Dataram images. Choose "Partition type" not "whole file"
because of the offset.
HTH

Shadow

unread,
Dec 21, 2019, 2:51:37 PM12/21/19
to
On Fri, 20 Dec 2019 20:00:10 -0500, Paul <nos...@needed.invalid>
wrote:

>> On Fri, 20 Dec 2019 17:54:40 -0500, "Mayayana"
>> <maya...@invalid.nospam> wrote:
>>
>>> I'm doing this on XP but that group seems to be rather quiet
>>> these days.
>>>
>>> Trying to set up ramdisk swap file on 32 bit. Have
>>> 16 GB RAM. Gavotte version didn't seem to work. ImDisk
>>> worked fine. But there's a catch. Virtual memory settings
>>> show that I set up 3 GB virtual memory on the ramdisk,
>>> drive R, but there's now a pagefile.sys on C drive that's
>>> taking up 3 GB. Settings show no swap file for C drive.

..............
>
>
>AFAIK, 4GB max size, for free. Even if the file versions
>don't match, it's still a fine addition to a WinXP system
>that has too much RAM.
>
>*******
>
> Dataram_RAMDisk_v4_0_5_RC0.msi
>
> 5,567,488 bytes
> MD5 = 1c9c709c647d979b1277f6d756dce265
> SHA1 = 95f123af724206b9c8dd61dd76bf388098fe9b54
>
>
> [Actually, try here. Took me a while to find this! Immediate download...]
> https://filehippo.com/download_ramdisk/14475
>
>
>Always check the checksums.

I always do. They match...
:)

Thanks for the info Paul. Dataram up and running.
I posted a few notes to Mayayana, which might help. I don't
see how using a swapfile in RAM will speed things up much for me,
since my system rarely uses it, but it certainly speeds the system up
if used for browser cache or TEMP folders ....
The only downside is Windows takes a little longer to shutdown
because it saves the image, so I use a small 128 --> 256MB image.
>
>HTH,

It did, very much so...
> Paul
[]'s

Mayayana

unread,
Dec 21, 2019, 5:08:14 PM12/21/19
to
"Shadow" <S...@dow.br> wrote

| I don't
| see how using a swapfile in RAM will speed things up much for me,
| since my system rarely uses it, but it certainly speeds the system up
| if used for browser cache or TEMP folders ....

It should reduce wear on SSDs, since the swap file seems
to be used even when it's not needed.


Mayayana

unread,
Dec 21, 2019, 5:10:29 PM12/21/19
to
"Shadow" <S...@dow.br> wrote

| 4) I had tried Dataram previously, but it crashed a lot. Not the
| version recommended by Paul though. So I'm going to keep track of the
| TBC(time between crashes). Softperfect only corrupted the image file
| once in 6 months, but required re-installing the driver and copying my
| browser profile from a backup.

I would never use it for anything but swap, so
I'm not worried about that. Using it for TEMP
or other folders seems dubious to me, and I don't
enable much browser cache.


Shadow

unread,
Dec 21, 2019, 6:24:58 PM12/21/19
to
Browsers do an incredible amount of writing, reading then
deleting files in the cache/profile folders, even with all the
"private browsing" and other settings. Can't find an article that
actually measures it, but I did read an article where it said it's
probably the biggest wear and tear on SSD's.
As to swapfile, yes, Windows uses it, even if you have oodles
of physical RAM. Not sure how much though.
Someone should measure them ... I don't know a software that
can measure disk usage by a program.
Trying Filemon. Filter = "palemoon".
Starting Palemoon. Less than a second to load. 881 accesses to
the disk in that time.
Typed in Majorgeeks.com. Press enter. Another > 400 entries.
It's a lot. But it didn't register ANY accesses to my profile
folder on RAMDISK Z:, All of them were to my C: drive
Filemon is obviously not the right tool for the job.
:(

Mayayana

unread,
Dec 21, 2019, 8:37:05 PM12/21/19
to
"Shadow" <S...@dow.br> wrote

| Trying Filemon. Filter = "palemoon".
| Starting Palemoon. Less than a second to load. 881 accesses to
| the disk in that time.
| Typed in Majorgeeks.com. Press enter. Another > 400 entries.
| It's a lot. But it didn't register ANY accesses to my profile
| folder on RAMDISK Z:, All of them were to my C: drive
| Filemon is obviously not the right tool for the job.

