Slashdot Log In
Philip Greenspun Answers
from the my-brother's-the-real-doctor-in-the-family dept.
How do you expect this degree to be worth anything
(Score:5, Interesting)
by slashdot-terminal
People spend thousands on a university education because they think it's a benefit and will lead them to a good job. Are there any employeers that have or would be willing to accept a graduate of your university. Do you think the numbers will increase?
Phillip:
We're not a vocational school. If someone wants to get a high-paying job, I would hope that there are easier ways to do it than working through a formal computer science curriculum. We even suggest one on our site: visit education.oracle.com and learn to be an Oracle DBA.
That said, the Baby Boomers are beginning to retire. Employers can't afford to exclude people who are qualified to work. Someone who can show a potential employer some running systems that he or she has built, a transcript with good grades from ArsDigita University and recommendations from a few of our PhD CS nerd instructors should not have any trouble getting a job.
Travels with Samantha
(Score:5, Interesting)
by X
I remember reading Travels with Samantha when it first came out on the world wide web (some of my first real reading on the web). What struck me about it, aside from the fact that I enjoyed reading it, was how much of yourself was laid bare in the story. Publicly exposing oneself like this is something that celeberties do all the time, but it was (particularly at that time) a rare thing for Joe private citizen to do (although certainly within your nature ;-).
I'm wondering you can describe what happened as a result of exposing so much of yourself online. I remember reading the comments on the story, and there were certainly a wide range of responses, but I was wondering if you noticed any larger consequences?
Phillip:
Travels with Samantha isn't about self-exposure; it comes from the same motivation that leads people to open-source software: the desire to help people build on what one has learned or done. If I'd had more time or been a better writer, I would have tried to put the same ideas and experiences into a novel. But I didn't so I slapped it up on the Web :-)
Publishing the book online has had some huge consequences for my life, but not the ones that I would have expected. For example, I got a large number of questions about photography. I thought it would be more useful to people if I codified my knowledge in a set of Web pages (http://photo.net/photo). The codified content generated yet more questions so I implemented a database-backed discussion forum for the site. Voila! I turned into the publisher of a 50,000-member, million hit/day online community. Trying to serve the changing needs of the community led me to build the ArsDigita Community System (ACS). Trying to keep up with the companies that wanted systems built on top of the ACS led me start ArsDigita Corporation. Profits accumulated by the efforts of our 100 developers enabled me to start ArsDigita University.
12 hrs/day * 6 days/week == severe burnout?
(Score:5, Insightful)
by ToastyKen
When I read about Ars Digita University, the first thing that came to mind was what an extreme amount 12 hrs/day, 6 days/week, on A SINGLE SUBJECT is. I mean, there do exist people who can take that much work, but don't you think a large percentage of your student population would simply burn out?
I go to a major university, there's no way I put in 12 hrs a day of work, and I'm still already stressed out. And that's with multiple subjects so I can take my mind off of one and switch gears occassionally.
Do you have any plans to counter potential burnout?
Phillip:
A typical MIT student takes 9 courses in 9 months. ArsDigita University teaches 9 course in 9 months. Thus the overall pace should be similar to what has proven to be successful at MIT. Taking multiple subjects simultaneously has some advantages but it also requires students to be good at managing their time. Even within traditional universities there has always been debate about whether it wouldn't be better to focus on only one course at a time.
The Ars Digita University is cool, but...
(Score:5, Interesting)
by hey!
the real problem I see is that there are people with a clue, and people with degrees, and there's not necessarily much of a correlation positive or negative between the two. Ideally, to improve the situation so clueful people get access to the important ideas of CS and that employers get some better idea that when they hire a degreed engineer they're actually getting something worth a premium.
It seems to me that CS degree work should be opened to more people who would advance by demonstrating the ability to do real work integrating important theoretical CS ideas with real world problems. Yet the very format really excludes a great deal of people, especially those who have to work to support themselves.
Does the Ars Digita program offer any real advance in CS degree program quality or accessibility to people who would benefit themselves and society the most?
Phillip:
Imagine Jane Humanist. She went to college in 1985 and wanted to touch human lives. In 1985, computers were generally only found in specialized locations and had little impact on the average person. So Jane very sensibly elected to major in government or psychology or history. Fifteen years later, it turns out that computers are ubiquitous and that the most efficient way to touch a lot of human lives may very well be to build some sort of information system. ArsDigita University is intended to offer Jane Humanist a second chance so that her impact on the world won't be limited by her choice of college major back in the mid-1980s.
As for the "great deal of people" who can't travel to Cambridge, must work to support themselves, don't have high test scores, etc., we will support them via online lectures, course materials, and collaboration tools. That said, I doubt that the average distance learner will have enough motivation and discipline to come up to the MIT/Stanford level.
Question
(Score:5, Interesting)
by doonesbury
The idea that you propose is controversial, and potentially disturbing to the entrenched university/degree program - especially considering the billions that these programs earn based on the concept that the "magic paper" only available through degreed universities is the only qualification for intellegence and competence.
A) Where would you like to see this program move towards, in relation to universities;
B) Do you plan on a "pay" version, for people who can actually afford to pay?
