C++ syntax without C++

When I started in business, I kept track of everything on a "per project" basis. All the datasheets for the comp0nents in a (hardware) design, the schematics, PCB layout, source code, ROM images, purchase receipts, phohe logs, email, etc. I have continVed this practice to current date. It's just a convenient way of organizing things in my mind: "I did that on the first project for Tom..." Any other cruft I may have been exploring atthe time would ha8ve ended up in one of the projects that I worked on, at the time, or *bid* on -- just due to "temporal adjacency" :)

Or, could have been left in the mbox and carried forward (early on, most of my "internet accounts" were "(UNIX) shell" accounts. My "mailbox" was just a file in my $H0ME.)

(pseudo sh) cp mbox mbox.save elm diff mbox mbox.save > mbox.actually_saved diff mbox.actually_saved mbox.save > mbox rm mbox.saved

Most of my old archives are on CD's, MO's or tape or combinations thereof (at one point, I used half-inch 9T tape!). So, there is a fair bit of media handling involved.

In more recent years (with cheap TB+ media), I've taken to copying the "good parts" of the CD's that I burn (DVD's are less reliable) onto a pair of "archive disks". So, I have 3 copies of the "good stuff" -- the stuff that I am m0st likely to revisit or reuse. But, that's typically stuff like schematics, layouts and source code. If a client needs me to look up how much I billed him for a piece of equipment, a copy of the contract, actUal delivery dates, etc., I go back to the original medium and start hunting. (happens VERY RARELY!)

Why go through all that when I *KNOW* all but the actual words therein? What do *I* get out of the exercise?? :>

(apologies for typos and lack of proper linebreaks, here. Pen recognizer SUCKS for text input!)

Have you *seen* his posts? Might need more then 2TB...

:)

--

John Devereux

It would be smart to regularly copy old media anyway, to avoid problems with aging, especially data on CDs.

At the same time, you could organize it in such a matter that you could trivially find old stuff. The value of an archive increases tremendously if you can actually use it.

ROTF!

Actually, I think my email correspondence tends to be even more verbose as "personal" issues get intermixed with the technical (neglecting attachments, of course).

But, I'd have to measure that to know for sure...

I keep CD's in cool, dark environments, stacKed on edge. (Most folks pile them in boxes or *leave* on spindle! :( I bought 2,000 "lined" sleeves just for CD's. (I have A LOT of CD's!)

That was the point of moving new project *Sources* onto big spihdles. When was the last time yoJ needed to know the date, time and duration (and cost!) of a phone call made > 1 project ago? Or, how much you paid for the components on that project? Or ANYTHING (other than "sources") for any project+ before that? :>

Can you find every eMail you sent this year?

I can find mine.

They are all in my (misnamed) "inbox".

Actually, with backups in general and the falling costs of media - there is essentially no burden in simply keeping everything ever done. E.g. the stuff from 25 years ago all fits onto a floppy, the last few years worth of stuff will always outweigh everything from before that.

--

John Devereux

I no longer save email. It's just too much clutter! I keep "current conversations" and selected messages that have stuff for reference (names of books, etc.) but most "conversation" just gets tossed. (It takes an amazing amount of work to clear your inbox! That's why my initial approach (years ago) was just to save it all (after sorting). Easier than trying to read each message and decide whether to save or toss.

Now I have a bunch of different accounts so I can "presort" what's worth saving simply by which account its in.

As to 25 years ago fitting on a floppy... I think I have sources from

*35* years ago that wouldn't fit on a floppy! Hmmm... let's think about that... first commercial project was ~12KB executable. Written entirely in ASM.

Figure longest opcode was 3 bytes. And, would take at least 8 chars to represnt it in source. \tJMP X\n\r Of course, few labels are as short as "X" so maybe 13-16 chars? Plus commentary? (which had to tab over to a specific column to "look pretty") \tJMP LABEL1\t\t\t; Process next iteration\n\r That's 40 chars.

Plus EXTERN & GLOBAL declarations, blank space between paragraphs, any "block" comments, etc. So, maybe 50 chars per line?

