Here is the documentation for a little module I threw together after one of our services did a denial-of-service attack against another of our services. The math for this simple trick works out very nearly.

I plan to upload this to CPAN very soon. Please let me know what you think.

NAME

SYNOPSIS

    use Proc::Governor();

    my $gov = Proc::Governor->new();

    while( ... ) {
        $gov->breathe();
        ... # Use resources
    }

    while( ... ) {
        my $res = $gov->work( sub {
            ... # Use Service
        } );
        ...
    }
[download]

DESCRIPTION

If you want to do a batch of processing as fast as possible, then you should probably also worry about overwhelming some resource and causing problems for other tasks that must share that resource. Fortunately, there is a simple trick that allows one to perform a batch of processing as fast as possible while automatically backing off resource consumption when most any involved resource starts to become a bottleneck (or even before it has become much of a bottleneck).

The simple trick is to pause between steps for a duration equal to how long the prior step took to complete. The one minor down-side to this is that a single strand of execution can only go about 1/2 maximum speed. But if you have 2 or more strands (processes or threads), then throughput is not limited by this simple "universal governor" trick.

It is also easy to slightly modify this trick so that, no matter how many strands you have working, they together (without any coordination or communication between the strands) will never consume more than, say, 60% of any resource (on average).

A typical pattern for batch processing is a client sending a series of requests to a server over a network. But the universal governor trick also works in lots of other situations such as with 1 or more strands where each is doing a series of calculations and you don't want the collection of strands to use more than X% of the system's CPU.

Note that the universal governor does not work well for resources that remain consumed while a process is sleep()ing, such as your process using too much memory.

Proc::Governor provides lots of simple ways to incorporate this trick into your code so that you don't have to worry about your code becoming a "denial-of-service attack", which also frees you to split your processing among many strands of execution in order to get it done as fast as possible.

METHODS

new()

    my $gov = Proc::Governor->new( {
        working     => 0,
        minSeconds  => 0.01,
        maxPercent  => 100,
        unsafe      => 0,
    } );
[download]

new() constructs a new Proc::Governor object for tracking how much time has recently been spent potentially consuming resources and how much time has recently been spent not consuming resources.

new() takes a single, optional argument of a reference to a hash of options. The following option names are currently supported:

working

If given a true value, then the time spent immediately after the call to new() is counted as "working" (consuming resources). By default, the time spent immediately after the call to new() is counted as "not working" (not consuming).

minSeconds

minSeconds specifies the shortest duration for which a pause should be done. If a pause is requested but the calculated pause duration is shorter than the number of seconds specified for minSeconds, then no pause happens (and that calculated duration is effectively added to the next pause duration).

The default for minSeconds is 0.01.

maxPercent

maxPercent indicates how much of any particular resource the collection of strands should be allowed to consume. The default is 100 (for 100%, or all of any resource, but avoid building up a backlog by trying to over-consuming any resource).

Note that percentages are not simply additive. Having 3 groups of clients where each is set to not consume more than 75% of the same service's resources is the same as having just 1 group. The 3 groups together will not consume more than 75% of the service's resources in total.

Say you have a group of clients, H, all set to not consume more than 50% of some service's resources and you have another group of clients, Q, all set to not consume more than 25% of that same service's resources. Both H and Q together will not add up to consuming more than 50% of the service's resources.

If Q is managing to consume 20% of the service's resources when H starts running, then H won't be able to consume more than 30% of the service's resources without (slightly) impacting performance to the point that Q starts consuming less than 20%.

     H   Q   Total
    50%  0%   50%
    40% 10%   50%
    30% 20%   50%
    25% 25%   50%
[download]

unsafe

You can actually specify a maxPercent value larger than 100, perhaps because you have measured overhead that isn't easily accounted for by the client. But doing so risks overloading a resource (your measured overhead could end up being a much smaller percentage of the request time when the service is near capacity).

So specifying a maxPercent of more than 100 is fatal unless you also specify a true value for unsafe.

beginWork()

    $gov->beginWork( $breathe );
[download]

Calling beginWork() means that the time spent immediately after the call is counted as "working" (consuming resources). Such time adds to how long the next pause will be.

If $breathe is a true value, then beginWork() may put the strand to sleep for an appropriate duration.

endWork()

    $gov->endWork( $breathe );
[download]

Calling endWork() means that the time spent immediately after the call is counted as "not working" (not consuming resources). Such time subtracts from how long the next pause will be.

If $breathe is a true value, then endWork() may put the strand to sleep for an appropriate duration.

work()

    $gov->work( sub {
        ... # Consume resources
    }, $which );
[download]

work() is a convenient shortcut that is roughly equivalent to:

    $gov->beginWork( $before );
    ... # Consume resources
    $gov->endWork( $after );
[download]

The value of $which can be:

    0   No pause will happen.
    1   A pause may happen before the sub reference is called.
    2   A pause may happen after the sub reference is called.
    3   A pause may happen before and/or after the sub is called.
[download]

If $which is not given or is undefined, then a value of 1 is used.