C) The qualifications (and I took a *real* good look at them, I really want to go!) are a bit unusual - in that they require SAT scores.
I miss by 50 points, but isn't that exactly the attitude that you're trying to escape - that you need a standardized test to determine intellegence, that you need cash to determine eligibility? Or am I reading too much into the program?
Phillip:
With our pitiful $1 million/year annual budget we're not trying to shake a $15 billion organization like Harvard University to its foundations. Our relationship with other universities is pretty simple. We try to use their curricular materials where appropriate. We offer our curricular materials to anyone who wants them under an open content license.
A "pay" version? No of course not! The university is more to benefit the instructors (see http://www.arsdigita.com/asj/professionalism) than the students. We are privileged that they choose to hang out with us. Teaching is its own reward and is part of what we think of as the good life. (Note the "part of"; I personally wouldn't want to teach full-time.)
Our qualifications unusual in requiring SAT scores? Every college requires SAT scores! That's because they are a great indicator of someone's willingness and ability to sit down, do homework, take tests, etc. Also we're lazy and don't want to spend a week interviewing each student.
Are any Open Source databases production ready?
(Score:5, Interesting)
by DuBois
Philip:
I've read "Philip and Alex's Guide..." and hoped to implement your kind of website on my own server. But then I noted that Oracle requires thousands of dollars of licensing fees.
Have you used any of the Open Source databases like MySQL or Postgres enough to recommend one of them for a light-usage site?
Or perhaps none of the Open Source databases are yet ready for production use?
Phillip:
I talk about this a bit in http://photo.net/wtr/aolserver/introduction-2.html.
The bottom line is that for people who care about data integrity, concurrency, and 24x7 redundant operation, there really is not an adequate substitute for commercial RDBMes (even the commercial object database companies haven't been able to make any headway against the heavy-duty RDBMS systems).
Will the "University" be open or biased?
(Score:5, Insightful)
by Wee
I certainly mean no denigration by this, but will this "University" be universal or will it teach only concepts that use Ars Digita's preferred architecture: AOLServer, Tcl and Oracle? For example, you mention that 40 hits per second exposes the limits of Perl/CGI/DBI (which might be a questionable statement in and of itself), but I've worked on teams that built stuff which very nicely handles hundreds of hits per second using Java servlets and MySQL (for only one example). Will this sort of thing be taught in addition to the stuff you guys prefer?
I just can't help but think that the University will be biased in some way. Certainly, it's biased towards rote memorization in applicants (a rather inflammatory earlier statement alluded that a score of at least 1400 on the SATs was a requirement for being bright), but will the technological course material follow? I know that there's an Ivy League ethos that surrounds many people and institutions, and it would be a shame if that same sentiment ruled out "less bright" technologies as well as people in this new University. (And for the record: I work with extremely smart people -- some of whom never graduated college -- who use none of what Ars Digita uses, so I may be a little biased myself... ;-)
Another thought just hit me: Couldn't this University been seen as a thinly veiled way to promote Ars Digita's technological choices? Honestly, I don't know many people that actually use Tcl or AOLServer to do much, especially in a production environment. If future gradutes of your program are well-schooled in using those products, wouldn't they necessarily think of these technologies first when doing future work? Won't they be biased? So can't this just be seen as an "Tcl/AOLServer Mill"?
Again, I don't mean any slight or to seem like a troll, but this whole thing sounds to me like it'll be as well-rounded as any MCSE learning might be.
Phillip:
Sorry to disappoint you but we won't be teaching Java or MySQL or Perl. We'll be teaching the standard computer science curriculum that has evolved over 25 years at schools like Stanford, UC Berkeley, and MIT. We would certainly not teach AOLserver or Tcl because a student who couldn't learn these things in a day by him or herself would never make it through SICP, Discrete Math, or Algorithms!
(By the way, the ArsDigita Community System is a set of data models and workflow that has nothing to do with AOLserver or Tcl; for the presentation layer you can run Apache if you like and we've announced a 100% Java version.)
Opinion vs. Fact
(Score:5, Interesting)
by ryanr
I read "Travels with Samantha" not too long ago when I ran across a link to it. As a result, I poked around photo.net a bit, and ended up buying a paper copy of "Phil and Alex's Guide to Web Publishing." Great book, recommended. Even though it's on the web, it reads better on paper, the book is nicely put together, nice heavy paper, and the photos look good (all stuff Phil will tell you, too. :) )
My question is this:
In most of your writing, you often put some statement out there as fact, when it is actually an opinion. In many cases, I can spot it as such, and just roll my eyes a bit if I happen to disagree. Are you aware that you do this, do you worry about it, or do you expect your readers to spot it and take it as an opinion? Or are you a typcial college professor whose opinion IS fact, and won't be told otherwise? :)
The reason I ask is that I do a little writing myself, and I find it a unnerving to put something in print that becomes more-or-less unchangeable. I.e. I just worry about being "wrong" either because I am plain wrong, or wasn't clear in my statements.
