TUHS

tuhs@tuhs.org

37 participants
6478 discussions

Re: [TUHS] Is it time to resurrect the original dsw (delete with switches)?

by norman＠oclsc.org

Not to get into what is soemthing of a religious war, but this was the paper that convinced me that silent data corruption in storage is worth thinking about: http://www.cs.toronto.edu/~bianca/papers/fast08.pdf A key point is that the character of the errors they found suggests it's not just the disks one ought to worry about, but all the hardware and software (much of the latter inside disks and storage controllers and the like) in the storage stack. I had heard anecdotes long before (e.g. from Andrew Hume) suggesting silent data corruption had become prominent enough to matter, but this paper was the first real study I came across. I have used ZFS for my home file server for more than a decade; presently on an antique version of Solaris, but I hope to migrate to OpenZFS on a newer OS and hardware. So far as I can tell ZFS in old Solaris is quite stable and reliable. As Ted has said, there are philosophical reasons why some prefer to avoid it, but if you don't subscribe to those it's a fine answer. I've been hearing anecdotes since forever about sharp edges lurking here and there in BtrFS. It does seem to be eternally unready for production use if you really care about your data. It's all anecdotes so I don't know how seriously to take it, but since I'm comfortable with ZFS I don't worry about it. Norman Wilson Toronto ON PS: Disclosure: I work in the same (large) CS department as Bianca Schroeder, and admire her work in general, though the paper cited above was my first taste of it.

3 years, 8 months

Re: [TUHS] Can't break line warning

by Douglas McIlroy

This may be due to logic similar to that of a classic feature that I always deemed a bug: troff begins a new page when the current page is exactly filled, rather than waiting until forced by content that doesn't fit. If this condition happens at the end of a document, a spurious blank page results. Worse, if the page header happens to change just after the exactly filled page, the old heading will be produced before the new heading is read. Doug

3 years, 8 months

Systematic approach to command-line interfaces

by Douglas McIlroy

> fork() is a great model for a single-threaded text processing pipeline to do > automated typesetting. (More generally, anything that is a straightforward > composition of filter/transform stages.) Which is, y'know, what Unix is *for*. > It's not so great for a responsive GUI in front of a multi-function interactive program. "Single-threaded" is not a term I would apply to multiple processes in a pipeline. If you mean a single track of data flow, fine, but the fact that that's a prevalent configuration of cooperating processes in Unix is an artifact of shell syntax, not an inherent property of pipe-style IPC. The cooperating processes in Rob Pike's 20th century window systems and screen editors, for example, worked smoothly without interrupts or events - only stream connections. I see no abstract distinction between these programs and "stuff people play with on their phones." It bears repeating, too, that stream connections are much easier to reason about than asynchronous communication. Thus code built on streams is far less vulnerable to timing bugs. At last a prince has come to awaken the sleeping beauty of stream connections. In Go (Pike again) we have a widely accepted programming language that can fully exploit them, "[w]hich is, y'know, what Unix is 'for'." (If you wish, you may read "process" above to include threads, but I'll stay out of that.) Doug

3 years, 8 months

Who is that in the framed photograph near the ceiling?

by norman＠oclsc.org

Steve Simon: once again i am taken aback at the good taste of the residents of the unix room. As a whilom denizen of that esteemed playroom, I question both the accuracy and the relevance of that metric. Besides, what happened to the sheep shelf? Was it scrubbed away after I left? And, Ken, whatever happened to Dolly the Sheep (after she was hidden to avoid upsetting visitors)? Norman Wilson Toronto ON No longer a subscriber to sheep! magazine

3 years, 8 months

Who is that in the framed photograph near the ceiling?

by Steve Simon

great picture. once again i am taken aback at the good taste of the residents of the unix room. -Steve

3 years, 8 months

Who is that in the framed photograph near the ceiling?