You can actually get a return value through work():

    my @a = $gov->work( sub { ...; get_list() }, $which );
    my $s = $gov->work( sub { ...; get_item() }, $which );
[download]

Note that scalar or list (or void) context is preserved.

Currently, if your code throws an exception, then endWork() does not get called. This is the same as would happen with the "equivalent" code shown above.

breathe()

    $gov->breathe( $begin );
[download]

Calling breathe() requests that the current process/thread pause for an appropriate duration.

Each of the following:

    $gov->breathe();
    # or
    $gov->breathe( 1 );
[download]

is actually equivalent to:

    $gov->beginWork( 1 );
[download]

While

    $gov->breathe( 0 );
[download]

will just pause but will not change whether $gov is counting time as "working" or as "not working".

pulse()

    $gov->pulse( $count, $begin );
[download]

pulse() is very much like breathe() except that it is optimized for being called many times before enough "working" time has accumulated to justify doing a pause. The meaning of $begin is the same as with breathe().

So, if you are making requests of a very fast service or are doing work in small chunks, then you can call pulse() directly in your loop and just pass it a value specifying approximiately how many calls to pulse() should be made before one of those calls does the work of calculating how long of a pause is called for.

For example, a request to our Redis service typically takes a bit under 1ms. So code to perform a large number of such requests back-to-back might be written like:

    my $gov = Proc::Governor->new( {
        maxPercent  => 70,
        working     => 1,
    } );
    my $redis = Redis->new(server=>...);
    while( ... ) {
        $gov->pulse( 20 );
        $redis->...;
    }
[download]

That is like calling breathe() every 20th time through the loop and is only the slightest bit less efficient (in run time) than if you had made the extra effort to write:

    ...
    my $count = 0;
    while( ... ) {
        if( 20 < ++$count ) {
            $gov->breathe();
            $count = 0;
        }
        ...
[download]

CROSS-OBJECT INTERACTIONS

A single process (or thread) can simultaneously use more than one Proc::Governor object. For example, each process (of a group) that makes a series of requests to a service and does significant local processing of the data from each request might want to both prevent overwhelming the service and prevent overwhelming local resources (such as CPU).

So you could have two Proc::Governor objects. One throttles use of local resources ($g_cpu below). The other throttles use of service resources ($g_db below).

    my $g_cpu = Proc::Governor->new( { maxPercent => 80 } );
    my $g_db =  Proc::Governor->new( { maxPercent => 30 } );

    $g_db->beginWork();
    my $db = DBI->connect( ... );       # DB work
    my $rows = $db->selectall_arrayref( ... );
    $g_db->endWork();

    for my $row ( @$rows ) {
        my $upd = $g_cpu->work( sub {
            process_row( $row );        # Local work
        } );
        $g_db->work( sub {
            $db->update_row( $upd );    # DB work
        } );
    }
[download]

The above code assumes that the local resources required for making requests of the database service are relatively low. And realizes that doing local computations do not use database resources.

If you set maxPercent to 100 for both Governors and each process spent about the same amount of time waiting for a response from the database as it spent performing local computations, then there might be no need for any pauses.

Note that only time spent doing "DB work" adds to how long of a pause might be performed by the $g_db Governor. And only time spent doing "Local work" adds to how long of a pause might be performed by the $g_cpu Governor.

Any pauses executed by either Governor get subtracted from the duration of any pauses of any Governor objects. So the $g_db Governor executing a pause also counts as a pause for the $g_cpu Governor (and thus makes the next pause that it performs either shorter or later or just not needed).

Time spent inside of Proc::Governor methods may also be subtracted from future pause durations. But the code pays more attention to keeping such overhead small than to providing highly accurate accounting of the overhead and trying to subtract such from every Governor object.

WHEN TO PAUSE

Say you have a service that is a layer in front of some other service. You want to ensure that your service can't become a denial-of-service attack against the other service. But you want to prevent a Governor pause from impacting clients of your service when possible.

You could implement such as follows:

    sub handle_request {
        my( $req ) = @_;
        our $Gov ||= Proc::Governor->new();
        my $res = $Gov->work( sub {
            forward_request( $req );
        }, 0 );                 # Don't pause here.
        give_response( $res );
        $Gov->breathe( 0 );     # Pause here; still idle.
    }
[download]

(Well, so long as your service architecture supports returning a complete response before the request handler subroutine has returned.)

If the other service is not near capacity, then the added pauses have no impact (other than perhaps preventing the number of active strands for your service from dropping lower). Be sure your service has an appropriate cap on how many strands it is allowed to keep active (as always).

