Bogus PPZIP support in WakePup2

Please post any bugs you have found
Message
Author
otropogo

Bogus PPZIP support in WakePup2

#1 Post by otropogo »

Have just tested WakePup2 option #2 "Iomega ZIP.disk drive, connected to the parallel port", and see no evidence of it working.

In fact, there's no indication that Wakepup2 looks for anything on the parallel port.

I tried WakePup2 on floppies created by both Puppy 3.01 Retro and Puppy 4.0.

For good measure, I also tried substituting the new autoexec.bat and config.sys files offered on December 2 in the thread "Yet Another Wakepup2".The previous files date from August and September of 2007,

If there's an earlier version of WakePup2 that actually DID support running Puppy from a parallel port ZIP, I'd like to try it. If not, I suggest this entry should be removed from the options menu.

I haven't tried option #3 "Backpack CD_ROM drive, connected to the parallel port", but I wouldn't count on it working either.

To be sure the problem wasn't in the ZIP drive, the cable, or the parallel port settings, I tested the ZIP under Windows 98SE on one of the three systems ( the other two don't have Windows installed)

In Win98SE, enabling the pp ZIP required running the Iomega "guest" program, after which it worked flawlessly.

User avatar
Crash
Posts: 453
Joined: Fri 09 Dec 2005, 06:34
Location: Melbourne, FL

#2 Post by Crash »

I'm not sure what is supposed to be in that part of the code sequence. When I did the code walk-through in preparation for creating the readme file, I noted the problem. I noted it in the readme file that I posted in January. I also posted a readme file for the 3.01 version in an obscure thread a while back. I'm re-posting it here. I'll look back to see where that part of the code disappeared, but I would appreciate it if anyone else that was involved in editing Wakepup2 could also help.
Attachments
wakepup2_301_readme.txt.gz
(8.16 KiB) Downloaded 481 times

otropogo

#3 Post by otropogo »

Crash wrote:I'm not sure what is supposed to be in that part of the code sequence. When I did the code walk-through in preparation for creating the readme file, I noted the problem. I noted it in the readme file that I posted in January. I also posted a readme file for the 3.01 version in an obscure thread a while back. I'm re-posting it here. I'll look back to see where that part of the code disappeared, but I would appreciate it if anyone else that was involved in editing Wakepup2 could also help.
Have just read your readme text for the first time. But, if anything, it deepens the mystery for me.

After posting my bug report, I managed to find an old WakePup2 floppy from Puppy 2.14, and tried it to compare it with the current version. There was no option to boot anything from the parallel port that I could see. The only options were F8 for a step-through boot, and F5 to skip config.sys and autoexec.bat.

The Kernel.sys on this disk is dated Feb.25 2005, and the config.sys is dated May 11, 2006. So it's a bit of a mystery that these parallel port capabilities were developed and lost again in the space of a year or so.

The other mystery is that when I choose Menu option #7 Pcmcia (experimental), the very last driver: ASPI CDROM Driver for DOS Version 4.01 Copyright 1990-1996 Adaptec, Inc.

finally detects my pcmcia_scsi host adapter and attached CDROM drive:

"host adapter 0, Target SCSI ID =5 LUN 0, and
SONY CD-RW CRX145S 1.0c

CD-ROM driver installed: I host adapter(s), 1 target(s)"

but when I "press any key to continue"

all WakePup does is look for Puppy on idehd, and reports:

"...file found on idehd, drive C:", and offers me a second menu with the Puppy Boot Modes, which doesn't have any option to boot from the freshly detected CDROM drive.

What sense is there in that? It should be asking whether I want to boot from the CD or from the hard drive.

BTW - I've also tried this "pcmcia" option in Wakepup2 with pcmcia flash card readers holding flash cards to which I'd installed Puppy with the Universal Installer, and none of these have ever been presented as boot options either.

User avatar
Crash
Posts: 453
Joined: Fri 09 Dec 2005, 06:34
Location: Melbourne, FL

#4 Post by Crash »

I searched all the intermediate versions of Wakepup2 that I saved last year while following the iterative development of the Wakepup2 that is now in version 3.01. I couldn't find any of them that actually executed code using option 2, so that is still a mystery to me.

Except for the version that I modified in December, I didn't really do any original authoring, just some debug suggestions. So I'm not sure what the people who did that part of the code had in mind. I don't have the hardware available to test it, either. I was mostly interested in the more mundane applications, such as booting to a USB flash drive. But I would like to get all the other parts of it to work too.