by William Cheswick

> I don't think anyone knows. Nobody relevant, I believe. > > -rob I understand that Dave Presotto bought that photo at a garage sale for $1. The photo hung in the Unix Room for years, at one point labeled “Peter Weinberger.” One day I removed it from its careful mounting and scanned in the photo. It bore the label “what, no steak?” The photo was stolen from a wall sometime after I left. The scanned image is at https://cheswick.com/ches/tmp/whatnosteak.jpeg ches

3 years, 8 months

Re: [TUHS] Depraz/Logitech Digimouse manual

by John Foust

At 07:24 PM 8/6/2021, Rob Pike wrote: >I don't think anyone knows. Nobody relevant, I believe. Indeed, a clipped and cleaned version in reverse image search on Google, Bing, Tineye and Yandex found nothing. Feels like a head shot for a theater major. - John

3 years, 8 months

Re: [TUHS] Depraz/Logitech Digimouse manual

by Rob Pike

I sent a picture (actually two at different resolutions; keep reading) to the list, but being images they are larger than the address space of a PDP-11 so not allowed here. Is it really necessary to have such a low message size limit in an era when I can buy a terabyte of storage for less than a hundred bucks? Here is a Google Drive link, for the adventurous. 20180123-UnixSkeleton.jpg <https://drive.google.com/file/d/1aS8ZmzwPUawIa8WXGoXOK9jDiYtJETGG/view?usp=…> -rob On Sat, Aug 7, 2021 at 7:44 AM Rob Pike <robpike(a)gmail.com> wrote: > I sent a higher-res version in which you can read all the text but it was > "moderated". > > This is the Unix room as of the year 2000 or so. > > -rob > > > On Sat, Aug 7, 2021 at 4:34 AM ron minnich <rminnich(a)gmail.com> wrote: > >> The story of the mice, one of which I gave to John: >> >> I ran a program called FAST-OS for LANL/Sandia for 6 years starting >> 2005. Think of it as "Plan 9 on petaflop supercomputers" -- it may >> seem strange now, but in that era when some top end systems ran custom >> kernels, there was a strong case to be made that plan 9 was a good >> choice. By 2011, of course, the Linux tsunami had swept all before it, >> which is why you no longer hear about custom HPC kernels so much -- >> though in some places they still reign. In any event, this program >> gave me 6 years to work with "the Unix room", or what was left of it. >> I had been in the Unix Room in 1978, and even met Dennis, so this >> prospect was quite a treat. >> >> We funded Charles Forsyth to write the amd64 compilers for Plan 9, >> which if you used early Go you ran into (6c 6a 6l); we also funded the >> amd64 port of Plan 9 (a.k.a. k10) as well as the port to Blue Gene. >> That amd64 port is still out and about. You can find the Blue Gene >> kernel on github. >> >> I had lots of fun spending time in the Unix room while working with >> the late Jim McKie, and others. I saw the tail end of the traditions. >> They had cookie day once a week, if memory serves, on Thursday at 3. I >> got to see the backwards-running clock, Ken's chess trophies, his >> pilot's license, pictures of Peter everywhere, a "Reagan's view of the >> world" map, the American Legion award for Telstar (which was rescued >> from a dumpster!), and so on. The "Unix room" was more than one room, >> all built on a raised floor, as I assume it was former old school >> machine room space. If memory serves, it filled the entire width of >> the end of the top floor of the building it was in (4th floor?) -- >> maybe 50 ft x 50 ft -- maybe a bit more. There was a room with desks, >> and a similar-sized room with servers, and a smaller room containing a >> lab-style sink, a very professional cappucinno machine, decades of old >> proceedings, and a sofa. I fixed the heavy-duty coffee grinder one >> year; for some reason the Italian company that produced it had seen >> fit to switch BOTH hot and neutral, and the fix was to only switch >> hot, as the neutral switch had failed; I guess in the EU, with 220v, >> things are done differently. >> >> It was fun being there. A few years later the whole room, and all its >> history, was trashed, and replaced with what Jim called a "middle >> management wxx dream" (Jim was never at a loss for words); Jim found >> some yellow Police crime scene tape and placed it in front of the >> doors to the new space. It was redubbed "the innovation space" or some >> such, and looked kind of like an ikea showroom. Much was lost. I tried >> to find a way to save the contents of the room; I had this dream of >> recreating it at Google, much as John Wanamaker's office was preserved >> in Philadelphia for so many decades, but I was too late. I have no >> idea where the contents are now. Maybe next to the Ark. >> >> One day in 2008 or so jmk took me for a tour of the buildings, and we >> at one point ended up high in the top floor of what I think was >> Building One (since torn down?), in what used to be Lab Supply. Nobody >> was there, and not much supply was there either. Finally somebody >> wandered in, and Jim asked where everyone was. "Oh, they closed lab >> supply, maybe 4 years ago?" >> >> Bell Labs had seen hard times since the Lucent