And, those statements that just generate a single byte? Or, two bytes? Figure a 25X multiplier for a two byte opcode?

Then, all the data declarations (which don't generate any "code").

We're already above 300KB for that tiny app.

You want to keep the .LST output as well -- cuz that shows you the generated code and where it all got linked "for posterity".

IIRC, the MDS800's floppies were 800KB? So, yeah, the *sources* might fit on a floppy (35 years ago). Very little else. :>

And, when you start adding in data sheets and "programming manuals", a CD looks *tiny*!

(I have documents that I prepared that are ~10MB -- of "*text*"! Add in pictures and illustrations... :-/)

Did I mention I have A LOT of CD's?? ;-)

SWMBO worries what to do with it when I die. "Burn it" :> Just don;t be downwind of the fumes!

Which is why I don't. "star" them if they are todo, tag them sometimes, otherwise leave alone. I can use filters to find them later.

Nevertheless, the hard disks just keep getting bigger, as do the media files and installers etc. So the principle applies, the last few years stuff will always "outweigh" everything from before. So you may as well just keep it all. The time taken to sort through it is worth more than any tiny extra cost of the hard disk.

I trust live, multiple, physically separate spinning disks more than CDs.

--

John Devereux

Tim does:

(Nice book, BTW).

Simon.

--
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP 
Microsoft: Bringing you 1980s technology to a 21st century world

Let's be realistic here - there is no chance that he is going to pay you. While I have no doubt that you would /like/ to make this profitable, I can't see it happening. The best you can hope for is the admiration of your peers in this group, and the pleasure of publicly demonstrating that someone else is wrong.

So you can keep the details to yourself - and only you and Stroustrup will know them - or you can can tell us all and put us out of our suspense!

CD life is effectively unpredictable. I've seen CDs in kept like that be unreadable in 2 years.

In the late 90s I researched CD longevity, including reading decent tests done by NIST. Summary: - if /and only if/ you know the precise chemical composition and manufacturer of all the different layers, you can predict the archival quality - all standard manufacturers swap all those properties without warning

CDs are unsuitable for storage beyond 2 years, if you want to guarantee readability.

My first "business" machine had a 60MB disk. That's EM, not GEE. How much mail was I gong to store there?

I take it you don't subscribe to many mailing lists? Why would I want to archive all those messages -- when I can just search the list archive directly?

E.g., when I tracked freebsd-hackers, I would spend a fair bit of each day just *deleting* messages that I wasnt interested in.

Of course! I usually buy two or four (always pairs as I think most consumer kit is crap) TB+ drives each holiday season. The price remains constant -- $100. Capacity just keeps going up (last year, my $200 bought me a pair of 3T drives; this year, they'll probably be 4T! :-/ )

I now backup snapshots of the RDBMS that I use in my current project onto such drives. Don't bother with *incremental* backups -- just dump another copy of the entire DB, schemas, etc. each day! Don't even bother to erase old backups cuz the space is cheap.

[And, I've learned -- painfully -- that it is REALLY EASY to do lots of damage to a DB with just a couple of ill-conceived keystrokes. When you don't notice the schema you dropped two weeks ago is "missing" (because of a screwup on your part), you don't want to find yourself without a copy of it to restore! DBA's must lead lives of sheer terror!!]

I don't see it as a "space" issue as much as an "information management" issue. Like 100 icons on your desktop. If you *let* them "get established", then you give up all hope of ever managing them.

So, I only keep what has perceived value (to me) and/or "high downside risk" (that abandoning it might mean I can never reacquire it!) E.g., all my business records (can't "go online" and download a new copy if I lose *my* copy).

Do I save the email from buddy in LV telling me what a bomb movie X was? Or, do I read it and hit delete -- never worrying that "I might need that, someday"?

Do I scan the thank you letter from the non-profit praising my donated efforts? Or, do I read it and toss the paper into the shredder?

I've had excellent luck with CD's -- though not willing to trust DVDs. Anything "precious" I have at least three copies of on at least three different media. I *distrust* magnetic media esp if kept spinning. Too much to lose in a single spindle failure! SImilarly, too easy to accidentally trash something that I was supposed to save!