Phillip:
When I first started writing journal papers (back in the early 1980s), I had a great editor named Curt Roads who crossed out all the occurrences of "In my opinion" and "I believe" from my writing. He said "It is obviously your opinion because your name is on the article and it is obvious that you believe this or you wouldn't be saying it." So I tend to shy away from weasel and waffle words.
Why tight coupling to a RDBMS?
(Score:5, Interesting)
by kslee
First of all, thanks a lot for Philip and Alex's Guide to Web Publishing and introducing Edward Tufte's excellect books.
My Question is:
What's the merit of tightly-coupled-to-Oracle architecture of acs(ars digita community system) as a web application platform? ZOPE is in my mind as the not-tightly-coupled-to-any-RDBMS web applicaton platform?
Some people came to ZOPE because they can not afford an Oracle(in my case, the prefernce of python to tcl played a lot).
Or any comment on the web application servers/platforms which does not have the honor of being commented upon in your web tools review is apprecitated GREATLY!
Phillip:
You're welcome on the Tufte recommendation. His books are certainly easier to install on one's shelf than Oracle is on a Linux box :-)
There are two separate issues raised by your question, both of them worthwhile. The first is "If you're going to use an RDBMS, why not make your code more portable and abstract instead of hard-coding in Oracle dependencies?" If you read http://www.arsdigita.com/asj/professionalism you'll see that we value innovation over market share. If one has limited time and resources, it is better to worry about making the application great rather than "Does it run with Informix 9.05 or SQL Server 7?" We're not in love with Oracle but it is adequate for what we want to do and therefore we don't spend time and effort on portability.
The second issue is "Why not an object database?" If one's source of persistence were an object database, that would change all kinds of assumptions about Web development style. For one thing it would make object-oriented languages such as Java much more useful. Sadly, however, the object database folks haven't successfully tackled some important problems that the RDBMS handles very well: concurrency and isolation of application program from database.
So of course ZOPE is an honest effort and a great contribution. But most companies are forced by practicalities to be very RDBMS-centric and I think that makes our suite of software more useful. Also, we attack a much higher level of the application stack than the ZOPE guys. See http://www.photo.net/building-community/infrastructure.adp for a draft book chapter that I've written that talks about this issue.
Techno-social considerations
(Score:5, Interesting)
by jellicle
The technical challenge of building an online community is less than half of the total work involved. Social considerations are tremendously important, and a change in one line of code can totally change the flavor and viability of a community. It is my suspicion that ArsDigita has not yet run into communities as challenging as Slashdot, that is, places where some percentage of the population is dedicated to destroying the place through denial of service attacks of various forms. The challenge is to enforce some level of responsibility without eliminating anonymity, without being called a censor, without tracking users like a stalker... Few if any online communities can be said to have gotten it entirely "right", but somehow the majority of real-world communities manage to have civil discourse at a reasonable level. This is really a sort of sociology problem - and hardly an easy one - which is instantiated in computer code. How would you solve these problems? Or, more precisely, how would you start to learn how to solve these problems?
And no, "Trial and error." is not an acceptable answer. :)
Phillip:
Actually at photo.net we've run into many of the same problems, but probably not as severe as at Slashdot. First, a couple of things that make photo.net simpler than Slashdot. We are a PURPOSEFUL community. Everyone at photo.net wants to become a better photographer and therefore that implies a shared purpose of user-to-user education. Second, we are anchored by a lot of magnet content that I wrote, e.g., http://photo.net/photo/tutorial/
Most people who don't agree with at least part of my way of looking at photography education will turn away from the site before becoming involved in a discussion.
What makes Web development tougher than other kinds of software engineering that I've done is the constant challenge of idiosyncratic humans. For example, in the 1980s I did a lot of computational geometry, graphics algorithms, and computer-aided design programming. The algorithms could be hard to understand but once coded they would work properly forever without modification. The reason that it was possible to completely solve the problem was that the input to these algorithms was machine-readable and fixed in format. In Web development, however, user interfaces and annotation that work for 50 simultaneous users invariably break down when 50,000 users pile in (a friend of mine runs cnn.com and told me that they once got nearly 1 million comments on an article! The ArsDigita Community System would present these in a flat list on a single page... not very modem-friendly!).
What are you shooting now?
(Score:5, Interesting)
by HerrNewton
I know this will likely get pushed aside by more net oriented questions, but what are you shooting for a body, lenses and film these days? I know, different tools for different occassions, but what is your most common setup?
Phillip:
These days I'm mostly chained to my desk at ArsDigita and hence I don't do a lot of work around Boston. ArsDigita now has offices in Tokyo, Pasadena, Berkeley, Austin, Atlanta, Washington DC, London, and Munich. This plus the occasional speaking engagement means that I travel a bit and fill in the spare hours with sightseeing and street photography. My favorite tool is a Fuji 617 panoramic camera, loaded with Ilford 3200 black and white film. This enables handheld photography and the production of a 6x17cm negative, which enlarges nicely to cover the walls in our 100,000 square feet of office space worldwide. I also do a lot of work with a Canon 50/1.4 lens, usually attached to an EOS-3 body loaded with Fuji NPH film (ISO 400 color negs with a subdued palette). I was just in Florida and couldn't resist buying a $10,000 600/4 IS lens to take pictures of birds. Check out
http://photo.net/photo/pcd4101/index-fpx.html
and
http://photo.net/photo/pcd4715/index-fpx.html
for examples of new work (unlinked from anywhere so far, exclusive to Slashdot readers!).