split, and it was clear >> it had not quite recovered, and Lab Supply was just one sign of it. I >> think the saddest thing was seeing the visitor center, which I first >> saw in 1976. In 1976, it was the seat of the Bell System Empire, and >> it was huge. There was a map of the US with a light lit for every >> switching office in the Bell Labs system. There was all kinds of Bell >> Labs history in the visitor center museum. >> >> The museum had shrunk to a much smaller area, and felt like a closet. >> The original transistor was still there in 2010, but little else.The >> library was, similarly, changed: it was dark and empty, I was told. >> Money was saved. At that time, Bell Labs felt large, strangely quiet, >> and emptied of people. It made me think of post-sack Rome, ca. 600, >> when its population was estimated to be 500. I have not been back >> since 2011 so maybe things are very different. It would be nice if so. >> >> As part of this tour, Jim gave me 3 depraz mice. I took one, gutted >> it, (sorry!), and filled its guts with a USB mouse innards, and gave >> it back to Jim. He then had a Depraz USB mouse. jmk's mouse did not >> have any lead in it, as John's did, however. The second I gave to >> someone at Google who had worked at the labs back in the day. The >> third mouse I gave to John, and he made it live again, which is cool. >> >> In spite of their reputation, I found Depraz mice hard to use. I have >> gone through all kinds of mice, and am on an evoluent, and as far as >> Depraz go, I guess "you had to be there". I don't recall if jmk used >> his "usb depraz" or it ended up on a shelf. Sadly, I can no longer ask >> him. >> >> I'll be interested to see what John thinks of the Depraz. >> >> ron >> >> On Fri, Aug 6, 2021 at 9:52 AM John Floren <john(a)jfloren.net> wrote: >> > >> > Ah, right. I opened the mouse because one of the encoders didn't seem >> to be working (it worked fine again this morning, who knows...) and >> discovered that there was something duct taped inside the plastic shell: >> > >> > http://jfloren.net/content/depraz/inside.jpg >> > >> > Peeling back the tape, I saw what I first took to be chunks of >> flattened beer cans: >> > >> > http://jfloren.net/content/depraz/reveal.jpg >> > >> > A closer look showed that they were the wrappers which cover the corks >> of wine bottles. Up into the 1980s, these were made out of lead, and by >> flattening five of them, a previous owner of the mouse was able to add >> quite a bit of extra weight to it: >> > >> > http://jfloren.net/content/depraz/wrapper.jpg >> > >> > >> > john >> > >> > ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ >> > >> > On Friday, August 6th, 2021 at 9:34 AM, ron minnich <rminnich(a)gmail.com> >> wrote: >> > >> > > john, don't forget to mention the beer can >> > > >> > > On Fri, Aug 6, 2021 at 9:29 AM John Floren john(a)jfloren.net wrote: >> > > >> > > > I stuck an Arduino on it and with surprisingly little code I have >> it acting like a 3-button USB mouse. >> > > > >> > > > The only problem is that the pointer doesn't move smoothly. It does >> OK left-to-right, and can move down pretty well, but going up is a problem. >> I think pushing the mouse forward tends to move the ball away from the >> Y-axis wheel, and the old spring on the tensioner just doesn't have the >> gumption to hold that heavy ball bearing in any more. >> > > > >> > > > john >> > > > >> > > > ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ >> > > > >> > > > On Wednesday, August 4th, 2021 at 9:12 PM, ron minnich >> rminnich(a)gmail.com wrote: >> > > > >> > > > > John, you can see that "stick a bird on it" -> "stick an arduino >> on >> > > > > >> > > > > it" -> "stick a pi on it" has gone as you once predicted :-) >> > > > > >> > > > > On Wed, Aug 4, 2021 at 8:59 PM John Floren john(a)jfloren.net >> wrote: >> > > > > >> > > > > > ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ >> > > > > > >> > > > > > On Wednesday, August 4th, 2021 at 6:12 PM, Henry Bent >> henry.r.bent(a)gmail.com wrote: >> > > > > > >> > > > > > > On Wed, 4 Aug 2021 at 20:52, John Floren john(a)jfloren.net >> wrote: >> > > > > > > >> > > > > > > > Having just been given a Depraz mouse, I thought it would >> be fun to get it working on my modern computer. Since the DE9 connector is >> male rather than female as you usually see with serial mice, and given its >> age, I speculate that it might have a custom protocol; in any rate, >> plugging it into a USB-serial converter and and firing up picocom has given >> me nothing. >> > > > > > > > >> > > > > > > > Does anyone have a copy of a manual for it, or more >> information on how to interface with it? If I knew how it was wired and >> what the protocol looked like, I expect I could make an adapter pretty >> trivially using a microcontroller. >> > > > > > > >> > > > > > > This might be of some help? >> > > > > > > >> > > > > > > >> https://www.vcfed.org/forum/forum/technical-support/vintage-computer-hardwa… >> > > > > > > >> > > > > > > -Henry >> > > > > > >> > > > > > This looks great, thank you! >> > > > > > >> > > > > > john >> >