Those are almost all reads, PM loading DLLs, loading
prefs, that sort of thing. A ramdisk wouldn't make a difference
there. And my understanding is that the wear on an SSD
comes only from writes. You might be right about cache,
though I have my PM cache limited to 10 MB. Maybe I should
just disable it. A site being cached for good reason is rare
these days because alomst every page is created on-call,
which means browsers ar never getting a 302, which means
they never use their cached pages. Caching is outdated.


Mayayana

unread,
Dec 28, 2019, 3:50:52 PM12/28/19
to
"Paul" <nos...@needed.invalid> wrote

| Dataram RAMDisk Driver
| V4.0.4.2
| Dataram RAMDiskVE Driver
| Dataram, Inc.
|

An update and warning on this:

I installed v. 3.5.130 for XP support. The most recent
versions don't support XP and a bit further back they
switched to .Net v. 4, which I prefer to avoid. It all
worked very well from the start and continues to work
well.

Except.... When I opened Visual Studio 6 I was getting
errors. It appears that the Dataram author is talented
with low-level operations but somewhat ignorant of
GUI programming. The program uses several ActiveX
controls, unnecessarily. And the installer makes an
unforgiveable mess of Registry settings. It actually
changes some specific interface settings without
registering the controls. I had to unregister and
re-register the system versions of those controls. It
was an amazingly messed up installer. And when I uninstall
I'll need to remember to watch out for corruption again.

The only thing
I can think of is that maybe the MSI was made by
a very messed up MSI maker. The only clue is that the
program used to make the MSI, as listed in the Summary
Info, is "Windows Installer Editor Standalone". I'm not
sure if that's an actual program or a default value used
when someone sets up their own MSI using something
like Wix.

At any rate, beware of all versions. The author clearly
doesn't know how to do a proper software install. There
could also be problems with the .Net version. I was going
to send them a note but they seem to be very evasive.
There's no email listed on their website. There's only an
option for paying customers to get a "support ticket".

Weird. But it works when nothing else did.


Shadow

unread,
Dec 29, 2019, 1:25:23 PM12/29/19
to
On Sat, 28 Dec 2019 15:50:09 -0500, "Mayayana"
<maya...@invalid.nospam> wrote:

>"Paul" <nos...@needed.invalid> wrote
>
>| Dataram RAMDisk Driver
>| V4.0.4.2
>| Dataram RAMDiskVE Driver
>| Dataram, Inc.
>|
>
>An update and warning on this:
>
> I installed v. 3.5.130 for XP support. The most recent
>versions don't support XP and a bit further back they
>switched to .Net v. 4, which I prefer to avoid. It all
>worked very well from the start and continues to work
>well.

I installed https://filehippo.com/download_ramdisk/4.0.0.5.-1/
as recommended by Paul.
Only thing that annoys me is I can't make it into a
"removable" drive, i.e. I get RECYCLER and System Volume Information.
No crashes so far.
(See below)
You can use lessMSI to "look inside" the installer and even
extract files:

https://sourceforge.net/projects/lessmsi.mirror/files/

It's written in VB6.(DataRam I mean).
Maybe it installed old libraries?
[]'s

Mayayana

unread,
Dec 29, 2019, 2:44:45 PM12/29/19
to
"Shadow" <S...@dow.br> wrote

| I installed https://filehippo.com/download_ramdisk/4.0.0.5.-1/
| as recommended by Paul.
| Only thing that annoys me is I can't make it into a
| "removable" drive, i.e. I get RECYCLER and System Volume Information.
| No crashes so far.

I also found no problems at all and appreciate the
good docs. It's exactly what I wanted. The only
problem is the installer. The version you have is using
.Net, so it may be OK. But be warned. The author
doesn't seem to know anything about proper software
installation methods.


| You can use lessMSI to "look inside" the installer and even
| extract files:
|
| https://sourceforge.net/projects/lessmsi.mirror/files/
|

I'm familiar with that. It's actually one of the few MSI
unpackers that actually works. But I daresay my own
is arguably the best. (As Walter Brennan used to say
in an old 70s cowboy show, "No brag, just fact." :)

http://www.jsware.net/jsware/msicode.php5#unpackx

______________________________________________

* Lessmsi: A wrapper around the MS Wix libraries, which
are an unnecessary wrapper around the system file,
msi.dll, which has all the functionality for handling MSIs.
(MSIs are so poorly designed, and so monstrously
complicated, that few people can work with them. So
Microsoft actually created a software package called
Wix, to deal with their software package for making
installers. Wix is also made with .Net, so it's a bloated,
unnecessary wrapper around Microsoft's incredibly
bloated installer calamity.)