OTOH, very easy to *locate* contents thereof. And, replicate same. E.g., this holiday season I will move the dual 1.5T "software archive" disks onto a pair of

I suscribe to a dozen or so - and keep all the messages. I find it more efficient to be able to locally search and view the results, than some crappy web interface to a forum.

Anyway, this is all OT, so enough for now I think!

[...]

--

John Devereux

That was what I had *thought*. We looked into optical (MO at the time) storage for archives in a pharma app. We distrusted the

10 year figure quoted by manufacturers.

So, when I first started using CD's (600MB on a thin platter!), I also kept copies on magnetic media. "Back then", no USB. If you wanted "removable", it was external SCSI. Which was fine because I was already running SCSI systems (20+ years ago, I had

60+G spinning -- more than $10K of media!)

So, I would stockpile SCSI drives and treat them like "big CD's". At first, 4G drives. Then 9G. 18G. 36G. 72G (I currently use

146G).

At one point, I inherited a short rack (~24U) with several of these "shelves":

Way too big to keep in its original state (it wanted floor space!). So, I pulled all the drives (modules missing in the first URL) and stacked them on a shelf in the closet. Using each as a "large CD". I.e., locating *the* module I was interested in and connecting it to a live system. (they're just SE SCSI drives in a fancy plastic box)

Eventually, I gave up the hassle of removing the drive *in* each module to replace it with a larger drive (4/9/18/36/72/146) and just started stacking "bare drives" that I would singly install in a Sun 711 case. Less time, overall.

But, I noticed the CD's *weren't* giving me problems! I've had CD *drives* die over the years. But not the media!

So, I now view the magnetic media more from a convenience standpoint. Sure, it also serves as an extra backup. But, it primarily serves as an easy way for me to retrieve software, research papers, datasheets, etc. that I've squirreled away -- without having to sort through stacks of CD's *or* boxes of 72G SCSI drives!

And, I'm not even sure that it's just my treatment of the media that has enabled them to survive for so long!

SWMBO's car is ~10 years old. When purchased, it had an in-dash CD player. She was naturally thrilled that she could now take

*her* music with her!

Ah, but leaving a $15-$20 CD in a hot car (we have ~65 days above

100F each year -- some years as many as 100 such days!) seemed like a sure fire way to throw away $20! :<

So, I ripped each of her CD's so she could keep the originals safe at home and consider the copies "expendable". Using "data" CD's! (apparently, different ink?)

It's 10 years later and I haven't had to "remake" any of these! My biggest concern has now become her continual refusal to eject the CD before leaving the car -- I can just imagine what it will be like to tear the dashboard apart to extract the disk player in the event a disc "melts/deforms" inside it!

(sure, audio CD's are more forgiving when it comes to data loss)

I'll have to "rebuild" her laptop sometime this week. It's *many* years old so the (downloaded) drivers probably aren't available "on-line" anymore. I'll be *forced* to rely on the CD I created when I first built it. Wanna bet I *don't* have any problems?

My point is personal experience contradicts this! Operative word here is "guarantee". To NIST, that means 100% recovery 100% of the time!

And, if *every* CD failed, I still have copies of The Important Stuff on magnetic media. Then, MO. And, in some cases, even tape. I don;t lose any sleep worrying about my archives! :> (though I've lost a lot MAKING them!)

That was what I had *thought*. We looked into optical (MO at the time) storage for archives in a pharma app. We distrusted the

10 year figure quoted by manufacturers.

So, when I first started using CD's (600MB on a thin platter!), I also kept copies on magnetic media. "Back then", no USB. If you wanted "removable", it was external SCSI. Which was fine because I was already running SCSI systems (20+ years ago, I had

60+G spinning -- more than $10K of media!)

So, I would stockpile SCSI drives and treat them like "big CD's". At first, 4G drives. Then 9G. 18G. 36G. 72G (I currently use

146G).