Regarding the problem where your pcmcia_scsi adapter is found but not searched: When the part of code is executed that searches for Puppy, it goes through a fixed search sequence, and if Puppy is found on any media earlier than the one you want to boot to, it will select it first. The search sequence in order is: idehd, usbhd, usbflash, and finally ide or usb cd.

On the experimental path, when you get to the "Select Puppy Boot Mode" screen, you should be able to select Option 6, "Boot Media". Then you can select the CD explicitly. I did very limited testing of the experimental path, but I'll go back to it and make sure what I'm saying really occurs.

otropogo

#5 Post by otropogo »

Crash wrote:....Regarding the problem where your pcmcia_scsi adapter is found but not searched: When the part of code is executed that searches for Puppy, it goes through a fixed search sequence, and if Puppy is found on any media earlier than the one you want to boot to, it will select it first. The search sequence in order is: idehd, usbhd, usbflash, and finally ide or usb cd.
Why, after Freedos has detected and enabled the pcmcia scsi adapter and its attached CDROM, should WakePup disregard it and look on the hard drive for puppy? Obviously the user chooses the PCMCIA boot option because he wants to boot from pcmcia. not from the hard drive.

What's also nuts is that when you choose the PCMCIA boot option, Freedos loads and tries a zillion SCSI drivers that have nothing to do with pcmcia scsi. Wouldn't it make more sense to load and try just those few that are written for pcmcia host adapters?

Crash wrote:.On the experimental path, when you get to the "Select Puppy Boot Mode" screen, you should be able to select Option 6, "Boot Media". Then you can select the CD explicitly. I did very limited testing of the experimental path, but I'll go back to it and make sure what I'm saying really occurs.
Well, if you look at the options presented, PCMCIA isn't one of them.

BTW - on the topic of the missing parallel port Zip support, I would point out that all iomega removables IIRC, have to be booted from partition #4. I know for certain that that's true of ZIP disks and Bernoullis. So if WakePup2 looks for Puppy on a ZIP anywhere but sdx4, it's not going to find it. Iomega uses a proprietary low level format, and partition #4 is the only one the user can access.

User avatar
Crash
Posts: 453
Joined: Fri 09 Dec 2005, 06:34
Location: Melbourne, FL

#6 Post by Crash »

I can't comment on the trail of logic involved since I am not the original author. But did you try the "5. CD Drive" option in the Boot Media menu? If that didn't work, there is a bug to be worked out. What drive letter does your pcmcia-scsi CD ROM show up as? I note that Wakepup2's CD search code only looks for drive X, Y, or Z.

/// Edited an hour later:

The reason that Wakepup2 simply executes the same search code is that there is no communication between the pcmcia.bat program and the autoexec.bat program that it passes control back to. With a little thought, pcmcia.bat could create an environment variable that it passes back to steer autoexec.bat the right way. The problem is that I don't have any hardware here to test it out. But first things first. Let's see if you can get the pcmcia-scsi CDROM to boot before I start hacking code.

otropogo

#7 Post by otropogo »

Crash wrote:... did you try the "5. CD Drive" option in the Boot Media menu?


Tried it just now."5. CD Drive" takes me right back to the Boot Options menu, from which I selected 6. Boot Media.

When I select #1. Normal on this menu AFTER going on the little #.6 Boot Media/#5. CD Drive tour, I get exactly the same menu of pup_savexx.2fs files from the hard drive as I do when I go to Normal boot the first time round.

But, after the tour, when I select:

"1 hda1 /pup_save-301Rxxx.2fs"

Wakepup2 reports (in red letters)

"pup_400.sfs not found.Dropping out to initial-ramdisk console..."

And when (always the optimist, and six minutes older) I select:

"0 none"

it displays the same error message in red, followed by:

/bin/sh: can't access tty; job control turned off

So evidently WakePup2 WAS able to read the CDROM in the pcmcia_scsi chain, or it wouldn't be looking for pup_400.sfs

Crash wrote:What drive letter does your pcmcia-scsi CD ROM show up as? I note that Wakepup2's CD search code only looks for drive X, Y, or Z.
Not sure what you mean by drive letter.It's been too long since I've run it under Windows to recall, except for Freedos, which doesn't supply a letter. But logically, it should be "D" under DOS, especially if I don't have any pcmcia flash readers inserted..