Lessmsi also requires .Net 4 or higher. That's OK if you
already need .Net 4. I don't. So I'd be installing a lot
of useless bloat to do a simple job. Lessmsi + Wix + .Net 4.
____________________________________________

* jsMSIx: My own version. Uses msi.dll functions directly
and incorporates functionality to unpack CAB files. It's a
total of 200 KB, with no extra dependencies and no need
to install.

jsMSIx not only unpacks an MSI but also writes a log
detailing the file list, Registry entries... basically everything
that the MSI will do when it's run.

For anyone who's interested, at my webpage there are
also VBScript versions, which also use msi.dll directly.
And there are various tools for working with MSIs, as
installers or as handy databases. (I use several. One is
a database to store all my old email in a searchable
format.)
I also have an MSI editor, made as an HTA, that allows
one to see the table view and add, delete, edit tables,
rows and columns.

I started working with MSIs in the early days because
I was working on text-to-speech software for a blind friend.
I needed SAPI5 speech support to do it. It turned out that
the only way to get that was to download 100 MB of SAPI
"merge modules" and then build an MSI installer that would
use about 6 MB of that slop. When I set about doing as
Microsoft advised I quickly realized that MSIs were a ridiculously
complex way to build a software installer. I would have spent
more time on the installer than on the software. But SAPI5
support wasn't available in a normal system update package.
So I had to figure out how MSIs worked, in order to get the
SAPI5 support separate from the MSI mess. That led to all the
tools I made. Included among those tools are also VBScript
tools to automate a non-MSI setup from an MSI installer.
I had to write that to get at sapi.dll and the handful of Registry
settings I needed for my own software.

Lessmsi has become well known because it's so-called
open source. But it you look at the code you'll see that
there's not actually much of it there. It just calls Wix
to do the job. But that doesn't stop the author from
adding a lot of legalese at the top of every file, claiming
that you can't use his code unless you include his
copyright notice and give him credit.

As you may have guessed, I don't have a lot of sympathy
for people who write a few lines of common wrapper code
and then make a big deal about how it's their personal
masterpiece. There are far too many people out there
hoarding code, and most of them hardly even know how
to write it.

It reminds me of the Monty Python skit about the scientist
who goes on TV to explain his new theory about dinosaurs.
The scientist clears his throat, repeatedly announces that he
has a theory, and generally wastes time with fanfare. Finally,
with the talk show host at the limit of his patience, the
scientist graces us with his amazing theory:

"The brontosaurus is small at one end, quite large in the middle,
then small again at the other end... That's my theory. Which is
mine. Which is my theory. Which belongs to me." :)

| It's written in VB6.(DataRam I mean).
| Maybe it installed old libraries?

I explained it already in my earlier post.
Using my own MSI unpacker I was able to see the
Registry settings that the installer changed. I'm not
sure that it used older controls but what it did do,
which is crazy, unacceptable behavior, was to change
only some CLSID keys for the specific functionality it
was using. Then it also wrote a bogus value so that
it could reverse the damage whem unistalled. It was
operating as though it were the only program that would
ever use ActiveX controls!

Example: The Microsoft common controls OCX is a
main system file that provides a number of GUI items:
listview, treeview, tabstrip, slider, progress bar, and
so on. The Dataram installer altered only the COM
settings for specific items it was using. For instance,
it might change the path of the tabstrip but leave the
path of the listview! That behavior is wrong and unstable
in numerous ways. The installer should have checked
whether the file was present, then checked whether
the present version was older, then installed the new
version *only* in that case and *only to the system folder*.
Then the file should have been registered. The whole thing.
It should have been told to register itself.
Instead, what Dataram did was to edit the registration
settings selectively in the Registry, without doing any
registration.

The way that works, in case anyone's curious and
doesn't know, is that COM libraries are self-registering.
If you call mscomctl.ocx to register itself then it will add
dozens, maybe hundreds, of settings into HKCR. Those
settings are necessary for it to work. Typically
there's a ProgID entry, like
HKEY_CLASSES_ROOT\MSComctlLib.TabStrip

That key then points to a key under HKCR\CLSID. That key
has numerous subkeys that point to such things as the
path of the file containing the tabstrip functionality,
as well as pointing to the path of the type library. It's
fairly complicated. The Dataram installer was going in
and changing only the file path, from system32 to the
Dataram folder. That was breaking things in other software.