TO-DO

A future version should have support for asynchronous processing. The shape of that interface is already sketched out, but the initial release was not delayed by the work to implement such.

So my main question, also to myself, is "How much speed are you willing to sacrifice for a new feature?".

Really. Lets assume you have a neat module that deals with your data, and it deals with it pretty well and reliable, but extending it with new features - some of them asked for by others - is getting harder and harder.

We now have git, and making a branch is easy, so you can implements the most requested new feature, or the one that most appeals to you and when you are done and all old tests and new tests have passed, you notice a speed drop.

What considerations do you make to decide whether to release the module with the new neat new feature and mention the slowdown (specified) or do you revert the change and note in the docs that the new feature would cause to big a slowdown.

Dear Monks, I am pretty new to programming. In the past i spent my time trying to be a pro overnight but lately realized my initial approach to learning was wrong. I spent most of my time doing irrelevant things on computer because i found it had focusing on one thing which is practicing programming. Lately i spend the whole day laying on my bed trying to slow down my life and rediscover myself. In doing this i only use the computer for a short period of time which is usually in the evenings and i now visit less websites. when i notice i am no longer able to focus i logout and relax. Why i do this is because i have come to realize that if i want to be a skilled programmer it is something i have to commit to and let other things go. In doing this i find myself learning more though my progress is quite slow but im pleased with the pace at which i am going. These are still early days of my approach. I hope i am not doing anything wrong in my approach. I will love to know how others relax and spend there day at home or at work and what advice do you all have for me. Thanks in advance.

Starting with PAUSE, I have uploaded several new modules to PAUSE. Each time, I get a message that says:

This distribution name can only be used by users with permission for the package <My::Package>, which you do not have.

The packages are in new namespaces. As I understand it, simply uploading a new module should allocate that namespace to me on a "first-come" basis. But it isn't.

This doesn't seem to matter to search.cpan.org, when it's working: it still indexes the module so that it can be found and downloaded via the cpan utility.

However that doesn't seem to apply to metacpan. It uses 02packages.details.txt which isn't being updated, presunably because of the PAUSE issue. Thus my modules are not appearing on metacpan in their search. Metacpan's help says:

MetaCPAN uses the PAUSE generated 02packages.details.txt file. If it's not in there, then the module author will need to fix this,

Does anyone know if it's fixable? I have mailed modules@perl.org a couple of times but no response.

This started life as a reply to Re^2: Which 'Perl6'? (And where?), but it seems too important to bury it down there in a long dead thread as a reply to an author I promised to resist, and whom probably will not respond. So I'm putting it here to see what of any interest it arouses.

1. Is concurrency appropriate? There are two basic motivations ... and 2) to speed things up. In the latter case, if the problem being tackled is really IO bound, turning to concurrency probably won't help.

   That is way too simplistic a view. If the problem is IO bound to a single, local, harddisk, and is uncacheable, then concurrency may not help.

   But change any of the four defining elements of that criteria; and it might -- even: probably will -- be helped by well written asynchronicity. Eg.

   1. If the IO data is, or can be, spread across multiple local physical drives; concurrency can speed overall throughput by overlapping requests to different units.
   2. If the disks are remote -- as in SAN, NAS, cloud etc. -- then again, overlapping requests can increase throughput by utilising buffering and waiting time for processing.
   3. If the drives aren't harddisks, but SSDs; or SSD buffered HDs; or PCI connected virtual drives; then overlapping several fast read requests with each slower write request can more fully utilise the available bandwidth and improve throughput.
   4. If the IO involved displays temporal locality of reference -- that is, if the nature of the processing is such that a subset of the data has multiple references over a short period of time, even if that subset changes over the longer term -- then suspending the IO for new references until re-references to existing cached data play out comes about naturally if fine-grained concurrency is used.

   And if some or all of the IO in your IO bound processing is to the network, or network attached devices; or the intranet; or the internet; or the cloud; -- eg. webserving; webcrawling; webscraping; collaborative datasets; email; SMS; customer facing; ....... -- then both:

   • Preventing IO from freezing your processing;
   • And allowing threads of execution who's IO has completed to continue as soon as a core is available -- ie. not also have to wait for any particular core to become available;

   Is mandatory for effective utilisation of modern hardware and networks; even for IO-bound processing.

   Only kernel(OS) threading provides the required combination of facilities. Cooperative multitasking (aka. 'green threads'; aka. Win95 tech) simply does not scale beyond the single core/single thread hardware of the last century.

2. The Problem with Threads.

   The problem with "The Problem with Threads", is that it is just so much academic hot air divorced from the realities of the real world.

   Only mathematicians and computer scientists demand total determinacy; and throw their arms up in refusal to work if they don't get it.

