Forgot your password?
typodupeerror
User Journal

Journal: Why branch prediction doesn't help

Journal by davecb

In the discussion about IBM putatively buying Sun, we were having a side-discussion about prefetches and branch prediction.

I had forgotten why my branch prediction performance experiments had failed ("confirmed the null hypothesis") and had to go back to my notes.

It turns out that mature production software tends to be full of small blocks of error-handling and debug/logging code, which is not often used. A Smarter Colleague[TM] and I set out to test the newly-available branch prediction logic, expecting to see a significant improvement. I manually set the branch prediction bits in a large production application, only to find no detectable improvement.

The test application was Samba, so we changed the driver script to only read a few files from a ram disk, to eliminate disk I/O overheads. Still no detectable advantage from predicting the branches correctly!

Then we tried just a single few functions, under a test framework that did no I/O at all. Still nothing.

Eventually we tracjked it down to the debug/log/else logic: the branches areound it were always taken, but the branch-arounds were long enough that the next instructions were in a different icache line, and the cache-line had to be fetched.

It turned out that we had reproduced in code what our HPC colleagues see in data: the cache doesn't help if you're constantly leaping to a different cache line!

--dave

User Journal

Journal: Capacity planning in six paragraphs

Journal by davecb

An acquaintance asked about what to measure, and what tools to use, expecting to hear about vmstar, sar or the like.

However, the really interesting measurements are of the application's performance: response time and transactions per second.

Imagine you have a web site which responds in 1/10 second on average, is known to be running on a single cpu (queuing center, to be precise) and is averaging 6 transactions per second (TPS)

From that you know that the maximum performance will be 10 TPS, because ten 1/10ths fit into one second. You also know you're at 60% of the maximum, a nice safe number.

Now correlate this with your average CPU usage, network bandwidth and IO bandwidth, and you have a little estimator for what resources are needed to maintain good performance.

You also know that things will start getting bad at >8 TPS, so if you expect more business in future, you need to add more queuing centers (CPUs) with the appropriate amounts of network and disk I/O bandwidth.

You can also now use both the resource usage figures and tools that all the other folks have suggested, and watch out for growth in each of them. If the trend in their use looks like it will soon get above the number that corresponds to 8 TPS, above, then and only then do you need to start buying resources.

--dave c-b

User Journal

Journal: J.K.Rowling wins $6750, and pound of flesh 17

Journal by NewYorkCountryLawyer
J.K. Rowling didn't make enough money on Harry Potter, so she had to make sure that the 'Harry Potter Lexicon' was shut down. After a trial in Manhattan in Warner Bros. v. RDR Books, she won, getting the judge to agree with her (and her friends at Warner Bros. Entertainment) that the 'Lexicon' did not qualify for fair use protection. In a 68-page decision (PDF) the judge concluded that the Lexicon did a little too much 'verbatim copying', competed with Ms. Rowling's planned encyclopedia, and might compete with her exploitation of songs and poems from the Harry Potter books, although she never made any such claim in presenting her evidence. The judge awarded her $6750, and granted her an injunction that would prevent the 'Lexicon' from seeing the light of day.
User Journal

Journal: U. Mich. student calls for prosecution of Safenet

Journal by NewYorkCountryLawyer
An anonymous University of Michigan student targeted by the RIAA as a 'John Doe', is asking for the RIAA's investigator, Safenet (formerly MediaSentry), to be prosecuted criminally for a pattern of felonies in Michigan. Known to Michigan's Department of Labor and Economic Growth -- the agency regulating private investigators in that state -- only as 'Case Number 162983070', the student has pointed out that the law has been clear in Michigan for years that computer forensics activities of the type practiced by Safenet require an investigator's license. This follows the submissions by other 'John Does' establishing that Safenet's changing and inconsistent excuses fail to justify its conduct, and that Michigan's legislature and governor have backed the agency's position that an investigator's license was required.
User Journal

Journal: ABA Judges Get an Earful about RIAA Litigations 5

Journal by NewYorkCountryLawyer
Well, I was afforded the opportunity to write for a slightly different audience -- the judges who belong to the Judicial Division of the American Bar Association. I was invited by the The Judges' Journal, their quarterly publication, to do a piece on the RIAA litigations for the ABA's Summer, 2008, 'Equal Access to Justice' issue. What I came up with was 'Large Recording Companies vs. The Defenseless : Some Common Sense Solutions to the Challenges of the RIAA Litigations', in which I describe the unfairness of these cases and make 15 suggestions as to how the courts could make it a more level playing field. I'm hoping the judges mod my article '+5 Insightful', but I'd settle for '+3 Informative'. For the actual article go here (PDF). (If anyone out there can send me a decent HTML version of it, I'll run that one up the flagpole as well.)
User Journal