I've never seen such a harebrained hack. Even just
registering the controls in its own folder would have been
against the rules of proper software installation. To essentially
hack the Registry settings for system libraries is worse.


Shadow

unread,
Dec 29, 2019, 6:29:29 PM12/29/19
to
RAMDisk.exe is actually a VB6 executable. The Microsoft Visual
C# / Basic.NET / MS Visual Basic 2005,7 and 10 files appear to be the
registration routine.

According to SoftOrganizer, the installer made 567 changes to
the registry so yes, very badly implemented. The author does not
believe in the KISS doctrine..(maybe he did in the later versions,
don't know).

As to the programs you mentioned, I have HTA and scripting
disallowed(security), and didn't manage to download jsMSIx. I use wget
for downloads, so I can keep a log of where I got each file.

Your page seems to have mistaken it for a download manager.

Mayayana

unread,
Dec 29, 2019, 9:06:33 PM12/29/19
to
"Shadow" <S...@dow.br> wrote

| RAMDisk.exe is actually a VB6 executable. The Microsoft Visual
| C# / Basic.NET / MS Visual Basic 2005,7 and 10 files appear to be the
| registration routine.
|

Yes. I know it's VB6. I think we're crossing wires here. :)

There's no .Net in Dataram's program. And the installer is
not .Net. It's just an MSI.
I was talking about Lessmsi being .Net. Lessmsi requires .Net
v. 4, which is not something I'm ever likely to need, so I'd
be installing probably 100s of MB of stuff I don't need (plus
the Microsoft Wix libraries) to run a very simple program
that shouldn't have any dependencies.

| As to the programs you mentioned, I have HTA and scripting
| disallowed(security), and didn't manage to download jsMSIx. I use wget
| for downloads, so I can keep a log of where I got each file.
|
| Your page seems to have mistaken it for a download manager.

Sorry about that. It will work fine if you just use
a browser. Or you can spoof the UA. (I'm surprised
you don't already do that.)

I do block a lot of downloader programs.
They tend to be very sloppily made and the people who
use them tend to just go around grabbing things. I got
tired of it once when one of them downloaded over
300 copies of a program in about 1 minute. And that
was back in the old days when traffic was expensive.

It really is amazing how little people pay attention
online. People often grab one of everything -- more stuff
than they'll ever look at. Or they "scrape" the whole site.
It's just manic consumption. My MSI Unpacker is my most
popular download. There are several options and there's
a lot of information on the webpage. But the typical time
for people to download it, from the time they load the
webpage, is 6-10 seconds. I don't know how they even
find the link that fast! Just grab and go.

It reminds me of those old TV shows where people would
win a chance to run around a supermarket and get everything
they can grab in 2 minutes. I don't know why people do that.
Maybe it's just a result of the endless possibilities online.
One link leads to another and people end up racing around.
I sometimes find myself getting strung out like that. I start
out researching something. One thing leads to another.
My search branches. At some point I have 15 webpages open,
have lost track of all the things I plan to read, and I'm starting
to get low blood sugar. :)

If you don't allow script or HTAs you can still use jsMSIx,
which is a standard EXE. But it's up to you. I make no
money from it. If you like Lessmsi that's fine.

It seems a shame, though, for someone as handy as you
to block those things. An HTA is just a webpage that runs
in IE with no security. Bu it does have one limitation: It
can only run locally. So you can't be tricked online. If you
click a link in IE to an HTA online you'll get a download
dialogue. And the default icon is like the generic DOS
executable icon. So it would be very hard to get tricked
by it. Similarly with VBS, someone like you can't really be
tricked into running VBS malware, such as an email attachment,
because you know it's script. The real risks are script in
webpages and script embedded in MS Office files when MS
Office is installed. And disabling .js, .vbs and .hta won't help
a bit with either of those risks.

I have 9 HTAs on my desktop, and probably 30 VBScripts.
I use them for all kinds of things. One of the most common
uses is a script I wrote to fix carriage returns in Unix files.
I can just drop a whole folder on it and it will convert all
CR characters in text files to CR LF, the normal Windows
style. Another script cleans all TEMP files. Yet another
encodes/decodes Base64. And I have a javascript
deobfuscator HTA that I wrote for parsing obfuscated
webpage code. (That's another example where there's a
popular OSS version, but it's not very good. I think it's called
something like jsBeautifier.)
Recently I wrote another HTA for automating youtube-dl, so
I don't have to deal with the tedium of a console window.
I even wrote a graphic editor as an HTA.