   The rest of the world -- you, me, mothers and toddlers, doctors, lawyers, spacemen, dustmen, pilots, builders, shippers, movers & shakers, factory workers, engineers, tinkers, tailors, soldiers, sailors, rich & poor men, beggars and thieves; all have to live in the real -- asynchronous -- world, where shit happens.

   Deliveries are late; machines break down; people are sick; power-outs and system-downs occur; the inconvenient realities of life have to be accepted, lived with and dealt with.

   The problem is not that threading is hard; the problem is that people keep on saying that "threading is hard"; and then stopping there.

   Man is very adept at dealing with hard and complex tasks

   Imagine all places you'd never have been; all the things you'd never have done; if the once wide-spread belief that we would suffocate if we attempted to travel at over 30mph.

   Too trivial an example for you? Ok. Think about heart transplantation. Think about the problems of disconnecting and reconnecting the (fragile, living) large bore pipes supplying and removing the pumped liquid; the wires carrying electrical control signals; the small bore pipes carrying the lubricants needed to keep the pump alive and removing the waste. Now think about the complexities of doing a pump change whilst keeping the engine running; the passengers comfortable and the 'life force' intact. And all the while contending with all the other problems of compatibility; rejection; infection; compounded diagnosis.

   Circa. 5000 coronary transplants occurred last year. Mankind is good at doing difficult things.

   Asynchronicity and non-determinism are 'solved problems' in almost every other walk of life

   From multiple checkouts in supermarkets; to holding patterns in the skies above airport hubs; to off & on ramps on motorways; to holding tanks in petro-chemical plants; to waiting areas in airports and doctors and dentists surgeries; to carousels in baggage claims and production lines; distribution warehouses in supply chains; roundabouts and filter-in-turn; {Add the first 10 things that spring to your mind here! }.

   One day in the near future a non-indoctrinated mathematician is going to invent a symbol for an asynchronous queue.

   She'll give it a nice, technical sounding name like "Temporally Lax Composer", which will quickly become lost behind the cute acronym and new era of deterministic, asynchronous composability will ensue.

   And the academic world will rejoice, proclaim her a genius of our time, and no doubt award her a Nobel prize. (That'd be nice!)

   And suddenly the mathematicians will realise that a process or system of processes can be deterministic, without the requirement for every stage of the process (equation) to occur in temporal lockstep.

   'Safety' is the laudable imperative of the modern era.

   As in code-safety and thread-safety, but also every other kind of predictable, potentially preventable danger.

   Like piety, chastity & sobriety from bygone eras, it is hard to argue against; but the world is full (and getting fuller) of sexually promiscuous atheists who enjoy a drink; that hold down jobs, raise kids and perform charitable works. The world didn't fall apart with the wane of the religious, moral and sobriety campaigns of the past.

   In an ideal world, all corners would be rounded; flat surfaces 'soft-touch'; voltages would be low; gases non-toxic; hot water wouldn't scald; radiant elements wouldn't sear; microwaves would be confined to lead-lined bunkers; there'd be no naked flames; and every home would be fire-proof, flood-proof, hurricane-proof, tornado-proof, earthquake-proof, tsunami-proof and pestilence-proof.

   Meanwhile in the real-world, walk around your own house and see all the dangers that lurk for the unsupervised, uneducated, unwary, careless or stupid and ask yourself why do they persist? Practicality and economics.

   Theoreticians love theoretical problems; and eschew practical solutions.

   When considering concurrency, mathematicians love to invent ever so slightly more (theoretically) efficient solutions to the 'classical' problems.

   Eg. The Dining Philosophers. In a nutshell: how can 6 fil..Phillo.. guys eat their dinners using 5 forks without one or more of them starving. They'll combine locks and syncs, barriers and signals, mutexs and spinlocks and semaphores trying to claw back some tiny percentage of a quasilinear factor.

   Why? Buy another bloody fork; or use a spoon; or eat with your damn fingers.

   The problem is said to represent the situation where you have 6 computers that need to concurrently use the scarce resource of 5 tape machines. But that's dumb!

   Its not a resource problem but a capital expenditure problem. Buy another damn tape machine and save yourself 10 times its cost by avoiding having to code and maintain a complex solution. Better still, buy two extra tape machines; cos as sure as eggs is eggs, it'll be the year-end accounting run; or the Black Friday consumer spending peak when one of those tape machines defy the 3 sigma MTBF and break.

   Threading can be complex, but there are solutions to all of the problems all around us in the every day, unreliable, non-deterministic operations of every day modern life.

   And the simplest solution to many of them is to avoid creating problems in the first place. Don't synchronise (unless you absolutely have to). Don't lock (unless it is absolutely unavoidable). Don't share (unless avoiding doing so creates greater problems).

