mxh

Two thoughts: First of all, I'm a Python fan, and would like to use it client-side. But the truth is that ECMAScript ain't that bad, it's just a little strange. I'm not at all convinced that it's a huge win to use Python over ECMAScript, although I'd choose to, given the opportunity.

Secondly, this is cool, but if it depends on users installing a plug-in to IE for anything to happen, it's unlikely to amount to much. I don't see devs using anything (except Flash) that forces their users to choose between installing something (scary!) or going to another site.

This is more like the "Top Five Best-Known Hackers of All Time". Surely, the real top 5 (with all due respect to RTM), whoever they are, never saw the inside of a courtroom.

"None of us have taken time off in the last three years..."

...

"Just don't forget that you really aren't going to get paid in cash this week. Or next week. Or next month."

I don't know when you throw in the towel on your own baby, but it certainly seem questionable to join a startup that's been in crunch mode for 3yrs and still can't make payroll.

This may be a silly question, but it's something that's bothered me ever since Scribd launched: How do you handle all the file format conversions?

It seems to be that you couldn't be wrangling all those formats yourself, as many are undocumented, and/or hideously complex, and you've got to do a lot more work than simply reading the files, and not much time to do it. So, are there some off-the-shelf components to do this work (with reasonable license terms), or is the problem easier than it seems?

I'm most interested in the MSFT formats, which seem the trickiest.

Marc seems to advocate raising a ton of money, and then spending it prudently. Which sounds great, but I've read that many VCs aren't that happy with that plan. They want their companies to roll the damn dice already, grow huge quickly, or flame out quickly (allowing the VCs to focus on their few 'home run' companies). The relevant quote being "once you take a few million of my dollars, the clock is ticking".

Aside from a passing reference to not giving up control, I don't see this concern addressed anywhere in the article. Would a post-huge-money board really be ok with a "save it for a rainy day" money management strategy?

The article makes reference to the fact that once a dating site becomes free, you don't have to scan user content for contact info any more. I think this is a buried lede; free sites aren't just paid sites with lower costs, they have different (and sometimes better) characteristics than their paid counterparts:

1.) Easier to signup - no need to force users to drag out their credit cards

2.) Enhanced anonymity - no need to tie an account to a real person

3.) Simpler operation - the site owner doesn't have to worry that the users are ripping him off/subverting his business model

4.) Lower expectations - how much can a user complain if he's getting something for free?

I think this could also be stated as:

If no one is doing what you're doing, it's probably not because no one thought of it before. There are two reasons an idea might never have been developed into a product before:

1.) Seems too hard 2.) Seems unwanted

Underlying these perceptions might well be the reality that the idea is too hard, unwanted, or both, but the perception is what matters.

So, if you're doing something new, most people will think it's either too hard, unwanted, or both. (Otherwise, lots of people would be doing it.) They might be right, but their opinions are not meaningful data points.

If you want to innovate, don't waste your time arguing.

The 2nd-to-last paragraph is what caught my attention ... if the legislation discussed passed, and if it were extended to other private partnerships, that seems like it would eat into the mgmt. fees of most VCs.

And that would seem likely to reduce the number of VC firms, at least. (Less profit = less motivation to enter the business.) It's very speculative, but interesting.

My definitions:

Web App: Application developed entirely using open standards, which (aims to) run in any browser, requiring no additional vendor plugins.

Desktop App: Application designed to launch from the user's desktop, without reference to a web browser.

Thick Client: Any desktop app that makes heavy use of the Internet, or any "Web App"-like thing that makes use of non-standard extensions to the browser.

Reasonable people may disagree on this, and the differences are pretty fuzzy, but my distinctions are based upon the development problems I feel each type of approach must address.

Web App: Works on (most any) platform, no installation, but you have to deal with cross-browser issues, and you're limited by the technology of the web (no local FS access, for instance.)

Desktop App: Tied to one architecture (or worry about JVM incompatibilities), installation hassles, but you get an incredibly rich set of capabilities.

Thick Client: If it's a desktop app that talks to the network, you've got all the desktop tradeoffs, plus richer data, plus worries about what happens when the user works offline. If it's a "Web App" using non-standard extensions (Flash, Silverlight) you've got to worry about whether the user has the right version installed, and licensing hassles and/or fees on the server side and/or in the toolchain. You're also constrained by the capabilities of the extensions, but, in practice, development seems easier than for a pure Web App.

I realize that these definitions mean that lots of things (YouTube) people think of as Web Apps aren't (to me, anyway), but I'm okay with that. It seems a crucial distinction that, once you're targeting the Flash client, you're not really solving the same set of (technical) problems as someone targeting a browser.

I thought this was interesting; it's always nice to see startups selling something other than advertising.

"Google runs entirely on OO languages (C++, Java, Python)"

There're some interesting questions about what makes a language an 'OO' language. For one, does a language need only to provide facilities for OO, or must it enforce OO approaches? For another, what flavor of OO must it support or enforce?

Most 'modern' languages will provide facilities that can be seen as supporting some form of OO; data structures are mandatory, and methods are an easy extension of them. However, most of these languages don't mandate an OO approach to coding; for instance, C++ (aka C/C++) doesn't and Python doesn't, although Java may be more dogmatic on the question. (A language that mandates OO approaches wouldn't be to my taste, because I don't believe those approaches to be appropriate for all problems.)