And since Wakepup2 is obviously accessing and enabling my pcmcia_scsi card, why doesn't it enable the pcmcia flash readers as well? The Sandisk CF reader should be especially easy. I gather it's nothing but a straight-through connector to an ISA controller.

In Puppy 3.01 Retro, the only version in which I've been able to enable it, the CD-ROM drive shows up as/dev/sr0

User avatar
Crash
Posts: 453
Joined: Fri 09 Dec 2005, 06:34
Location: Melbourne, FL

#8 Post by Crash »

I hate to make you do the troubleshooting for me, but I am only attempting to emulate the pcmcia scsi CD problem with the hardware I have.

Can you edit the autoexec.bat file and go to the line that reads:

driver\SHSUCDX /D:?IDE-CD,X,,2 /d:?USB-CD,Z,,1 /QQ

and get rid of the " /QQ"?

also, add a pause command after it. So the whole thing looks like:

driver\SHSUCDX /D:?IDE-CD,X,,2 /d:?USB-CD,Z,,1
pause

This way, you will see what the CD ROM driver finds. On mine, I see some title information and then this information:

Drive Driver Unit
X: IDE-CD 0
Y: IDE-CD 1

See what you get. It will give a clue where things aren't working.

/// Edited a few hours later:

Also, what are the contents of \linld\puppy.lin on the floppy? They should look like this:

root=/dev/ram0
PMEDIA=idecd

otropogo

#9 Post by otropogo »

Crash wrote:I hate to make you do the troubleshooting for me, but I am only attempting to emulate the pcmcia scsi CD problem with the hardware I have.

Can you edit the autoexec.bat file and go to the line that reads:

driver\SHSUCDX /D:?IDE-CD,X,,2 /d:?USB-CD,Z,,1 /QQ

and get rid of the " /QQ"?

also, add a pause command after it. So the whole thing looks like:

driver\SHSUCDX /D:?IDE-CD,X,,2 /d:?USB-CD,Z,,1
pause

This way, you will see what the CD ROM driver finds. On mine, I see some title information and then this information:

Drive Driver Unit
X: IDE-CD 0
Y: IDE-CD 1

See what you get. It will give a clue where things aren't working.
OK. Done that, and now, after running option #7. PCMCIA, and "pressing any key to continue", the following message is displayed:

SHSUCDX installed.
Drives Assigned
Drive Driver Unit
Z: USB-CD 0

Press any key to continue...

Then I get the Select Puppy2 Boot Mode menu, select #6. Boot Media, then #5, CD Drive, and I can hear the CD drive power up, then the Select Puppy 2 Boot Mode menu redisplays, and I select #1 Normal, the CD drive powers up again (briefly), the screen displays "loading from USBCD" (or words to that effect), then I get the pup_savexx menu from hda, and chose #0 None.

and get the same red error message as previously:

pup_400.sfs not found. Dropping out to initial-ramdisk console...

/bin/sh: can't access tty; job control turned off

/// Edited a few hours later:
Crash wrote:IAlso, what are the contents of \linld\puppy.lin on the floppy? They should look like this:

root=/dev/ram0
PMEDIA=idecd
on my WakePup2 floppy (the unedited one generated by Puppy), these lines read:

root=/dev/ram0
PMEDIA=idehd

Should I change idehd to idecd or to usbcd?

otropogo

#10 Post by otropogo »

Crash wrote:.... what are the contents of \linld\puppy.lin on the floppy? They should look like this:root=/dev/ram0 PMEDIA=idecd
As posted previously,. the second line read

PMEDIA=idehd

when I first checked it.

But after removing the "/QQ" in autoexec.bat and rebooting, that line has now changed to:

PMEDIA=usbcd

Unfortunately, WakePup2 still can't boot from the Live CD on the pcmcia scsi chain.

While were at it, if you could tell exactly how to manually create a bootable install of Puppy 4.0 on a flash card and/or ZIP disk (copying from the LiveCD), we could give booting from those a try with WakePup2 as well.

Puppy 4.0's universal installer doesn't work at all on my IDE ZIP drive, and I doubt it works any better on USB cards.

User avatar
Crash
Posts: 453
Joined: Fri 09 Dec 2005, 06:34
Location: Melbourne, FL

#11 Post by Crash »

Overall, things look pretty good. The fact that SHSUCDX sees your CDROM is good. As you execute the program, Wakepup2 overwrites the puppy.lin file, and indeed it is updating it with what it is being told.