   But equally, don't throw the baby out with the bath water. Flames are dangerous; but oh so very useful.

3. Futures et al are the future. There are much simpler, safer, higher level ways to do concurrency. I haven't tried Paul Evans' Futures, but they look the part.

   And therein lies the very crux of the problem. Most of those decrying threads; and those offering alternative to them; either haven't tried them -- because they read they were hard -- or did try them on the wrong problems, and/or using the wrong techniques; and without taking the time to become familiar with and understand their requirements and limitations.

   Futures neither remove the complexity nor solve the problems; they just bury them under the covers forcing everyone to rely upon the efficacy of their implementation and the competence of the implementors.

   And the people making the decisions are taking advice from those thread-shy novices with silver bullets and employing those with proven track records of being completely useless at implementing threaded solutions.

   The blind taking advice from the dumb and employing the incompetent.

4. Perl 5 "threads" are very heavy. This sometimes introduces additional complexity.

   The "heaviness" of P5 threading is a misnomer. The threads aren't heavy; the implementation of shared memory is heavy. And that could easily be fixed. If there was any interest. If there wasn't an institutionalised prejudicial barrier preventing anyone even suggesting change to improve the threading support; much less supporting those with the knowledge and ideas to take them forward.

   They've basically stagnated for the past 8 or more years because p5p won't allow change.

I like to go to “networking events” now and again (if the price is cheap), but sometimes it is depressing to watch the hawkers.

The thing that they are selling are little TLA’s (three-letter acronyms) to attach after your name.   And, apparently, they’re not cheap.   But one comment that was announced to the crowd, by someone who I’m pretty-much sure was a “shill,” went something like this:

“The very best thing I did for my career was to become a “SCRUM Master Level II”!!   I had to dig to find the money, but you know, you have to ‘invest in yourself to get ahead.’   You don’t even have to know how to program.”

(Blink ...)

And then there was this other guy – let’s be kind and say he didn’t exactly look like Chuck Norris – who nevertheless styled himself a “third-degree black belt master” about some sort of management-theory or another . . .

So, is “the blind leading the blind” an accepted business-practice these days?   Even though the person who runs the shop where I get my car serviced might not be holding an grease-gun when he introduces himself to me, I do want to know that he has held one in the fairly recent past.   And, while I know that jobs like “car servicing” require a tremendous amount of intellectual study these days, at the end of the evening it’s all still all about:   a machine.   You can’t abstract-away the experience of having actually done this and replace it with a little piece of very-expensive paper.   You can’t learn to swim by reading a book about it.

Now, I, likewise, do not wish to here “abstract away” the value of intellectual study, or employee training, or personal self-education.   That’s not my point.   But I feel like I was watching swimsuits being sold to drowning people.   A little friendly-conversation around that room showed me that the length of the courses was usually “over the weekend,” and the prices of those weekends were high – $1000 to $3300 (USD).   Ouch.   If I had thought that I could just buy a bottle of “SCRUM sauce” and pour it over my career (with or without saying, “Shazam!”), I might at one time have been fairly-easily convinced to do so.   Maybe I’m just too old and gray to believe it now.   I do not see a credible value-proposition here, despite the intense sales pressure.

So ... can I now ask the Monks for what is your perspective here?   Whether you, yourself, bought a Golden Ticket, or (I think more-likely) hired one, or didn’t, what were your experiences and insights into this matter?

I don't know what the full back story is behind this, but I've noticed that CPAN is going down a lot lately.

As the first port of call for anyone searching Perl documentation, I think this reflects really badly on the community - I mean, if I wasn't sold on Perl, started looking into it and hit errors on the language's main (edit: "main module documentation") web site, I'd be thinking something along the lines of, "If they can't keep up a web site, how can we rely on Perl as a technology?".

Is there no way that maybe the error pages could be updated if the server situation is not easily fixable? Surely it would be better to display more than just "504 Gateway Time-out" ??? Maybe something more descriptive, in a formatted HTML page that links to other Perl resources (Perlmonks, perl.com. perldocs.net etc).

It a little embarrassing when I'm excitedly discussing a cool aspect of Perl and then try to send that person the CPAN page, only to hit an error message. What can be done to fix this?

(edited title for clarity)

The primary use I can think is that for someone who likes to give Perl talks and wants to give a talk (about anything), they may consult this list to find something that strikes their fancy.

Perl Talk Topics Bank on Github

Another user could be someone who would give a talk if they knew that something about which they felt comfortable talking was of interest to other people, they may step up.

The immediate problem being solved here is a local one - at Houston.pm we're constantly asking the mailing list for topics or speakers to volunteer. Even though it is a local topics bank (act locally, right?), I'd like to get ideas from anyone willing to contribute.