Journal: eBay beats Tiffany's in trademark case 2

Journal by NewYorkCountryLawyer
Tiffany's has lost its bid to hold eBay liable for trademark infringement of Tiffany's brands taking place on eBay. After a lengthy bench trial (i.e. a trial where the judge, rather than the jury, decides the factual questions), Judge Richard J. Sullivan has issued a 66-page decision (PDF) carefully analyzing the facts and legal principles, ultimately concluding that 'it is the trademark owner's burden to police its mark, and companies like eBay cannot be held liable for trademark infringement based solely on their generalized knowledge that trademark infringement might be occurring on their websites'.
Earth

Journal: Save time, money and 100 million trees

Journal by Jahf

Just a note about an application (Windows only though) I've started using: GreenPrint http://www.printgreener.com/. This marks possibly the first time I've ever sent a software recommendation to my friends and family. If its good enough for them, it might be good enough for some of you!

GreenPrint adds a step between hitting the "Print" button and actually sending the task to the printer. This step allows you to weed out the wasted space before printing and/or saving. It is very easy to use.

How often do people waste paper when printing prototypes, brochures, etc? You can easily prevent 20% of your total printing. Save closer to 50% when printing "1-page" web pages that so often have that one nearly blank page at the end. Save over 50% when printing things with lots of "hidden" text. Ever seen a spreadsheet print 4 blank pages for 1 "real" page (or more)?

There is a free version (GreenPrint World) that is supported by unobtrusive ads, but is not free for commercial use (ie, in the office). C|Net's download.com reports it as verified free of spyware/malware. Home Premium is corporate licensable. Volume licensing is available for the Enterprise version; I don't have a clue what their terms are. Home Premium and Enterprise have additional features. All run on Windows 2000, XP and Vista.

NOTE: I have no interest in the program other than it being a smart way to save trees, time, and money. This isn't spam it is a real review. If you want to see a demo before installing, go here: http://printgreener.com/tutorial.html

  • Do you ever:
  • need to print only SPECIFIC pages, not everything?
  • want to define rules for things like "don't print blank pages!"?
  • want to remove images from your document?
  • prefer to save a PDF file instead of printing?
  • GreenPrint will allow you to:
  • remove entire pages (easily, graphically)
  • print only text (whole document or selected pages)
  • print only images (whole document or selected pages)

NOTE: Don't just use the right-click menu for these, use the "Tools" menu for more control

  • Once done you can:
  • Print to any of your configured printers
  • Save to a PDF (replacing tools like CutePDF)
  • Email immediately as a PDF (I didn't get this working but you can email saved PDF files)
  • GreenPrint will also:
  • Automatically remove blank and "header/footer only" pages
  • Automatically remove pages via configurable criteria (ex: less than X lines, etc)
  • Alert you before removing pages automatically
  • Track the amount of money saved (and define costs on a per-printer basis)
  • Track how many pages you have saved
  • Handle printer configuration (like you have now)
  • Collate pages
  • PS. As with any 1.00.00 software there are bugs. I reported two today to their support address. I got replies back in under an hour.

User Journal

Journal: Dow Jones MarketWatch likens RIAA to the Mafia 11

Journal by NewYorkCountryLawyer
According to commentator Therese Polletti at Dow Jones MarketWatch, "the RIAA's tactics are nearly as bad as the actions of mobsters, real or fictional. The analogy comes up easily and frequently in any discussion of the RIAA's maneuvers." Among other things she cites the extortionate nature of their 'settlement negotiations' pointed out by Prof. Bob Talbot of the University of San Francisco School of Law IP Law Clinic, whose student attorneys are helping private practitioners fight the RIAA, the illegality of the RIAA's use of unlicensed investigators, the flawed evidence it uses, and the fact that the RIAA thinks nothing of jeopardizing a student's college education in order to make their point, as support for the MAFIAA/Mafia analogy.
User Journal

Journal: Class action complaint against RIAA available online 4

Journal by NewYorkCountryLawyer
Recommended reading for all interested in the RIAA's litigation war against p2p file sharing is the amended class action complaint just filed in Oregon in Andersen v. Atlantic. This landmark 109-page document (pdf) tells both the general story of the RIAA's campaign against ordinary folks, and the specific story of its harassment of Tanya Andersen, and even of her young daughter. The complaint includes federal and state RICO claims, as well as other legal theories, and alleges that "The world's four major recording studios had devised an illegal enterprise intent on maintaining their virtually complete monopoly over the distribution of recorded music." The point has been made by one commentator that the RIAA won't be able to weasel its out of this one by simply withdrawing it; this one, they will have to answer for. If the relief requested in the complaint is granted, the RIAA's entire campaign will be shut down for good.
User Journal