... But ...

I question why the pcmcia.bat program wants to call the SCSI CDROM a USB-CD, since I think it more resembles an IDE-CD (although to tell you the truth it's been a long time since I had a SCSI drive in my computer). At the risk of sounding stupid, I suggest you change the reference in pcmcia.bat to make the Kernel think it is an IDE CD drive:

Change the line in pcmcia.bat from:

devload /q driver\aspicd.sys /d:USB-CD

to:

devload /q driver\aspicd.sys /d:IDE-CD

( Just change "USB-CD" to "IDE-CD". )

Then again, maybe that won't do anything. This is getting into how the Kernel behaves with the "PMEDIA=" parameter, and this is where my knowledge starts to get pretty thin. Maybe others can comment about this.

... On the subject of putting the files manually on Flash media: I have very good luck just copying vmlinuz, initrd.gz, pup_400.sfs, zdrv_400.sfs, and a USBFLASH marker file to the Flash media, then just letting Wakepup2 do its default bootup.

... It is a little late to mention it, but from your earlier description, it looks like you have an "IDEHD" marker file on you hard drive already. It may be less confusing to temporarily delete or rename it, since that is probably why Wakepup2 initially saw the hard drive rather than the CDROM. With it out of the way, the CDROM will probably be recognized simply by selecting the "Normal Boot" mode without going through all the other strangeness.

John Doe
Posts: 1681
Joined: Mon 01 Aug 2005, 04:46
Location: Michigan, US

#12 Post by John Doe »

otropogo wrote:
Crash wrote:.... what are the contents of \linld\puppy.lin on the floppy? They should look like this:root=/dev/ram0 PMEDIA=idecd
As posted previously,. the second line read

PMEDIA=idehd

when I first checked it.

But after removing the "/QQ" in autoexec.bat and rebooting, that line has now changed to:

PMEDIA=usbcd
as you've noted, the original contents are irrelevant. puppy.lin is overwritten just after :writeoptions in the autoexec.bat, all prior to being called..
otropogo wrote:Unfortunately, WakePup2 still can't boot from the Live CD on the pcmcia scsi chain.
i'll try to tinker with it a bit also (I have one pcmci cd drive here). i think the problem was with needed himem to load the pcmcia drivers and that killed LINLD.COM or something along those lines.

the drivers for PCMCIA should be there under "driver/CARDWIZ" and the zip drivers should be under "driver/zipdrv"

if you look under config.sys there is some commented out logic that might work for pcmcia. i think the logic for the zipdrive driver is hidden in the old 1.x series wakepup. I think i have a copy around somewhere. i don't have a zip drive so never implemented that logic. would be great to see it working.

great thread. lets see how far we can push it for 4.1.

otropogo

#13 Post by otropogo »

John Doe wrote:...
otropogo wrote:Unfortunately, WakePup2 still can't boot from the Live CD on the pcmcia scsi chain.
i'll try to tinker with it a bit also ... i don't have a zip drive so never implemented that logic. would be great to see it working.

great thread. lets see how far we can push it for 4.1.
Thanks for your interest. I think being able to dry run Puppy from a PCMCIA slot or even an external ZIP, whether off the parallel port or off a SCSI pcmcia host adapter would be a great boost for Linux among people with older laptops.

I think pcmcia support on the boot floppy is the key. I managed to install Suse9.2 on this machine using Suse's pcmcia floppy to access the pcmcia_scsi chain. So it's certainly doable. And now that flash cards up to 4GB have gotten dirt cheap, that would give older notebooks much greater flexibility.

John Doe
Posts: 1681
Joined: Mon 01 Aug 2005, 04:46
Location: Michigan, US

#14 Post by John Doe »

otropogo wrote:I think pcmcia support on the boot floppy is the key.
agreed.

further more, whether the pcmcia is USB or CD or "other".

the only two laptops I have; one has a 16bit slot and the other has a 32bit slot. it's just a mess for my mind to handle between the DOS drivers and binaries as well as the binary switch on the card.

i seem to recall getting to a point where the pcmcia/usb devices would go with DOS but not with puppy.