Eventually, I'd like to make this a list that is appropriate for any event or meeting where people talk about Perl and the interesting things we're doing with it.

If you're uncomfortable with forking, editing, and issuing a pull request; please just create an issue with the topics you'd like to add.

Hope to see you contribute to the topics list.

Thank you!

Not everything that can be counted counts and not everything that counts can be counted

-- William Bruce Cameron

What's measured improves

-- Peter Drucker

Three recent events got me thinking about software metrics again:

• Management use individual KPIs to reward high performers in Sales and other departments. They are contemplating doing the same for Software Developers.
• Performance Appraisals often seem subjective. Would metrics make them more objective? Or do more harm than good?
• Larry Maccherone was in town recently promoting his company's approach to Agile metrics.

I'm interested to learn:

• How does your company reward Software Developers? Are the rewards team-based, individual-based, department-based, whole-company based? How well does it work?
• Do you have Performance Appraisals? Do they use metrics? Do your Software Developers/Teams have KPIs?
• Do you use metrics to improve your Software Development Process?

I've done a bit of basic research on these topics, which I present below.

Software Metric Gaming

Key performance indicators can also lead to perverse incentives and unintended consequences as a result of employees working to the specific measurements at the expense of the actual quality or value of their work. For example, measuring the productivity of a software development team in terms of source lines of code encourages copy and paste code and over-engineered design, leading to bloated code bases that are particularly difficult to maintain, understand and modify.

-- Performance Indicator (wikipedia)

"Thank you for calling Amazon.com, may I help you?" Then -- Click! You're cut off. That's annoying. You just waited 10 minutes to get through to a human and you mysteriously got disconnected right away. Or is it mysterious? According to Mike Daisey, Amazon rated their customer service representatives based on the number of calls taken per hour. The best way to get your performance rating up was to hang up on customers, thus increasing the number of calls you can take every hour.

Software organizations tend to reward programmers who (a) write lots of code and (b) fix lots of bugs. The best way to get ahead in an organization like this is to check in lots of buggy code and fix it all, rather than taking the extra time to get it right in the first place. When you try to fix this problem by penalizing programmers for creating bugs, you create a perverse incentive for them to hide their bugs or not tell the testers about new code they wrote in hopes that fewer bugs will be found. You can't win.

Don't take my word for it, read Austin's book and you'll understand why this measurement dysfunction is inevitable when you can't completely supervise workers (which is almost always).

-- Joel Spolsky on Measurement

The anecdotes above are just the tip of the iceberg. I've heard many stories over the years of harmful gaming of metrics. It is clear that you should not introduce metrics lightly. It seems best to either:

• Define metrics that cannot be effectively gamed; or
• Win people's trust that metrics are being used solely to improve company performance and will not be used against anyone.

Performance Appraisals

At a recent Agile metrics panel discussion, I was a bit surprised that everyone agreed that their teams had some "rock stars" and some "bad apples". And that "everyone knew who they were". And that you didn't need metrics to know!

That's been my experience too. I've found that by being an active member of the team, you don't need to rely on numbers, you can simply observe how they perform day to day. Combine with regular one-on-ones plus 360-reviews from their peers and customers and it is obvious who the high performers are and who needs improvement.

Though I personally feel confident with this process, I admit that it is subjective. I have seen cases where two different team leads have given markedly different scores to the same individual. Of course, these scores are at different times and for different projects. Still, personality compatibility (or conflict) between the team lead and team member can make a significant difference to the review score. It does seem unfair and subjective. Can metrics be used to make the performance appraisal process more objective? My feeling is that it would do more harm than good, as indicated in the "Software Metric Gaming" section above. What do you think?

Software Development Process Metrics

Lean-Agile City runs on folklore, intuition, and anecdotes

-- Larry Maccherone (slide 2 of "The Impact of Agile Quantified")

It's exceptionally difficult to measure software developer productivity, for all sorts of famous reasons. And it's even harder to perform anything resembling a valid scientific experiment in software development. You can't have the same team do the same project twice; a bunch of stuff changes the second time around. You can't have two teams do the same project; it's too hard to control all the variables, and it's prohibitively expensive to try it in any case. The same team doing two different projects in a row isn't an experiment either. About the best you can do is gather statistical data across a lot of teams doing a lot of projects, and try to identify similarities, and perform some regressions, and hope you find some meaningful correlations.

But where does the data come from? Companies aren't going to give you their internal data, if they even keep that kind of thing around. Most don't; they cover up their schedule failures and they move on, ever optimistic.

-- Good Agile, Bad Agile by Steve Yegge

As pointed out by Yegge above, software metrics are indeed a slippery problem. Especially problematic is getting your hands on a high quality, statistically significant data set.