I make HTAs because it's so powerful, fun, and relatively
simple. Webpages can host most of the typical things that
appear in a program window: text windows, buttons, dropdown
selectors, etc. And COM gives script vast powers. Of course,
that's why IE became dangerous and it's why webpages are
no longer safe. I rarely allow script online and haven't used
IE online since 2000. But offline it's a gem of ingenuity. COM
and ActiveX were really a brilliant invention. MS just made the
mistake of trying to use it online.

That's basically why MS invented HTAs in the first place. When
they began to see that ActiveX was a disaster online it was
a problem because a lot of corporate admins were using IE
to make simple but powerful utilities to run on their networks.
So MS forked the security. They gradually started restricting
script and ActiveX in IE, while at the same time allowed HTAs
free reign. All it takes is naming a file HTA instead of HTML.
Then it ges run by mshta instead of iexplore.exe. But it's the
same thing. Both of those are just thin wrappers around an
IE browser window.


J. P. Gilliver (John)

unread,
Dec 29, 2019, 9:19:21 PM12/29/19
to
In message <qubm36$rjk$1...@dont-email.me>, Mayayana
<maya...@invalid.nospam> writes:
[]
> Recently I wrote another HTA for automating youtube-dl, so
>I don't have to deal with the tedium of a console window.
[]
Did I miss your announcement of this, or haven't you made one?
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

"Bother," said Pooh, as he fell off the bridge with his stick.

Mayayana

unread,
Dec 29, 2019, 9:39:19 PM12/29/19
to
"J. P. Gilliver (John)" <G6...@255soft.uk> wrote

| > Recently I wrote another HTA for automating youtube-dl, so
| >I don't have to deal with the tedium of a console window.
| []
| Did I miss your announcement of this, or haven't you made one?


https://www.jsware.net/jsware/scrfiles.php5#ytdlh

I didn't post it. No one seemed to be interested, and
I thought I remembered you saying that you like the
console window. So I figured probably no one was curious.

Few people seem to share my distaste for typing obscure
commands in DOS windows. But for me, if I have to do it
more than once than it's worth writing a mouse-based
alternative.



J. P. Gilliver (John)

unread,
Dec 30, 2019, 1:27:45 AM12/30/19
to
In message <qubo00$4ia$1...@dont-email.me>, Mayayana
<maya...@invalid.nospam> writes:
>"J. P. Gilliver (John)" <G6...@255soft.uk> wrote
>
>| > Recently I wrote another HTA for automating youtube-dl, so
>| >I don't have to deal with the tedium of a console window.
>| []
>| Did I miss your announcement of this, or haven't you made one?
>
>
>https://www.jsware.net/jsware/scrfiles.php5#ytdlh

Thanks.
>
> I didn't post it. No one seemed to be interested, and
>I thought I remembered you saying that you like the
>console window. So I figured probably no one was curious.
>
> Few people seem to share my distaste for typing obscure
>commands in DOS windows. But for me, if I have to do it
>more than once than it's worth writing a mouse-based
>alternative.
>
>
>
(3)
I don't like the console window or typing that much, I just wasn't sure
a GUI would save me _that_ much, since I still have to paste the URL,
and I know how to do that in a console window (alt-space, E, P). But I
tried it: downloaded and scanned the .zip from the above, and extracted
the folder: C:\Program Files\JSWare\Youtube-dl Helper . Did "take
ownership" on it to make sure it could save its settings. Then wasn't
sure how to run it, but double-clicking on the .hta seemed the best
guess, so I did, and your GUI opened. I pasted in D:\MOVIES\X\!pending
as both path to Youtube-dl.exe and download folder. Then, in absence of
a close button, I just clicked the X. It closed, and I saw ytdlsets.dat
appear. I opened it (incidentally, do you know how to stop "Always use
the selected program to open this type of file" being pre-ticked when
you click an unknown filetype? I usually try a text editor first, but
don't want that associated. Someone told me how under XP, but I've
forgotten and it's probably different for 7 anyway). It (ytdlsets.dat)
contained three CRLF pairs (I checked with a hex viewer). I opened ytdlh
again, and it hadn't remembered (.exe or download locations). OK, I
thought, maybe I have to actually use it to make it remember, so I found
a short one to try - https://www.youtube.com/watch?v=85DQB0C_eZY -
pasted that in (having also refilled the other two), and clicked Get
File. Error sound and "Youtube-dl.exe not found." popup. Checked: it
_is_ there. Tried it the old-fashioned way - still worked. (Didn't we
accept grotty video 37 years ago!) Closed and reopened GUI - all boxes
empty. The .dat file timestamp had changed, but still only 6 bytes big.