Journal: EFF travels to Arizona to argue Howell case

Journal by NewYorkCountryLawyer
Although based in San Francisco, and only an amicus curiae in the Phoenix, Arizona, case of Atlantic v. Howell, the Electronic Frontier Foundation is sending its senior intellectual property lawyer Fred Von Lohmann to Phoenix to argue the Howell case, on behalf of the defendant, who is not represented by counsel. Due to the RIAA's attempt to take advantage of Mr. Howell's being undefended to try to convince the judge that merely 'making files available for distribution' -- i.e., just having them on one's computer in a manner that is accessible to sharing -- and that copying files from one's cd onto one's computer in mp3 format is itself "unlawful", EFF filed an amicus brief in January. Now it's taking the unusual step of actually sending someone to the courthouse to orally argue the motion.
User Journal

Journal: Should RIAA's investigator have to disclose backup? 12

Journal by NewYorkCountryLawyer
A technology battle is raging in UMG v. Lindor in Brooklyn over whether the RIAA's investigator, SafeNet (formerly known as MediaSentry), which has produced certain *txt printouts, now needs to disclose its digital files, validation methodology, testing procedures, failure rates, software manuals, protocols, packet logs, source code, and other materials, so that the validity of its methods can be evaluated by the other side. SafeNet and the RIAA say no, the information is "proprietary and confidential". Ms. Lindor says yes, if you're going to testify in federal court the other side has a right to test your evidence. A list of what is being sought is here (pdf). MediaSentry has produced 'none of the above'. "Put up or shut up" says one commentator to MediaSentry. What do you say?
User Journal

Journal: Connecticut Judge rejects RIAA 'making available' theory 3

Journal by NewYorkCountryLawyer
A federal judge in Connecticut has rejected the RIAA's "making available" theory, which is the basis of all of the RIAA's peer to peer file sharing cases. In Atlantic v. Brennan, in a 9-page opinion (pdf), Judge Janet Bond Arterton held that the RIAA needs to prove "actual distribution of copies", and cannot rely -- as it was permitted to do in Capitol v. Thomas -- upon the mere fact that there are song files on the defendant's computer and that they were "available". This is the same issue that has been the subject of extensive briefing in two contested cases in New York, Elektra v. Barker and Warner v. Cassin. Judge Arterton also held that the defendant had other possible defenses, such as the unconstitutionality of the RIAA's damages theory and possible copyright misuse flowing from the record companies' anticompetitive behavior.
User Journal

Journal: U. of Maine legal clinic fights RIAA; first in country

Journal by NewYorkCountryLawyer
"A student law clinic is about to cause a revolution" says p2pnet. For the first time in the history of the RIAA's ex parte litigation campaign against college students, a university law school's legal aid clinic has taken up the fight against the RIAA in defense of the university's college students. Student attorneys at the University of Maine School of Law's Cumberland Legal Aid Clinic, under the supervision of law school prof Deirdre M. Smith, have moved to dismiss the RIAA's complaint in a Portland, Maine, case, Arista v. Does 1-27, on behalf of 2 University of Maine undergrads. Their recently filed reply brief (pdf) points to the US Supreme Court decision in Bell Atlantic v. Twombly, and the subsequent California decision following Twombly, Interscope v. Rodriguez, which dismissed the RIAA's "making available" complaint as mere "conclusory", "boilerplate" "speculation". The 2 students represented by Cumberland join the 8 students represented by a prominent Portland law firm, bringing to 10 the number of University of Maine students fighting back in this case.
User Journal

Journal: RIAA objects to Oregon AG's request for information 2

Journal by NewYorkCountryLawyer
The RIAA is apparently having an allergic reaction to the request by the State Attorney General of Oregon for information about the RIAA's investigative tactics, in Arista v. Does 1-17, the Portland, Oregon, case targeting students at the University of Oregon. See The Oregonian, December 1, 2007 ("UO suspects music industry of spying") and p2pnet, November 29, 2007 ("RIAA may be spying on students: Oregon AG"). Not only are the record companies opposing the request (pdf), they're asking the Judge not to even read it. (pdf) Commentary: "RIAA scorns Oregon University request".

The bogosity meter just pegged.

Working...