The findings in this document were extracted by looking at non-attributable data from 9,629 teams

-- The Impact of Agile Quantified by Larry Maccherone

Larry Maccherone was able to solve Yegge's dataset problem by mining non-attributable data from many different teams, in many different organisations, from many different countries. While I found Larry's results interesting and useful, this remains a slippery problem because each team is different and unique.

Each project's ecosystem is unique. In principle, it should be impossible to say anything concrete and substantive about all teams' ecosystems. It is. Only the people on the team can deduce and decide what will work in that particular environment and tune the environment to support them.

-- Communicating, cooperating teams by Alistair Cockburn

By all means learn from Maccherone's overall results. But also think for yourself. Reason about whether each statistical correlation applies to your team's specific context. And Larry strongly cautions against leaping to conclusions about root causes.

Correlation does not necessarily mean Causation

The findings in this document are extracted by looking for correlation between “decisions” or behaviors (keeping teams stable, setting your team sizes to between 5 and 9, keeping your Work in Process (WiP) low, etc.) and outcomes as measured by the dimensions of the SDPI. As long as the correlations meet certain statistical requirements we report them here. However, correlation does not necessarily mean causation. For example, just because we show that teams with low average WiP have 1/4 as many defects as teams with high WiP, doesn’t necessarily mean that if you lower your WiP, you’ll reduce your defect density to 1/4 of what it is now. The effect may be partially or wholly related to some other underlying mechanism.

-- The Impact of Agile Quantified by Larry Maccherone

"Best Practices"

There are no best practices. Only good practices in context.

-- Seven Deadly Sins of Agile Measurement by Larry Maccherone

I've long found the "Best Practice" meme puzzling. After all, it is impossible to prove that you have truly found the "best" practice. So I welcomed Maccherone's opening piece of advice that the best you can hope for in a complex, empirical process, such as Software Development, is a good process for a given context. Which you should always be seeking to improve.

A common example of "context" are business and economic drivers. If your business demands very high quality, for example, your "best practice" may well be four-week iterations, while if higher productivity is more important than quality, your "best practice" may be one-week sprints instead (see the "Impact of Agile Quantified Summary of Results" section below for iteration length metrics).

Team vs Individual Metrics

From the blog cited by Athanasius:

(From US baseball): In short, players play to the metrics their management values, even at the cost of the team.

Finding the Right Metrics

More interesting quotes from this blog:

The same happens in workplaces. Measure YouTube views? Your employees will strive for more and more views. Measure downloads of a product? You’ll get more of that. But if your actual goal is to boost sales or acquire members, better measures might be return-on-investment (ROI), on-site conversion, or retention. Do people who download the product keep using it, or share it with others? If not, all the downloads in the world won’t help your business.

In the business world, we talk about the difference between vanity metrics and meaningful metrics. Vanity metrics are like dandelions – they might look pretty, but to most of us, they're weeds, using up resources, and doing nothing for your property value. Vanity metrics for your organization might include website visitors per month, Twitter followers, Facebook fans, and media impressions. Here's the thing: if these numbers go up, it might drive up sales of your product. But can you prove it? If yes, great. Measure away. But if you can't, they aren't valuable.

Good metrics have three key attributes: their data are consistent, cheap, and quick to collect. A simple rule of thumb: if you can't measure results within a week for free (and if you can't replicate the process), then you’re prioritizing the wrong ones.

Good data scientists know that analyzing the data is the easy part. The hard part is deciding what data matters.

It has taken me a while to come around to the position that using code such as that in the title (or similar) is really cargo-cult programming. Given that seems to be the case, then I was somewhat dismayed to see that Moo automatically includes such when it generates a constructor.

I initially got around this by using before new to detect the presence of a ref instead of a string and issuing an appropriate croak message. (tip of the hat to the folks on #moose for that!)

However, several others on #moose, notably some of the authors of Moo, said that was not a wise practice, and indeed, I found that the code was position dependent in relation to where BUILDARGS is defined. Specifically, the code for before new must come after BUILDARGS is defined or else the return value from that gets blown away.

After asking a couple of questions, and getting the expected response of "Well, write it yourself!" I decided to do just that. I have a working patch to Moo that disables the code in question, replacing it with an exception.

My question at this point is this: Is the ability to force a constructor to be a "class only" method a viable feature to have in Moo (and be extension, Moose)?

The solution I arrived at after reading the code for Moo and a few other MooX modules led me to use an option passed into Moo's import method. This was mainly due to the fact that other than a significant refactoring of Method::Generate::Constructor it does not seem feasible to alter the code being generated.

While I am waiting for a review of my proposed patch, I'd like to hear the thoughts of my fellow monks if this is a desirable option to have.

I have just finished converting DateTimeX::Fiscal::Fiscal5253 from classic OO Perl to Moo. I would greatly appreciate any comments that result from taking a look at the changes. The test suite required almost no changes (mainly tightening up) so I am fairly confident that I have it close to being correct.