At one point, I inherited a short rack (~24U) with several of these "shelves":

Way too big to keep in its original state (it wanted floor space!). So, I pulled all the drives (modules missing in the first URL) and stacked them on a shelf in the closet. Using each as a "large CD". I.e., locating *the* module I was interested in and connecting it to a live system. (they're just SE SCSI drives in a fancy plastic box)

Eventually, I gave up the hassle of removing the drive *in* each module to replace it with a larger drive (4/9/18/36/72/146) and just started stacking "bare drives" that I would singly install in a Sun 711 case. Less time, overall.

But, I noticed the CD's *weren't* giving me problems! I've had CD *drives* die over the years. But not the media!

So, I now view the magnetic media more from a convenience standpoint. Sure, it also serves as an extra backup. But, it primarily serves as an easy way for me to retrieve software, research papers, datasheets, etc. that I've squirreled away -- without having to sort through stacks of CD's *or* boxes of 72G SCSI drives!

And, I'm not even sure that it's just my treatment of the media that has enabled them to survive for so long!

SWMBO's car is ~10 years old. When purchased, it had an in-dash CD player. She was naturally thrilled that she could now take

*her* music with her!

Ah, but leaving a $15-$20 CD in a hot car (we have ~65 days above

100F each year -- some years as many as 100 such days!) seemed like a sure fire way to throw away $20! :<

So, I ripped each of her CD's so she could keep the originals safe at home and consider the copies "expendable". Using "data" CD's! (apparently, different ink?)

It's 10 years later and I haven't had to "remake" any of these! My biggest concern has now become her continual refusal to eject the CD before leaving the car -- I can just imagine what it will be like to tear the dashboard apart to extract the disk player in the event a disc "melts/deforms" inside it!

(sure, audio CD's are more forgiving when it comes to data loss)

I'll have to "rebuild" her laptop sometime this week. It's *many* years old so the (downloaded) drivers probably aren't available "on-line" anymore. I'll be *forced* to rely on the CD I created when I first built it. Wanna bet I *don't* have any problems?

My point is personal experience contradicts this! Operative word here is "guarantee". To NIST, that means 100% recovery 100% of the time!

And, if *every* CD failed, I still have copies of The Important Stuff on magnetic media. Then, MO. And, in some cases, even tape. I don;t lose any sleep worrying about my archives! :> (though I've lost a lot MAKING them!)

Good anecdote :)

That's the rational course of action.

Ah, no! It's a chance for him to MAKE SOME MONEY! ****If**** he has the courage of his own convictions! If his *reputation* is worth anything to *him* and he's so CERTAIN of his beliefs that he would MALIGN me publicly, you would think he would be *eager* to jump at the chance!

OK, maybe $100K is out of his price range. But, heck, as it was a SURE THING (for me!), why not shoot for the moon? Had he the same sense of surety, he'd have been eager to see me humiliated AND obligated to this financial outcome!

I guess he's just "all talk". The $1000 figure scared him off, too.

Maybe he just "waits tables" so a day's pay is only $100? And, he probably wouldn't be able to risk *that*, either!

"All talk"

Hmmm.. the term "self-aggrandizing" comes to mind. Where have I heard that before?

So, now that we have established that his reputation isn't worth more than 3 figures (*if* that)... [sort of the same ballpark as streetwalkers, eh?]

Not interested in "admiration". And, I've known all along that "someone else is wrong" :>

Actually, *his* "errata" suggestion was the fastpath to The Truth. As BS hasn;t written *lots* of titles, it was simple to find the one:

From the second edition changes, note:

"pg246 in example 1,2,3, and 5 swap X and X& in the return types. in example 4 swap postfix and prefix in the comments."

Note that this is BS's "testimony" -- not anything *I've* manufactured!

Unfortunately, he doesn't credit ANY of the folks who may have reported these "errors" -- and I'm sure a great number of them came from "customers" and not his own eyes!

No big deal. I don't need to see my name in lights! :>

This is consistent with my mention of "backwards" -- SWAP! And, *5* instances on a single page (not a "simple typo" -- though it is easy to see how he could have missed the first mistake and *copied* it blindly).

Ah, but there's no PROOF that *I* was involved in this! I could have FABRICATED this possible explanation/bug after-the-fact just to provide a foundation *under* my claims! "Bunkum".