What am I doing wrong?
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

If you carry on hating, you're the one who's damaged.
- Sir Harold Atcherley, sent to the Burma/Siam railway in April 1943

Mayayana

unread,
Dec 30, 2019, 9:19:57 AM12/30/19
to
"J. P. Gilliver (John)" <G6...@255soft.uk> wrote

|Then wasn't
| sure how to run it, but double-clicking on the .hta seemed the best
| guess

Woops. I guess I should make that more clear in
the readme file. I don't know why I didn't list the
actual file name.

| It closed, and I saw ytdlsets.dat
| appear. I opened it (incidentally, do you know how to stop "Always use
| the selected program to open this type of file" being pre-ticked when
| you click an unknown filetype? I usually try a text editor first, but
| don't want that associated. Someone told me how under XP, but I've
| forgotten and it's probably different for 7 anyway).

I don't know if there's a specific setting to do that
for all unknown files. I don't deal with it much. I like
to have "Open with Notepad" and several other similar
options on right-click, so I use those. I almost always
expore an unknown file with either "Open with Notepad"
or "Open with HxD".

| It (ytdlsets.dat)
| contained three CRLF pairs (I checked with a hex viewer). I opened ytdlh
| again, and it hadn't remembered (.exe or download locations). OK, I
| thought, maybe I have to actually use it to make it remember, so I found
| a short one to try - https://www.youtube.com/watch?v=85DQB0C_eZY -
| pasted that in (having also refilled the other two), and clicked Get
| File. Error sound and "Youtube-dl.exe not found." popup. Checked: it
| _is_ there. Tried it the old-fashioned way - still worked. (Didn't we
| accept grotty video 37 years ago!) Closed and reopened GUI - all boxes
| empty. The .dat file timestamp had changed, but still only 6 bytes big.
|
| What am I doing wrong?

It's saving the settings file but doesn't have any
settings to save.

Note the browse button. When you browse to select
the file/folder it confirms it's valid then stores it, for
use in the program and to be saved at close. If you just
paste in the path the program doesn't know you did that.
There's no "Save Setting" button.

It's interesting how my preconceptions always turn up
when people try software I've written. I think the first
such revelation was the first time I asked someone to
try a program and they crashed it by clicking every
button multiple times. It had never occurred to me that
someone might do that! But many people just keep clicking
until they see something happen. So from then
on I had to make sure that the code for every button
could deal with multiple clicks.

My idea here was that it's cleaner and simpler not to
have a "Save Setting" button. It's designed to be as
easy as possible. But I didn't account for people pasting
in a path when they can browse to it instead. :)

Why would it need a Save button, you might ask?
To not have a Save button and not require the Browse
button would mean I don't know when you've entered
text that you want to be the setting value. I'd have to
check and confirm the validity of the text in those
textboxes every time you click "Get File". That would
be very sloppy. Or I'd need to constantly monitor
changes to the textbox. That's inefficient and prone
to possible errors. Or I'd need to make the settings a
separate window so that I'd know when you entered
a new setting. It's just an HTA. I wasn't getting that fancy.
Sorry, sir, you'll just have to use it the easy way. :)

What's with the exclamation point? That's allowed?
I've never tried it before.


Mayayana

unread,
Dec 30, 2019, 9:35:13 AM12/30/19
to
"Mayayana" <maya...@invalid.nospam> wrote
| "Shadow" <S...@dow.br> wrote
|
|| RAMDisk.exe is actually a VB6 executable. The Microsoft Visual
|| C# / Basic.NET / MS Visual Basic 2005,7 and 10 files appear to be the
|| registration routine.
||
|
| Yes. I know it's VB6. I think we're crossing wires here. :)
|
| There's no .Net in Dataram's program. And the installer is
| not .Net. It's just an MSI.

I realized later that you're probably talking about the
newer Dataram Ramdisk. As I noted above, I'm using the
older version, which has no .Net and supports XP. Later
versions added .Net. Still later versions dropped XP support.
The basic functionality is actually a driver. But the GUI
uses VB6 or .Net, depending on the version. From what
I can gather, the version I have seems to be a thrown-
together VB6 GUI that mainly just existing as a front-end to
the INF file, so that the RAMDisk can be installed or
uninstalled conveniently.