You can find the Moo branch here on GitHub: https://github.com/boftx/DateTimeX-Fiscal-Fiscal5253/tree/moo-delegation (The current release is in the master branch as one might suppose.)

I would like to get some feedback before I release even as a devel version since I am fairly certain it will pass the CPAN testers without a problem as it is passing a make disttest on my platform.

Update: Of special interest is that I had to place a modifier on "new" in order to force "new" to be a class-only method. That is, to ensure that one could not call it as $obj->new. This works as desired, but what is interesting is that the call to before new must come after the code for BUILDARGS or else the returned arg list from that gets blown up somehow before the object is instantiated.

As in every discussion there is always an out of framework post, not particularly related to other posts but a person asking needed to start somewhere so there you have it. This is that type of a post.

So I finally decided it is time to move on. I am not young enough to change the job and start all over with my 90' programming skills (like someone nicely phrased it in response to one of my previous posts here) and not old enough or should I say not arrogant enough to steal someones else's position in the company I currently work at. So I decided to start my own business. Since I've been in system administration and search engines for a while now and know a thing or two about it I decided to frame my business idea around it. However for that I need a financial injection and/or an access to a machine (hopefully not too expensive) where I can set up my Internet business. Since this is not a typical internet sales framework I cannot just set up a web site and start selling. I need a full unix/linux environment with ca. 100GB of disc space 4-8 cores 4-8 GB of RAM and an Internet domain. Is there any service that you know that can rent something like that? Second I never programmed any money transfer app. And have no idea how to do it right now so is there any perl module that can be used to do something like that? Or how do people setup something like that so a client can safely make a money transfer. Essential I need some basic first hand advice on how to start an online business. How to register it even if I am looking at some donation based system like here at perlmonks. I have been googling a lot about the subject but there are bunch of advices like

1. Have an Idea.
2. Register a company
3. Make a website
4. Make a gazillion dollars

Well maybe it is just me, but this is so vague that I cannot even begin to rationalize what the author meant or what was he/she thinking when writing something like that. So if there is anyone here that has some experience with it and willing to share I would gladly take some advice. (Positive or negative, both are welcomed)

cheers,

baxy

A decade or so ago, I was a conlanger. I liked to construct a language spoken by a virtual people I didn't however, in contrary to J.R.R. Tolkien, shape regarding their culture. Obviously I had a plenty of mind capacity for such things, and they called me quixotic. No more. Now I prefer programing and hobbies and a real life, too. I remember that the grammar part was in favour compared to developing a comprehensive lexicon so the language is talkable. Inflection and word order and features I liked in other languages however were far more interesting. The best part about it I found were its 35 cases, actually two case systems interacting like cogwheels.

I just programmed a word generator as a kind of tribute to that young nerd's passion. It respects how words can be shaped in a language and how not. In English for instance there is no word like quirge, but it still seems more English than, er, dampfschiff which is by the way, believe it or not, actually a word in a spoken language. These unwritten principles are characteristic for every language. Despite of their being unwritten, you can approximate these principles by probabilistic shares and the rand() of perl.

Even if you are not interested in conlanging, you might want to study this example how curried closures can lead to efficient code, i.e. code of which I am quite sure it's efficient, until someone will prove me wrong. For those who do not know yet: Closures are anonymous subroutines that cling, for their own life-time, to any used lexical variables which would have been garbage-collected otherwise on going out of their respective scopes. Higher Order is when you pass these subroutine references to other functions which call them back when appropriate. Learned that once from the High Order Perl book by Mark J. Dominus (available online but if you can effort it, consider buying it).

Update: Corrected name of the Higher Order Perl author. Further confused Currying and Higher Order. Will have to re-read myself the very book I cited.

-- flowdy

Hi,

I got laid off a couple of months ago and am in the midst of the whole job search process. With that in mind, one thing I am wanting to have a good idea of is my marketability.

I have been programming in Perl for close to 15 years now, but I think my use of the language has been pretty limited in scope.

Now, my last job included an application called CA eHealth. I did a ton of Perl programming involving this application. (In fact, I largely lost my job due to my employer’s decision to replace this application.)

There is one particular kind of scripting that I just love, but I am frankly unsure if it is sufficiently marketable. What I love doing is having Perl read data from a number of sources (databases, flat files) and produce desired data. I just LOVE that kind of coding. I love the whole portion in between those two points. Data aggregation, coming up with one or more nested hash data structures, looping through the hashes and writing out data as needed. Sometimes needing to do math and/or stats as part of the data aggregation. Mining through lines of text and db queries, and so on.

My question is, is just the above thing I love to do marketable? I honestly do not know, but I sure hope so.


Tony