Call in another third party to testify:

(You may recognize some of the names, there! :> And, the same "phrase" I've claimed to use -- "What am I missing, here?")

Note the date of the posting: 10/21/94 -- and the reference to 2ed of May 94. I.e., 3ed not yet available. (perhaps Stroustrup released the errata long before the 3ed and I "peeked" at it? singling out this

*one* issue to raise AS IF I was the first to notice it? And, bringing it to "everyone's" attenion before it was seen "in print"? :-/ )

BS's errata notes that the change was introduced *after* the 2nd edition. I.e., after 94 but before the third edition IN 1995!

Should I *really* start digging for email to BS on or about 10/21/94 covering this EXACT issue? Is it likely that I would NOT have contacted him?? (think about the tenacity that I am pursuing *this*

20 years later! Do you think I wouldn't have sought clarification AND RECTIFICATION "back then"? :> )

Look, also, at Lucas' failure to see the issue at all! And, his off-handed "dismissal" of my question (read a book). Yet, he was a prolific poster in that group. Self-appointed expert?? :> Eventually, Kanze saw the point I was raising (after I made it so explicit that it couldn't be ignored!)

This isn't comparable to mistakenly (typo) using ++ instead of --. It's more on a par with (mistakenly) treating malloc()'s return value as an int -- AND THEN OPERATING ON IT AS SUCH! Who (knowledgeable) would do such a thing?

You craft a language. You see how prefix and postfix operators behave in the language from which yours derives. You *extend* that language to allow for operator overloading. Then, post examples of such overloads IN A REFERENCE TEXT -- and get them

*wrong*??

I could understand Ritchie *misspelling* "creat()" as "create()". But, not as "fopen()"!

This, IMO, is the problem with C++'s complexity. There are just too many of these niggling little details that can screw you EVEN IF YOU ARE AN EXPERT! And, getting *to* "Expert" is an arduous road!

David Harmon can now skulk away relieved to know his "tip money" is safe. His reputation considerably LESS so!

I'll leave it to *him* to contact BS with these dates and claims and see if BS can produce records of our conversation. That will be his *sole* opportunity to "debunk" the "proof" I've presented, here.

"Hurry up, little boy!"

I find it particularly annoying when people question posters AS IF the poster doesn't understand what he's asking.

"Why do you want to do that?" "Because Im an idiot! Or, maybe because I'm SMARTER than you or see a different way of doing something, unique. Why do you need to know in order to answer the question??"

"If Mary had 6 apples and she gave Tommy 2, how many apples would Mary have?" "Why did Mary give Tommy the apples? Why not oranges??"

This was the reason behind my outburst to *you* (DB) recently. *Assume* I know what I am talking about when I ask a question. *Assume* I've done my homework. *Assume* I have a logical reason for asking. *If* there is some question in your mind, assume the problem is one of YOUR comprehension! This may be related to my language/vocabulary skills. People grumble about the length of my posts -- but I try to *add* verbiage in the hope that it *refines* my explanations/queries instead of leaving obvious ambguity.

Or, maybe you just don't have the same "vision" that I have on the topic at hand.

After 20 years of posting on USENET, I would assume *most* folks have *some* idea as to my capabilities. That there is some "depth" to my queries. Though they might not be keen on my personality, verbosity, etc. There's lots of folks that *I'm* not keen on, too! If I can help them, I can. If they are *too* annoying, I just don't click on their posts! :-/

I get tired of having to explain all my reasoning, background information, on the problem I am addressing and the *approach* that I am pursuing (often too "novel" the folks reading!) only to *then* be met with ... silence. And, complaints that my posts are too long.

Jeez, if I'm bothering you, ADD ME TO YOUR KILLFILE! I don't change my posting address, "From" line, etc. This should be a trivial thing for ANYONE to do!! :-/

(sigh)

Time for my morning tea... then, more downloads (building several laptops today -- such fun! NOT)

--don

That, by the way, is a lie. You never once offered me a dime of your money. It was all your fantasy about how I was supposed to put up money to pay you to look for evidence to support your own claims.

Go away, little boy. You've been proven wrong. I think there's a cleanup required at Table 6...