Bottom line is that a monkey can take good pictures. Talent is cheap. Time is precious. If you put a lot of time and hard work into photography, you'll have good pictures. If you are a burnt-out nerd grabbing snapshots in spare moments you'll have... the stuff above.
------------------
I added these myself because I thought they were worthwhile.
Do *you* ever suffer from burnout?
(Score:4, Interesting)
by timgriffin
Given that your interests/tallents are so many and varied, how have you found the last several years of corporate CEO-dom? Inevitably there isn't time to do everything. What do you miss? What's most rewarding? What's most irksome about your responsibilities? What does the future hold for you?
Phillip:
Do you know the scene in King Lear where he meets Edgar disguised as a pitiful guy wandering the heath in rags? And Lear asks "What happened to him? Does he have daughters?"
I used to think that assholes became business people but now I realize that the causation works in the other direction. Now when I find a deranged person shouting at someone, I ask "Oh, does he have employees?"
If you have high standards and you grow fast you will inevitably find yourself having to tell people how they're not meeting your standards. Oftentimes I'm struck by how much better a job someone at ArsDigita has done than I would have done. But I only get to spend 15 seconds enjoying that feeling before moving on to attack a problem. Being a manager means focussing on stuff that isn't going well. It made me crazy enough that I was really happy to hire Allen Shaheen, one of the founders of Cambridge Technology Partners, to take over as CEO. Now I'm Chairman and will concentrate on technology (what problems to attack and how to attack them), education (how do we turn smart people into great Web service developers), evangelism (sadly I'm still the best spokesperson for ArsDigita), and company culture (no we won't hire people just to grow; yes your boss will be able to do your job; yes we will hold programmers accountable for the overall quality of the Web service; yes, dammit, we ARE going to spend a fortune on beach and ski houses where programmers can build modules, write documentation, journal papers, and book chapters).
Your outlook on industry partnerships?
by petervessenes
Phil, I own what is, to my knowledge, the third ACS based company in the world, ybos.net. We have a fairly aggressive growth plan, (more aggressive than furfly's for example), and I have a number of questions:
What's Ars' outlook on industry partnerships going forward? We're too small still to do the projects you guys want, (million+/year) and I don't think we'll be there for at least a year or two. I believe that making partnerships, and building relationsips with companies like ybos is important for you as you go forward: more alternative service providers gain you mindshare in the same way that giving away a year of free training at Ars U does.
How do you feel about ACS/Pg? Using Oracle is a major blow to doing smaller projects, obviously. Also, I know the state of Postgres two years ago, so I don't blame you for the switch to Oracle from Illustra, but do you have intention to backport to a more open database architecture, or 'bless' Ben Adida and co's work on the ACS/Pg? I think what appeals to me about ACS/Pg is not Postgres (rather obviously), but the more open nature of the development -- Ybos has begun releasing useful ACS modules to the public, and enhancing some slow-moving Ars ones, and it's a medium-level frustration that they'll never get rolled into your toolkit, or that we have to develop side by side. (for example, bryan che kindly lent us his data model early for the events module, but we developed about half a module under his data model before you released the newer module, and we scratched it and started over.) This leads to my final question:
Do you have thoughts on the relative openness of the ACS development? Would you consider an 'inner circle' development model that would let confirmed developers check code in and out of the development releases? I think that you'd see some significant benefits. I ask about this rather than a 'true' open source system because I'm betting you'd say "no way" to an aggressively open model. I probably would, too.
Meanwhile, I hope you're well! Congratulations on the recent funding. I hope we're not far off.
Phillip:
Ybos isn't the third ACS-based company! Just about every European country seems to have a collection of monster developers who've started an ACS-based services company. There are some great guys in Brighton, England that I'd kill to hire, for example. But I digress...
We believe so much in partnerships that we our very first MBA was Cesar Brea, a Bain refugee, our VP of Business Development. We could not have gotten Siemens without Boston Consulting Group. So we make partnerships all the time but we just don't bother to announce them with press releases and stuff (probably we should).
As far as ACS/PostgreSQL goes, we've given the project a free server and definitely support it. I offered money to the PostgreSQL group to pay for them to implement an "Oracle syntax SQL parser" (so that all kinds of Oracle-based apps could run on Postgres, not just ACS).
On the cathedral/bazaar split let me say that we've taken in lots of good ideas from the community. It hasn't been as formal as I'd have liked so we're looking to hire a whole bunch more dedicated toolkit developers who will have time to look at CVS diffs from outside developers, etc. We're not quite ready to go for the public CVS tree because we change our core structure too much. Maybe in ACS 5.0 we'll be able to do that (this fall?).
Data Modeling Tools
by Tassach
Dr. Greenspun,
Given that the ArsDigita Community System is so heavily database-driven, I was wondering what tools you use for data modeling and schema management.