My guess is that the author(s) is someone knowledgeable
about low level functionality like drivers but just grabs
what's handy for the GUI and installer. Or maybe told the
new guy to do it. :)


Shadow

unread,
Dec 30, 2019, 1:37:16 PM12/30/19
to
Probably.

Yes, as I said, I installed the version Paul recommended.
Later versions limited the size of the RAMDisk in the free versions,
and later still it appears support for XP was dropped.
The licensing routine was probably written by a third party,
and he chose a crappy compiler.

As to your other post, I'm a lazy "GUI frontend" programmer,
though I do my stuff in Pascal (Lazarus).
My frontend to wget saves a log, my frontend to youtube-dl has
radio boxes for the quality of the video.
I managed to DL your little program with -U "useragent"
"utility_link" typed into my wget GUI.
I quarantine ANYTHING for two weeks, then submit it to Jotti.
One of my "security" measures, nothing personal.
I'll give you feedback. Later.

Paul

unread,
Dec 30, 2019, 2:56:44 PM12/30/19
to
DataRAM, as a company, is a "seller of DIMMs".

The RAMDisk software product, is a sideline.

My suspicion is, the product is provided by a "contractor",
and a dedicated staff is not likely to exist in the
DataRAM building. I would not try to imagine a "quality"
department beating on a local developer, trying to make
him or her do the right thing.

And the reason you use a contractor, is ramdisks
are hard, and it's taken an inordinate amount of
time for a RAMDisk to be as capable as that one
is. If you have any experience with older RAMDisks,
the difference is night and day. The first RAMDisk
was sample code Microsoft provided, and that put people
down wrong paths for a long long time. They tried to
iterate that code, to no good effect. It's a bit
like the "AmCap effect", where the AmCap code took
an inordinate amount of time to be augmented by
a developer and beat into decent shape. In a sense,
sample code seems to do an awful lot of damage, human
nature being what it is.

Paul

Mayayana

unread,
Dec 30, 2019, 4:00:43 PM12/30/19
to

"Shadow" <S...@dow.br> wrote

| I managed to DL your little program with -U "useragent"
| "utility_link" typed into my wget GUI.
| I quarantine ANYTHING for two weeks, then submit it to Jotti.
| One of my "security" measures, nothing personal.
| I'll give you feedback. Later.

Sounds fine. If I don't hear back I'll alert
the authorities to check on you. :)

I actually started doing hashing on the program
installers, so that people can check that. But I
don't use any kind of functionality that lends itself
to hacking my site, so I don't really worry about that.
No SQL. No Wordpress. No backend functions of
any kind other than PHP includes.




Shadow

unread,
Dec 30, 2019, 5:49:55 PM12/30/19
to
On Mon, 30 Dec 2019 15:59:58 -0500, "Mayayana"
<maya...@invalid.nospam> wrote:

>
>"Shadow" <S...@dow.br> wrote
>
>| I managed to DL your little program with -U "useragent"
>| "utility_link" typed into my wget GUI.
>| I quarantine ANYTHING for two weeks, then submit it to Jotti.
>| One of my "security" measures, nothing personal.
>| I'll give you feedback. Later.
>
> Sounds fine. If I don't hear back I'll alert
>the authorities to check on you. :)

Please don't. It's the "authorities" that abduct, torture and
kill you here.
Just remind my wife not to buy roses for my funeral. I'm
allergic.
;)
>
> I actually started doing hashing on the program
>installers, so that people can check that. But I
>don't use any kind of functionality that lends itself
>to hacking my site, so I don't really worry about that.
>No SQL. No Wordpress. No backend functions of
>any kind other than PHP includes.

Yes, I saw the checksums.

https://virusscan.jotti.org/en-US/search/hash/08112aff0c9d0d7a72e06cffec843b207ef5dc652b4eef62f1c5d0dd13341659

It's a good precaution.

J. P. Gilliver (John)

unread,
Dec 30, 2019, 8:13:40 PM12/30/19
to
In message <qud129$1km$1...@dont-email.me>, Mayayana
<maya...@invalid.nospam> writes:
>"J. P. Gilliver (John)" <G6...@255soft.uk> wrote
>
>|Then wasn't
>| sure how to run it, but double-clicking on the .hta seemed the best
>| guess
>
> Woops. I guess I should make that more clear in
>the readme file. I don't know why I didn't list the
>actual file name.