User avatar
Aitch
Posts: 6518
Joined: Wed 04 Apr 2007, 15:57
Location: Chatham, Kent, UK

#15 Post by Aitch »

Hi Guys

I seem to recall BarryK saying that USB is a 'SCSI type' device

However, I suspect otropogo's problem is one of speed matching the boot process, as Crash appears to have figured the SCSI recognition - applause

I am in process of getting some real SCSI hardware to Barry to sort out booting to/from SCSI hardware to be fully implemented into Puppy
Apparently there is a bit of code in the kernel needs tweaking, though since you've got further than before, maybe just timing?

Those PCMCIA scsiCDs are quite a bit slower to get going [or at least used to be on my old W98 one] so there maybe a need for some 'sleeps' in there somewhere, in the scripts?

Possible?

Only other thing, on my normal scsi hardware I have to assign LUN, device number, could there be a setting jumper on the CD to be set?

Aitch

otropogo

#16 Post by otropogo »

Aitch wrote: However, I suspect otropogo's problem is one of speed matching the boot process,


Don't think so. After installation of the proper pcmcia_scsi module aha152x_cs, my SCSI drive runs just fine under Puppy 3.01 Retro.

Unfortunately, the module doesn't appear to be available for 3.01 Standard or 4.0 (not yet, at least)

Worse, Wakepup isn't working properly. In fact, both FreeDos and WakePup2 can see the adapter and drive, and Wakepup even reads the version number on the CD. There's just a glitch in the coding that causes it to fail to boot from it.
Aitch wrote: am in process of getting some real SCSI hardware to Barry to sort out booting to/from SCSI hardware to be fully implemented into Puppy


Aha! So YOU'RE the anonymous SCSI donor Barry mentions on his blog.
Unfortunately, he seems to be leaning toward dropping SCSI from Puppy altogether. If so, I think he should make an exception for pcmcia_scsi, where the host adapters are quite simple (no BIOS, no Wide SCSI, no jumpers, etc.), and the drivers are quite small.

BTW - can you tell me where you found Barry's mailing address? He needs some LS-120 disks to test and fix the LS-120 support, and I'm all set to mail him two shrink-wrapped disks, but don't have an address.

Aitch wrote:Those PCMCIA scsiCDs are quite a bit slower to get going [or at least used to be on my old W98 one] so there maybe a need for some 'sleeps' in there somewhere, in the scripts?


Are you sure you mean "PCMCIA" SCSI? And slower than what?

Non-Cardbus PCMCIA is 16 bit, so it's bound to be slower than 32 bit controllers. Ever try installing Windows from a parallel port CDROM? I've used both, and I guarantee you SCSI is faster.

On my desktops, writing to a SCSI ZIP is considerably faster even than writing to DVD-RAM.


Aitch wrote:... on my normal scsi hardware I have to assign LUN, device number, could there be a setting jumper on the CD to be set?


That's strange. On my main desktop, that I usually use with Win98SE, I've got two hard drives, two SCSI hard drives, one IDE DVD burner, and one SCSI ZIP drive installed. The SCSI drives run off an Adaptec AHA2940UW host adapter, and I've never had to input any of thos settings. You just have to make sure no devices are jumpered to the same ID.

I've just booted Puppy 4.0, and was going to demonstrate how smoothly the scsi support works, but ran into a big catch - my USB IR mouse doesn't work, my serial touchpad mouse (on the keyboard) doesn't work. The wired USB mouse I then added doesn't work either.

I rebooted the system and added a PS2 mouse, and now the keyboard itself doesn't work, so I couldn't even get through the Xorg configuration!

Happily, removing the USB IR mouse (leaving the wired USB mouse and the PS2 mouse on, got the keyboard back. But now, ONLY the touchpad mouse works, not the USB or the PS/2! Good enuff...

One last sour note: I tried changing the mouse type in Setup from serial to USB. I then restarted X, and am totally mouseless again....
,
On the positive side, Pmount shows both of the SCSI hard drives and the SCSI ZIP100 drive, and there's no problem accessing any of them. The ZIP is shown as sde4, the hard drives as sdf1 and sdg1



Even FreeDOS and WakePup2 can read the essential info off the SCSI chain. When I installed Suse2 with the same setup, it didn't require me to input any of it. I do remember that in the past, there have been some distros, such as Red Hat, and maybe Mandriva, where the PCMCIA boot floppies weren't set up properly, and memory location had to be given IIRC.

But with an adapter like the apa1460x, it's simple - the chain is LUN 0, the adapter is ID 0, and the devices' IDs are read according to their settings. If two devices are set to the same number, your system will likely lock up. Certainly the SCSI chain will lock up. But that's a pretty unlikely occurence on a laptop.