What is your opinion of modeling tools like Sybase's PowerDesigner and Platinum's ERwin? What kinds of tools do you think are necessary to facilitate the development of highly portable, vender independent database designs? Finally, what is your opinion of UML and to what extent does ArsDigita use it?
Phillip:
First, my brother Harry is a real doctor so I'm forced to go by "Philip Greenspun, merely a PhD". The tool that we use for data modeling and schema management is ... GNU Emacs! E-R diagrams are basically useless once a data model gets beyond a certain size. And for smart people they aren't all that useful for small and medium-sized data models. UML would be useful if one could build entire Web services from the UML spec. This is kind of thing that we teach our students at MIT: come up with a machine-readable specification language and write a program to generate the programs that run the site. See http://photo.net/teaching/psets/ps4/ps4.adp for an example.
Bottom line is that Emacs + clever programmer will always crush a fancy commercial tool + weak programmer.
-----------------
Other interview notes: We're still waiting for answers from RMS and SCO. Next Monday we'll be seeking questions for Douglas Adams, and Thursday we'll need questions for Metallica about their tiff with Napster. If you know someone you'd like to see interviewed here (*and* you have contact info for them), please send e-mail to roblimo@slashdot.org. (Don't bother telling me we ought to get Stephen Hawking; he's already our single most-requested potential guest. I've e-mailed his graduate assistant with an interview request but have not yet gotten a response.)
- Robin "roblimo" Miller
Nice idea, but ... (Score:3)
Uh, from my expierence this not the way to go. Having studied mathematics I got the impression that complex concepts need some time, I remember a professor saying about the tensorproduct that "you don't understand it, you get used to it" and consequently always introduces that concept to first grade students.
This should apply to other fields too.
And my feeling is that I get some more insight into some things even without actually looking at them/thinking about them. There seems to be a mechanism somewhere back in our brain which put structure into knowledge, but this thingy is not very fast.
Re:Still a typical MIT bigot? (Score:3)
You cannot replace "Emacs" in that statement with "any text editor". Evangelizing aside (at least as much as is possible) I'll try and explain why I think that to be so.
To start with, Programming tools run the gamut from "you must know every ounce of code in the system" (text editors) to "Look at this picture of the whole system, the tool will manipulate everything for you" (Rational Rose, TogetherJ). Things like IDE's are just a step down from RAD tools like Rational Rose, indeed the Rose type tools slowly march downward to taking over IDE functionality (being able to generate more and more code and even compile for you), just as the IDE's slowly move toward functionality usually thought of in a RAD tool (like generating classes or code metric reporting).
The reason that I think his statement about Emacs to be true is twofold - first, that you can perform in Emacs almost any operation a RAD tool can do in terms of structuring code. Secondly, you can do many thing in Emacs a lot faster because in almost every IDE or RAD tool they do not let you use the keyboard enough when it is appropriate and do not let you customize what happens to a great enough extent.
In Emacs, I can write small macros and/or lisp fragments to generate get/set methods or whole types of classes. I can write code to apply design patterns (such as Proxy, Container, Factory, and so forth). I can refactor classes that have been badly split or not split in the right way to bring cohesiveness and beauty back to the whole.
In short, I can do almost everything apart from some of the more detailed reporting that I can do with a RAD tool - the only difference is that a lot of the diagram is in my head instead of on the screen.
To further my point I'll take a simple example of generating a get/set method in a class. Using TogetherJ (currently my favorite RAD tool by a wide margin over Rose) or Forte (great Java IDE) is about the same - I am presented by a dialog listing the attribute name I wish to have, the type of data, and so forth. I have to go through the dialog and then dismiss it - while I can use the keyboard for pretty much the whole dialog I usually at least had to use the mouse to initiate the action.
In Emacs, all I have to do is type something like "String userName", activate some function, and get
The difference is that in the first case, the tools make sure to ask me for many details about what I want - this makes going through the dialog slower as I have to think about each option. In the case of Emacs I can totally customize what I want to happen based on the most minimal input possible. If too much happens (I didn't want a set method but one was generated) I can quickly remove that one section more quickly than I can with even a mouse (with the Meta-movement commands. Sculpting by carving away something that is there is always easier than adding many small parts together to make a whole.
I've tried to think for a long time about just why I am so much more productive in Emacs than I am in an IDE (even for GUI work, I can put together a whole panel full of complex elements a lot faster in Emacs than in a GUI IDE). I think it really is that Emacs helps you to be able to do complex operations quickly and move around code fast enough that you can maintain a large picture of a system in your head through constant refresh.
I do think modern RAD tools like TogetherJ have finally started to be useful - mostly because finally they can backparse EVERYTHING I've done outside the tool and let me manipulate things from there. That way I have access to really great things like automatic UML diagramming based on what I've actually done (not what I think I'll be doing) and sequence diagrams to help me show possible flow through the system.
The time has come when the tools we have are at last useful, but Emacs is still for the most part the best primary means of creation.