3 years, 8 months

Re: [TUHS] Depraz/Logitech Digimouse manual

by norman＠oclsc.org

The mouse with wine-bottle lead foil in the top may have been my fault. I did that to two of them--at home and in my office--because I found a little more pressure made the ball track better. I've never been an alcohol-consumer; the lead came from a friend, who used to save it (back in the 1980s) to mail to Republicans. Apparently he had, many years before, registered to vote in a Republican primary solely to oppose a particularly-poor candidate. That somehow got him on a GOP mailing list that sent him endless funding appeals with post-paid envelopes. He used to fill the envelopes with lead and drop them in the mail, in the hope that he would cost the party even more in excess postage than they were already spending to send the funding pitches. By the time I was at Bell Labs, he had moved to Canada, and was no longer receiving unwanted political funding pitches, but he was glad to save a few bits of lead for me when I thought of the trick and asked him. Only too glad, it turned out; he kept saving it and saving it and saving it even though I neither needed nor wanted any more. Eventually I managed to get the message through to him. He has since moved back to the US. He is still fond of wine. I don't know what he does with the cork wrappers. Norman Wilson Toronto ON

3 years, 8 months

Re: [TUHS] Depraz/Logitech Digimouse manual

by John Floren

I uploaded the high-resolution one to https://jfloren.net/content/unix_skeleton.jpg if anyone wants to check it out in all its glory. Thanks, Rob, this is a great picture. I don't think things were *too* different by the time I visited for IWP9 in 2007, but it's been a long time and I guess I didn't take any pictures then. john ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ On Friday, August 6th, 2021 at 2:44 PM, Rob Pike <robpike(a)gmail.com> wrote: > I sent a higher-res version in which you can read all the text but it was "moderated". > > This is the Unix room as of the year 2000 or so. > > -rob > > On Sat, Aug 7, 2021 at 4:34 AM ron minnich <rminnich(a)gmail.com> wrote: > >> The story of the mice, one of which I gave to John: >> >> I ran a program called FAST-OS for LANL/Sandia for 6 years starting >> 2005. Think of it as "Plan 9 on petaflop supercomputers" -- it may >> seem strange now, but in that era when some top end systems ran custom >> kernels, there was a strong case to be made that plan 9 was a good >> choice. By 2011, of course, the Linux tsunami had swept all before it, >> which is why you no longer hear about custom HPC kernels so much -- >> though in some places they still reign. In any event, this program >> gave me 6 years to work with "the Unix room", or what was left of it. >> I had been in the Unix Room in 1978, and even met Dennis, so this >> prospect was quite a treat. >> >> We funded Charles Forsyth to write the amd64 compilers for Plan 9, >> which if you used early Go you ran into (6c 6a 6l); we also funded the >> amd64 port of Plan 9 (a.k.a. k10) as well as the port to Blue Gene. >> That amd64 port is still out and about. You can find the Blue Gene >> kernel on github. >> >> I had lots of fun spending time in the Unix room while working with >> the late Jim McKie, and others. I saw the tail end of the traditions. >> They had cookie day once a week, if memory serves, on Thursday at 3. I >> got to see the backwards-running clock, Ken's chess trophies, his >> pilot's license, pictures of Peter everywhere, a "Reagan's view of the >> world" map, the American Legion award for Telstar (which was rescued >> from a dumpster!), and so on. The "Unix room" was more than one room, >> all built on a raised floor, as I assume it was former old school >> machine room space. If memory serves, it filled the entire width of >> the end of the top floor of the building it was in (4th floor?) -- >> maybe 50 ft x 50 ft -- maybe a bit more. There was a room with desks, >> and a similar-sized room with servers, and a smaller room containing a >> lab-style sink, a very professional cappucinno machine, decades of old >> proceedings, and a sofa. I fixed the heavy-duty coffee grinder one >> year; for some reason the Italian company that produced it had seen >> fit to switch BOTH hot and neutral, and the fix was to only switch >> hot, as the neutral switch had failed; I guess in the EU, with 220v, >> things are done differently. >> >> It was fun being there. A few years later the whole room, and all its >> history, was trashed, and replaced with what Jim called a "middle >> management wxx dream" (Jim was never at a loss for words); Jim found >> some yellow Police crime scene tape and placed it in front of the >> doors to the new space. It was redubbed "the innovation space" or some >> such, and looked kind of like an ikea showroom. Much was lost. I tried >> to find a way to save the contents of the room; I had this dream of >> recreating it at Google, much as John Wanamaker's office was preserved >> in Philadelphia for so many decades, but I was too late. I have no >> idea where the contents are now. Maybe next to the Ark. >> >> One day in 2008 or so jmk took me for a tour of the buildings, and we >> at one point ended up high in the top floor of what I think was >> Building One (since torn down?), in what used to be Lab Supply. Nobody >> was there, and not much supply was there either. Finally somebody >> wandered in, and Jim asked where everyone was. "Oh, they closed lab >> supply, maybe 4 years ago?" >> >> Bell Labs had seen hard times since the Lucent split, and it was clear >> it had not quite recovered, and Lab Supply was just one sign of it. I >> think the saddest thing was seeing the visitor center, which I first >> saw in 1976. In 1976, it was the seat of the Bell System Empire, and >> it was huge. There was a map of the US with a light lit for every >> switching office in the Bell Labs system. There was all kinds of Bell >> Labs history in the visitor center museum. >> >> The museum had shrunk to a much smaller area, and felt like a closet. >> The original transistor was still there in 2010, but little else.The >> library was, similarly, changed: it was dark and empty, I was told. >> Money was saved. At that time, Bell Labs felt large, strangely quiet, >> and emptied of people. It made me think of post-sack Rome, ca. 600, >> when its population was estimated to be 500. I have not been back >> since 2011 so maybe things are very different. It would be nice if so. >> >> As part of this tour, Jim gave me 3 depraz mice. I took one, gutted >> it, (sorry!), and filled its guts with a USB mouse innards, and gave >> it back to Jim. He then had a Depraz USB mouse. jmk's mouse did not >> have any lead in it, as John's did, however. The second I gave to >> someone at Google who had worked at the labs back in the day. The >> third mouse I gave to John, and he made it live again, which is cool. >> >> In spite of their reputation, I found Depraz mice hard to use. I have >> gone through all kinds of mice, and am on an evoluent, and as far as >> Depraz go, I guess "you had to be there". I don't recall if jmk used >> his "usb depraz" or it ended up on a shelf. Sadly, I can no longer ask >> him. >> >> I'll be interested to see what John thinks of the Depraz. >> >> ron >> >> On Fri, Aug 6, 2021 at 9:52 AM John Floren <john(a)jfloren.net> wrote: >>> >>> Ah, right. I opened the mouse because one of the encoders didn't seem to be working (it worked fine again this morning, who knows...) and discovered that there was something duct taped inside the plastic shell: >>> >>> http://jfloren.net/content/depraz/inside.jpg >>> >>> Peeling back the tape, I saw what I first took to be chunks of flattened beer cans: >>> >>> http://jfloren.net/content/depraz/reveal.jpg >>> >>> A closer look showed that they were the wrappers which cover the corks of wine bottles. Up into the 1980s, these were made out of lead, and by flattening five of them, a previous owner of the mouse was able to add quite a bit of extra weight to it: >>> >>> http://jfloren.net/content/depraz/wrapper.jpg >>> >>> >>> john >>> >>> ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ >>> >>> On Friday, August 6th, 2021 at 9:34 AM, ron minnich <rminnich(a)gmail.com> wrote: >>> >>> > john, don't forget to mention the beer can >>> > >>> > On Fri, Aug 6, 2021 at 9:29 AM John Floren john(a)jfloren.net wrote: >>> > >>> > > I stuck an Arduino on it and with surprisingly little code I have it acting like a 3-button USB mouse. >>> > > >>> > > The only problem is that the pointer doesn't move smoothly. It does OK left-to-right, and can move down pretty well, but going up is a problem. I think pushing the mouse forward tends to move the ball away from the Y-axis wheel, and the old spring on the tensioner just doesn't have the gumption to hold that heavy ball bearing in any more. >>> > > >>> > > john >>> > > >>> > > ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ >>> > > >>> > > On Wednesday, August 4th, 2021 at 9:12 PM, ron minnich rminnich(a)gmail.com wrote: >>> > > >>> > > > John, you can see that "stick a bird on it" -> "stick an arduino on >>> > > > >>> > > > it" -> "stick a pi on it" has gone as you once predicted :-) >>> > > > >>> > > > On Wed, Aug 4, 2021 at 8:59 PM John Floren john(a)jfloren.net wrote: >>> > > > >>> > > > > ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ >>> > > > > >>> > > > > On Wednesday, August 4th, 2021 at 6:12 PM, Henry Bent henry.r.bent(a)gmail.com wrote: >>> > > > > >>> > > > > > On Wed, 4 Aug 2021 at 20:52, John Floren john(a)jfloren.net wrote: >>> > > > > > >>> > > > > > > Having just been given a Depraz mouse, I thought it would be fun to get it working on my modern computer. Since the DE9 connector is male rather than female as you usually see with serial mice, and given its age, I speculate that it might have a custom protocol; in any rate, plugging it into a USB-serial converter and and firing up picocom has given me nothing. >>> > > > > > > >>> > > > > > > Does anyone have a copy of a manual for it, or more information on how to interface with it? If I knew how it was wired and what the protocol looked like, I expect I could make an adapter pretty trivially using a microcontroller. >>> > > > > > >>> > > > > > This might be of some help? >>> > > > > > >>> > > > > > https://www.vcfed.org/forum/forum/technical-support/vintage-computer-hardwa… >>> > > > > > >>> > > > > > -Henry >>> > > > > >>> > > > > This looks great, thank you! >>> > > > > >>> > > > > john

3 years, 8 months

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

TUHS