No problem (-:
>
>| It closed, and I saw ytdlsets.dat
>| appear. I opened it (incidentally, do you know how to stop "Always use
>| the selected program to open this type of file" being pre-ticked when
>| you click an unknown filetype? I usually try a text editor first, but
>| don't want that associated. Someone told me how under XP, but I've
>| forgotten and it's probably different for 7 anyway).
>
> I don't know if there's a specific setting to do that
>for all unknown files. I don't deal with it much. I like
>to have "Open with Notepad" and several other similar
>options on right-click, so I use those. I almost always
>expore an unknown file with either "Open with Notepad"

Notepad (or Notepad+ - often both) are often my first choice too), but I
don't have an "open with Notepad" shortcut in my right-click menu (I do
for IrfanView, though I rarely use it - I think it's there because it's
offered by IV's install, and since IV's code is quite good at working
out what _misnamed_ files are, I le[f]t it).

I suppose I could add an open with Notepad, but on a more general note,
I'd like it if "always open this type of file with" was unchecked by
default. Not so much for new unknown filetypes (though definitely there
too), but when I use Open With and then select something, for a type I
already do have an association for, but want to try something new, but
would rather not have what I'm trying associated from then on with that
filetype, perhaps as my normal association works fine for _most_ files
of that extension. Someone did post the tweak (to stop "Always use ,,,"
being ticked by default) under XP.

>or "Open with HxD".
>
>| It (ytdlsets.dat)
>| contained three CRLF pairs (I checked with a hex viewer). I opened ytdlh
>| again, and it hadn't remembered (.exe or download locations). OK, I
>| thought, maybe I have to actually use it to make it remember, so I found
>| a short one to try - https://www.youtube.com/watch?v=85DQB0C_eZY -
>| pasted that in (having also refilled the other two), and clicked Get
>| File. Error sound and "Youtube-dl.exe not found." popup. Checked: it
>| _is_ there. Tried it the old-fashioned way - still worked. (Didn't we
>| accept grotty video 37 years ago!) Closed and reopened GUI - all boxes
>| empty. The .dat file timestamp had changed, but still only 6 bytes big.
>|
>| What am I doing wrong?
>
> It's saving the settings file but doesn't have any
>settings to save.
>
> Note the browse button. When you browse to select
>the file/folder it confirms it's valid then stores it, for
>use in the program and to be saved at close. If you just
>paste in the path the program doesn't know you did that.
>There's no "Save Setting" button.
>
> It's interesting how my preconceptions always turn up
>when people try software I've written. I think the first

Yes; it never occurred to me that I _have_ to use the browse button!
[]
> My idea here was that it's cleaner and simpler not to
>have a "Save Setting" button. It's designed to be as

I'm fine with that.

>easy as possible. But I didn't account for people pasting
>in a path when they can browse to it instead. :)
>
> Why would it need a Save button, you might ask?
> To not have a Save button and not require the Browse
>button would mean I don't know when you've entered
>text that you want to be the setting value. I'd have to

Perhaps make those boxes not pasteable (-:? (Like I think it's the
bottom box, next to the Play button, which was greyed for me.)

>check and confirm the validity of the text in those
>textboxes every time you click "Get File". That would
>be very sloppy. Or I'd need to constantly monitor
>changes to the textbox. That's inefficient and prone

Yes, I think definite justification for making those boxes not editable!

>to possible errors. Or I'd need to make the settings a
>separate window so that I'd know when you entered
>a new setting. It's just an HTA. I wasn't getting that fancy.
>Sorry, sir, you'll just have to use it the easy way. :)

When I was trying it out, I happened to have the folder - where I had
ytdl - open in Explorer anyway, so cut and paste _was_ easier than
browsing (-:!
>
> What's with the exclamation point? That's allowed?
>I've never tried it before.
>
It's certainly allowed in a file (or folder) name - I think even under
8.3. I use it as a quick and dirty way to bring a file to near the top,
when explorer has reverted to sort by name, which (for practical
purposes of explanation it's IMO valid to say) it has a tendency to do.
>

--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

When I went to see Biddy Baxter [Blue Peter's editor] and told her I was
pregnant, her first reaction was 'Oh good, another viewer'. - Janet Ellis, RT
2016/2/27-3/4
0 new messages