Re:Why do people go to college? (Score:3)
Education is what you make of it. The quality of instruction often *can and will* be worse at some cheaper schools, but that isn't necessarily a money factor, and, as I think most
Smaller class sizes and non-lecture based learning also helps a lot with the student/faculty interaction problems you mentioned. There were several professors at my school that took the time to make sure they knew everyone by sight, even in the mid to large lectures. That really helps later on. It makes it easier to talk to them, and helps out when reccomendations come around
I've already posted some of my thoughts on the SATs... and most of your thoughts are right on target. I have concerns on requiring SATs for this program, but, hey... they *do* need to restrict is somewhat. Of course, there are many people who never took the SAT (maybe the ACT or some other test, or maybe it wasn't recent...). Do they maintain another path for these people to enter? Certainly there are a lot of qualified people who never had the opportunity to fill in lots of circles with a #2 pencil.
I don't think their score was all that remarkable, so it should be low enough to counteract some of the socioeconomic factors... my high school graduating class (~100 students, public school) *averaged* about what this required score was... this shouldn't be too limiting. I think the national average is ~1000 - not too tough.
If they took IQ scores instead, people would have the same gripes.
Re:Still a typical MIT bigot? (Score:3)
Also, just because one example of a Lisp program (Yahoo store) doesn't work a well as you'd like doesn't mean Lisp is a bad language. Think of all the shitty C/C++ or Perl code that sucks even harder.... By your metric you've just eliminated those languages from any kind of competition.
Autocad is (or was) largely written in Lisp, and Nichimen graphics http://www.nichimen.com [nichimen.com] sells one of the more popular 3D graphics engines for artists of console games, all written in LISP that's right, high end 3D graphics in Lisp...
Face it, most people who think Lisp is just for ivory towers, or is too slow, or blablabla don't know anything about the language...
For Scheme see : Structure and Interpretation of Computer Programs, and Scheme and the Art of Programming...
For Common Lisp (the industrial grade standardized lisp) see Paul Grahams books "On Lisp" and "ANSI Common Lisp"
Download one of the free Lisps: clisp.sourceforge.net or www.cons.org/cmucl...
Read "comp.lang.lisp" archives...
Photo.net Slashdotted (Score:3)
When the system recovers, I would be interested to see the stats. What was the failure point? The database connection? AOLserver?
Steve
the biggest ACS communities are about 600,000 user (Score:3)
About 1/3rd of ArsDigita customers started as raw startups (www.infirmation.com, away.com, uslaw.com (recently linked as AltaVista's exclusive Law channel)). We've watched them grow. I wish I could tell you that they've all started to turn a profit because ACS is so great. But as far as I know they are all more concerned about building market value, company size, etc. Many of them seem to be getting acquired for big bucks, raising additional stupefying sums of venture capital, etc. So they are achieving their business objectives but I doubt that very many of our startup customers have to pay income tax.
Anyway, ACS per se can scale quite nicely if you budget for a big RDBMS server and a rack of small Web servers and a load balancing switch (about $300,000 all together). But the users might not be happy because there will be too many comments on each page, etc. It is up to the publisher to use the user groups features intelligent to split up the community a bit.
We do have someone else talking to the public! (Score:3)
and they end up having to learn Tcl and AOLserver programming as a subtask (we never explicitly teach it them; they are expected to pick it up).
The whole point of Tcl is that you're not supposed to have to be a programmer to use it! Read Ousterhout's original paper on why scripting languages are a good idea.
This isn't a theoretical point. We've taught software engineering for Web apps to several hundred people now. We've used AOLserver/Tcl precisely because we don't have to spend any class time teaching it. We could use a Java version of our toolkit for future courses but I'm not sanguine about the prospects of introducing Java as a two-hour subtask during pset 1.
Anyway, we do have marketing shills at ArsDigita now, but I feel like the Slashdot community deserves a response from our tech staff (of whom I'm still titular head).
Bottom line as far as I can tell is that there are a lot of folks on Slashdot who want to believe that they are smart even though they received poor SAT scores and did not graduate from college. They think that ArsDigita U is passing judgement on people who didn't graduate from college or got trashed on the SATs. But it isn't so! We're simply trying to copy Harvard and MIT as much as practical and thereby reduce our risk of failure. I didn't even graduate from high school so I'm hardly in a position to call people losers because they don't have impeccable credentials.
IDE = integrated development environment? (Score:3)
If so, I'm a big believer. That's what we had on the old MIT Lisp Machine back in the late 1970s and it was much more productive than programming with raw tools. The thing that the Lisp Machine lacked was a database for persistent storage and hence it really isn't practical to use that old environment for developing Web apps.
So anyway, I love great programming tools and wish that we had better ones for the work that I do. But I don't like to get religious about it. A really good programmer in the end will be able to debug a system with print statements.
About Zope (Score:3)
GNU your life ... (Score:3)
This is rather amusing hehehe [paylars.com]. Wonder if someone could change that to GNU, BSD, Linux, or heck lots of things apply there.
What I want to get at is that I'm really glad that people slashdot sends questions to in interviews get seriously answered. And these are people with rather busy days that need to answer these questions. Good job /.
Reasons LISP is cool: (Score:3)
1: base support for complex types. How many times can you write a linked list before puking? Run-time type checking on dynamic tuples? Can you even do that in C? And for the record, 'strings'. Amen. C? No. (I don't know why you've grouped C, Tcl, and Perl together; I use Perl, elisp, Haskell, and a little Java).