I don't know what happens if you try to run two SCSI chains off two pcmcia host adapters in one laptop. But I think that's pretty academic.
OTOH, being able to run a ZIP drive, a CDROM burner, and a SCSI scanner of the one pcmcia slot can be useful, as can connecting a SCSI hard drive. I've done all of that under Windows98SE. The SCSI chain is pretty useful if you don't have USB.

User avatar
Aitch
Posts: 6518
Joined: Wed 04 Apr 2007, 15:57
Location: Chatham, Kent, UK

#17 Post by Aitch »

Aha! So YOU'RE the anonymous SCSI donor Barry mentions on his blog.
Unfortunately, he seems to be leaning toward dropping SCSI from Puppy altogether. If so, I think he should make an exception for pcmcia_scsi, where the host adapters are quite simple (no BIOS, no Wide SCSI, no jumpers, etc.), and the drivers are quite small.

BTW - can you tell me where you found Barry's mailing address? He needs some LS-120 disks to test and fix the LS-120 support, and I'm all set to mail him two shrink-wrapped disks, but don't have an address.

Anonymous, me?

No, I know Barry posted about some anonymous Aussie donor, but my efforts became high profile on the forum when getting ebay payments/collections in Oz went sour, & Barry posted comments about how p**d off he was!

Seems to be leaning towards.............he gives the impression that he was a postie & SCSI bit him on the A*se :lol:

He needs all the encouragement we can give him to support ALL SCSI, booting and all - after all, apparently it's just enabling already included kernel modules
Sage [another SCSI user] has posted elsewhere, that it would enable the loads of commercial SCSI kit currently being landfilled to be put back into use, often with lifetime guarantees on H/Ds

Re the LUN thing, I think we are on the same understanding, I probably didn't explain it very well, but it does seem to depend on the SCSI bias as to whether the jumpers or autoassign is used, in my experience

When I said slower I was referring to an old SCSI-1 external CD writer I had, which had a pcmcia scsi adapter, and I installed a pcmcia to pci adapter into a desktop - I believe this was 8bit SCSI!
It was pitifully slow compared to my IDE CD, but that was 40x and the external was only 2x so probably explains it, though it did take a long time to spin up as I recall
I have SCSI working super well in 2.14/&R/3.01, haven't tried 4 yet but hope it still works as I've just acquired a Trimm 8bay array which connects via 68p external so I can install some more drives :D
The SCSI chain is pretty useful if you don't have USB.
It's pretty useful if you do, too
I also have firewire400
I find both of them chain better than usb over distance, but keep meaning to get one of those active extension cables
I suppose really I just like playing with all sorts of old kit - space is running short, as it tends to need lots of bench/floorspace and wifi is too insecure :lol:

Address to send LS120s;
Barry Kauler
PO Box 359
Perenjori
WA 6620
Australia

Good Luck

Aitch :)

otropogo

#18 Post by otropogo »

Aitch wrote:[... Barry ... needs all the encouragement we can give him to support ALL SCSI, booting and all - ... it would enable the loads of commercial SCSI kit currently being landfilled to be put back into use, often with lifetime guarantees on H/Ds.
I fully agree, but if the bit about huge driver space requirements is apt, then perhaps a buffet type approach for WakePup2 and other boot floppies (like GRUB) would make sense.

BTW - for years I've been using a number of SCSI drives, all acquired used. And only one has ever burned out. And that was probably due to inadequate cooling arrangements on my part. These drives were the Rolls Royce line of hard drives, designed to run constantly for years. For the type of personal desktop use most of us would impose, they're probably more reliable than most new IDE drives marketed today.
Aitch wrote:....When I said slower I was referring to an old SCSI-1 external CD writer I had, which had a pcmcia scsi adapter, and I installed a pcmcia to pci adapter into a desktop - I believe this was 8bit SCSI!

It was pitifully slow compared to my IDE CD, but that was 40x and the external was only 2x so probably explains it, though it did take a long time to spin up as I recall.
Absolutely! I've had some experience with a 2x SCSI CDROM on a pcmcia adapter, and I would say it was somewhat like a horrific flashback to loading programs off cassette tape.