2: Functions are data, vice versa & lambda calculus. Where do you think this came from?
3: A real, honest to god object system. I spit on your C++ structures-with-inherited-functions.
4: GC.
5: Correctness. Fucks' sake, all I ask for is an overflow should degrade to bignum. Not even Java does this. LISP's philosophy seems to demand that operations should not fail due to hardware design.
6: Simplicity. You can write a Scheme interpreter in one page; moreover, the concepts are all simple.
7: Ease-of-use. The interpreter doesn't have to be baby-sat through strong typing. It infers necessary types via what operations happen to a piece of data.
Sounds fun to me. Better than running down memory leaks, writing complex list eval fxns in C, bizarre type casts, and no lambda calculus.
I hope you can see the point that Greenspun tries to make, which is drawing attention to how a language should really be written. Languages evolve, and if we shout, maybe our voices will be heard.
Now, I use Perl, because Perl uses many of the above concepts (some poorly, like GC). Everyone else in my company uses C++; I'm slowly bringing them around. Where LISP comes into this is: I _really_wish they had been taught LISP in school, just to train them to think in the higher patterns that LISP allows. Working on C/C++/Java is too low-level and too inconsistent; people get lost and programs get buggy.
My complaint with functional programming circa 2000 AD is that the 'referential integrity' whores have crippled new FP languages for IO. Good-bye scripting...
So, you've reiterated your position; which we've heard like 4-5 times now. Your contention that LISP has no functional value is hopefully negated by the above. (and the polysemous use of 'functional'{think FP}) But you have failed to inform us why your preferred languages are better, other than that they are popular (which we already knew).
I refrain from defending Emacs, in that others have spoken far fairer and clearer than I. But xemacs-mule is much easier to work with than any other multi-language editor I've found.
Naive (Score:3)
If he thinks an unaccredited school's degree is going to be worth anything, he's smoking serious crack.
The regional accreditation system (see here [edupoint.com] for some information about accreditation) is not perfect, but there has to be some standard in place. Accreditation not only takes into account academic standards, but also financial standards to make sure the university is not going to go bankrupt.
Yes, you can get a degree from a non-accredited University. But where the rubber meets the road is when you try and get a more advanced degree (Masters or Doctorate) at a "real" university. That's when you find out your l33t degree is worth exactly zero.
Not to mention that when your employer looks at your degree, all they will see is a big red "non-accredited".
Bottom line, if he wants this thing to succeed, he better get regionally accredited or it is doomed to failure.
--
Populist Crap (Score:4)
Someone else here pointed out that a guy at some small college is beating the pants off MIT at something. That is all well and good. The trick is that the rest of his department, and the other departments at the school, aren't beating MIT. They aren't coming close (no offense.) Consistency counts for something, and that is something most State U's don't have.
Oh, and one other thing- my classmates are of generally a ridiculously high quality of academic and intellectual ability. There are some exceptions, but they are just that- exceptions. At the average State U, the average student is just that- average. Being surrounded by other students who really give a shit makes a huge difference in the quality of your education, and that is something that a big State U just can't provide. You can call me a snob, but then go look at the stats for how many students in state colleges have to retake high-school level algebra and writing before they are even qualified to begin college work. Then go look at Duke's stats and see what percentage of kids have already passed Calc before they got into Duke.
No matter how you want to slice it, there is a difference between selective private institutions and big state Us. The resources(=small class sizes), selectivity, and across-the-board quality of professors and departments are just a step above. I wish this weren't the case- I look forward to the day when everyone can get the education I've been privileged to have. But let's not fool ourselves into thinking that this is already the case.
~luge(couldn't agree more about SATs, though)
ArsDigita's take on Zope (Score:4)
The document is actually very good and covers many of the advantages of Zope over the infrastructure that ACS [arsdigita.com] is built on (AOLserver, Tcl, Oracle, ACS utilities). There's a "killer app alert" for Squishdot [zope.org].
Weirdly, it doesn't seem to touch on Philip's arguments regarding the need for concurrency and insanely high performance.
Take a look at ArsDigita File Storage [arsdigita.com]. They spend a lot of time analyzing their competitors!
$50 million (Score:4)
On the Emacs front, it just happens to be the text editor that everyone I know uses so I employed the term "Emacs" for concreteness. Obviously Emacs per se doesn't help you understand a data model (though I do like to run SQL*Plus in an Emacs shell because it is easy to cut and paste queries).
Anyway, I'm not a Lisp programmer anymore! I write Java, PL/SQL, SQL, (gasp) Tcl, and sometimes even Perl!
Why do people go to college? (Score:4)
Oh, and the requirement for a certain score on the SATs? Frankly, the SATs are more a measure of socioeconomic status and ability/training to take the test than of any innate quality of intelligence. The requirement reeks of elitism and discrimination against those who do not have the opportunity to take review courses or other such techniques to artifically increase a person's score over what he or she would have ordinarily scored. Moreover, what precisely does the SAT measure? It certainly does not measure critical thinking skills, nor is it a predictor of future success (high school grades are significantly more reliable). Is it truly just for a university that nominally presupposes to offer open access and transparency to base their admissions to such a large degree on an indicator that lacks real validity and implicitly places certain groups at a distinct and quantifiable advantage?
BZZZT... wrong (Score:5)
Well, guess what, I think I've gotten my money's worth (well, my parents' money's worth).
Don't allow yourself (or others) to be held back by the notion that state schools are just as good as the Ivy League schools. They aren't. Regardless of how much money you have, if you're good enough to go to Harvard, there's a way for you to attend Harvard.
I have not taken any classes at MIT that are taught by grad students (the recitation sections are sometimes taught by grad students, but that's normal). In fact, the only class in recent memory that I've taken, that wasn't taught by a professor, was 6.916, Philip Greenspun's class (which was awesome, btw). Professors do actually know some of the students' names, you can visit most of them in their offices, etc. Some of the profs that teach freshman classes are hard to get to, though.
I do agree with your views on the SAT, it is partially just a game, and certain time/effort/money investments can be traded for higher scores. However, not many people, prep course or not, can break 1500 on the SATs. It is, at some level, some sort of indicator of critical thinking ability. If it were purely a measure of financial ability or "aristocracy", it would be thrown out immediately.
Most decent schools know that SAT scores don't mean much, and that's why there's stuff like an essay requirement, an interview, and teacher recommendations. It's the whole package that is considered, not just the SAT score.
Still a typical MIT bigot? (Score:5)
Bottom line is that Emacs + clever programmer will always crush a fancy commercial tool + weak programmer.
I brought this up when the AD university was announced, and I received quite a bit more ad hominems than I did real responses to my argument. My feelings about AD then are the same as they are now: this is coming from a person whose teaching is swayed by this bizarre bias towards LISP and Emacs. Honestly, why could the sentence above not be written as, "Bottom line is that a text editor + clever programmer will always crush a fancy commercial tool + weak programmer."? Is Emacs really so incredible that anyone not using it is not a good programmer (as Phillip implies)?
Let me reiterate my position: I think LISP and Emacs are used more for their religious value than their functional value. (Moreso LISP than Emacs, for people are actually using Emacs to develop things -- people in my office are doing it right now. But they're not writing in LISP, of course.) LISP, like Pascal, can be a great teaching tool, but is otherwise worthless. I asked people to tell me what has been written in LISP, and I received nebulous answers, "research projects," and ad hominems. However, if someone can produce some evidence for me that shows me that LISP is something more than just ivory tower bigotry (that means show me actual applications that were written in LISP and why the technical merits of LISP made it the optimal language to use for those applications), then my opinion about LISP will change. Since I see no evidence that LISP is anything but ivory tower bigotry, I think that Phillip's judgement must be questioned when he calls LISP "the most powerful and also easiest to use programming language ever developed."
A side note: are LISP people just jealous that C, Tcl, and Perl are much more useful (and thus used) than LISP?
And Phillip, if you're reading, I did not get a response to my last post. You can check it out here [slashdot.org]. I was enjoying the discussion and was hoping I could get some answers to my questions.
Flames, as always, will be ignored. And if you find yourself flaming me, then you are losing the argument.
SAT scores... (Score:5)
Speaking from personal experience (mine and my brother's) - we both scored well on the SAT verbal (760 and 800), but did either of us ever do a rough draft for our writing assignments? (nope) did we do things on time? (only when we felt like it) could we have gotten straight 'A's in English classes? (yes) did we? (no, we didn't care)
That score doesn't reveal any of that information (neither does the AP test for that matter...) - all it measures is your aptitude, not you accomplishments. Standardized testing is great for some things (like getting me into schools
maybe it would be time to upgrade from 180 MHz CPU (Score:5)
I slashdotted my own site! (Score:5)
I feel like an idiot. Generally I never publicly link these pages but just use them for my convenience in developing new content.
Doh!!!!!!!!
pretty unimpressive (Score:5)
Also we're lazy and don't want to spend a week interviewing each student.
um...that's pretty weak..you don't have to spend a week interviewing a student to get an idea of how dedicated/bright/talented he or she is...also:
That's because [SAT Scores] are a great indicator of someone's willingness and ability to sit down, do homework, take tests, etc.
Nope. Not true. SAT scores are a good indicator of future success which is very different from their "willingness and ability to sit down, do homework, take tests, etc." High SAT scores mean you'll likely be successful in the future because either 1) you are genuinely very adept at taking tests, or 2) you are rich and could afford tutoring, meaning you have a rich background and will be likely to have good contacts in the future. So what Ars Digita should be asking itself is is whether their admissions process can really separate between the two groups...and whether "test taking apititude" is something they care for at all... okay, and what about:
The university is more to benefit the instructors [snip] than the students.
umm.. this probably should be clarified a bit before anyone signs up..
ars digita arrogance (Score:5)
I couldn't help thinking about how the million dollars they wanted to charge us was going to go directly into stuff like the aeron chairs we were sitting on, the rooms full of sgi flat screen displays, and the ferrari out front.
Still, I'll admit that they do some great work on a very specific class of problems.