But nowadays people are tossing high end external SCSI CD drives, like my Sony Spressa Professional CD-RW burner, probably because it doesn't do DVD. I paid $5 in a thrift store for mine, complete with manuals, software,scsi_scsi and scsi_usb cables.. Hell, the enclosure and power supply are worth much more than that.

Thanks for Barry's address. I got it from him too, but it never hurts to have a reference check, as typos do happen. Will be mailing him a disk today.

User avatar
Aitch
Posts: 6518
Joined: Wed 04 Apr 2007, 15:57
Location: Chatham, Kent, UK

#19 Post by Aitch »

Oh dear, it's not looking good for the LS120.....

http://www.puppylinux.com/blog/?page=1

posted 22nd May [though in UK it's still 21st] so hot off the press!!

I'd respond if possible, it is a blog!

Aitch

otropogo

#20 Post by otropogo »

Aitch wrote:Oh dear, it's not looking good for the LS120.....

http://www.puppylinux.com/blog/?page=1

posted 22nd May [though in UK it's still 21st] so hot off the press!!

I'd respond if possible, it is a blog!

Aitch
Not possible. After composing the entry below, I had to copy a security code (done), enter a password (that too), and supply the name of the Puppy Mascot (who the hell knows?), so my entry was rejected.

And besides, the input window for his blog is even more awful than this one.

If Barry and his clan are interested in my comments, they'll have to do it here below:


I disagree strongly with the contention that ZIP or LS-120 installation is worthless to potential Puppy users. To begin by countering Barry's size and speed arguments:

a) using the current kernel of Puppy 4.0, there is room on even a ZIP 100 for an install, so long as you use vfat format

b) the LS-120 is 20% larger, and three times faster than a floppy

c) yes, both are very slow to load and to run off disk, however there are laptops that allow no other way to load Puppy (or any other Linux) due to lack of internal IDE CD-ROM or USB. Testing the live CD is even more of a challenge on such laptops.

With enough memory to run in RAM, a bootable ZIP or LS-120 disk (whether self-booting or by means of a boot floppy) would be most useful on such machines (assuming the glitch that fails to load Puppy into RAM from the LS-120 is fixed)

I myself have stuck with Puppy from ver. 2.17 only because I was able to install and run it from an LS-120, even though, because of the above mentioned glitch, it ran only off the disk.

I finally installed Puppy 3.01 Retro to the hard drive, but still face the same problems when attempting to try newer versions of Puppy on that laptop.

And for the rest:

Yes, testing new Puppies on old laptops would certainly be easier using removable flash memory.

Unfortunately, there appears to be no functional boot floppy support for pcmcia at present, whether for flash readers or pcmcia_scsi cards (and I'm speaking here of devices that run fine with the proper modules once Puppy is loaded ).

So if you have a laptop without either bootable USB or internal CD drive, you won't be able to use flash memory to try a new version.

And apparently, pcmcia support on a boot floppy is not easy to get working. Knoppix never had it, and DSL has a floppy pcmcia image that doesn't work. And this even though it's has been done successfully before - most recently in Suse 9.2

So I strongly suggest not throwing out any resources that might enable potential linux users to get Puppy up on their old laptops. LS-120 support makes Puppy unique. Zip parallel port boot support would be another unique feature to crow about.

And besides, I just airmailed a brand-spanking new LS-120 disk to Barry this morning. It's heading toward Canada's highest airport right now, and should be reaching him in a mere ten days or so.

OTOH, if you're NOT going to fix it, then yes - get rid of the menus and documentation that say you can do it, so unsuspecting users don't waste hours and days knocking their heads against a wall.

But my final word is, are you also going to give up the claim to run on low-powered, low-resource PCs? If I had a 2GB 2MHz laptop, with USB, I doubt I'd be running Puppy on it. Mandriva, Ubuntu, Suse, and Fedora are much more likely candidates. And there's plenty of competition in the so-called lightweight camp too - Feather, Antix, Xubuntu, Vector Lite, DSL. ...

It seems crazy to abandon precisely the qualities that established Puppy's niche, just because most of the regular users don't need them anymore. There are lots of old machines like mine out there. The outfit in Calgary, Alberta, from which I bought my CF-25 Toughbook had an office in Panama City for two years through which they sold off much of the discarded hardware they collected in Alberta.

Who knows how many thousands of potential Puppy users are out there in the third world with their old laptops, just waiting for a broadband connection...? Hell, Cuba alone must have thousands of them. I know I left a little Ambra there a few years back.

Post Reply