So, it's a bit of a non-sequitur to proclaim that the virtues of OO are demonstrated by the use of 'OO languages'; unless those languages mandate OO approaches you'd need to look at how those languages are being used in practice.

... and this level of abuse is earned just by using a standalone tool in officially un-approved-of ways. It seems that using a fully-locked-down platform (e.g. F8, eventually Silverlight/PopFly or Flex) is an invitation to even greater headaches.

I understand the enthusiasm for these things - they promise to make development easier. But given that cell carriers take advantage of their devs, and that game console manufacturers take advantage of their devs, I wonder why more people aren't more worried about 'platform' taxes appearing in the future.

Very true. "A friend of mine" was writing some ad copy for a job req. "his" employer was trying to fill, and wanted to use the word "hacker" in the headline. (E.g. "Looking for a great hacker", or somesuch.) The (oursourced, non-technical) HR dept. would not permit the use of that word, as it sounded too much like advertising for someone to break the law.

I've always believed that these terms told you a lot about the speaker, but I was shocked to see such a clear, real-life case.

Am I the only one who parsed that URL as "wait errant"?

Interesting .... thanks for the link. Though, in the context of that report, a 'new business' need consume only 15 or more hours per week to 'count'. Such low-intensity enterprises aren't really what the YC (or VC) crowd is interested in (whatever their other virtues), so I think there's a bit of an apples-and-oranges problem with the 550K/mo number, even assuming there are no deeper statistical problems with the report.

Wait .... 550K businesses a month = 6.6M businesses a year. There are 300M _people_ in the US (including children, and others not immediately suited for the entreprenurial life). If each business was started by only one person, that would mean that about 1 out of every 45 people started a business in any given year.

That seems high. May I ask the provenance of the 550K/mo number?

David vs. Goliath is a nice story, but it's always better to have someone bigger on your side. MSFT has a huge warchest, and therefor a big advantage going against the FSF. I'd look for support for the cause of free software from the likes of GOOG, IBM, and YHOO (assuming the latter isn't bought by MSFT); companies that make a lot of money by not paying the MSFT tax.

Now, could MSFT acquire/bribe all its deep-pocketed natural adversaries in a patent fight? My opinion is not; they don't have _that_ much money, and the GPL limits their ability to engage in cross-licensing maneuvers.

Still, this story makes me want to chant: "Hey, hey, ho, ho, software patents have got to go." I can't think of a damn thing s/w or business process patents have ever contributed to innovation. (Counterexamples welcome.)

The square brackets are SQL syntax; they let you reference column names containing 'funny' characters, in this case whitespace.

It's definitely not ASP. BTW, ASP doesn't really 'look' like anything, since it's just an environment for the actual languages (vbScript, ECMAScript, C#, whathaveyou) that drive the server side.

I wish I didn't know that.

Scribd _is_ obvious google-bait. Google's "all the world's information" bit makes something like scribd almost irresistible to them. Of course, google may try (and/or succeed in) rolling their own, first.

But kudos to the scribd guys for delivering a simple twist on Napster/Flickr/YouTube (share music/photos/video) that no one else had done. I expect they'll do very well out of it.

I suppose it depends how much you're into formal methods, etc. The YC, for instance, was developed (if memory serves) to obviate the need for assignment WRT recursion, because assignment makes it much harder to reason about programs.

In principle, CS theory can help you prove that programs are correct; Dijkstra was big on this.

"It turns out he wasn't noticing the space between the 'o' and the 'I' in 'Do It'; in the sans-serif system font we were using, a capital 'I' looked very much like a lower case 'l', so he was reading 'Do It' as 'Dolt' and was therefore kind of offended."

"He [said], 'I'm not a dolt, why is the software calling me a dolt?'"

It's a good thing they resisted the urge to make the obvious joke, and just changed the software.

Just to clarify: The whole point of a healthy industry (i.e. a vibrant, competitive one) is to provide low-cost, high-quality products to the consumer. Standards are a means to that end. Vendor lock-in, on the other hand, stifles competition, and produces lower quality at a higher price.

Consider how unfortunate it would be if, for instance, the WWW were built not on W3C standards, but on proprietary technology; say IIS, IE, and an undocumented network protocol. In that case, innovation would slow remarkably. Competitive browers (Firefox, Safari) could not be developed because they couldn't talk to the IIS servers (the protocol is secret, remember) and competitive servers (Apache) could not be developed for the same reason. The only way competitive products could emerge would be for some firm to develop a client/server/protocol stack, and market the hell out of it until it achieved sufficient penetration to be successful.

It's a lot cheaper to compete when the protocol is open, competition is what produces great product, and that's why I think it best that solutions built around open standards not just thump, but crush any proprietary alternatives. Multimedia is a bit of a disaster in this regard, but I see no reason RIA need go the same way.

I can't get on board with the notion that standards are only good for lowest-common-denominator sorts of things. The internet seems to work ok, for instance. As for people going 'above and beyond', to the degree that such activity represents the 'extend' part of E/E/E, standards can't prevent it, but responsible people ought to shun it.

As for MSFT's ability to cram product: they can bundle with an OS that has 90%+ of the market. I don't think their ability to guarantee their products penetration (not necessarily use, not necessarily sales, but availability) is a matter for serious dispute. If they want Silverlight to be a viable platform, they just have to roll it into IE 7.X or IE 8.0.

An excellent put-down of vendor lock-in. For the good of the industry, I hope he's right, and the standards-based web continues (for the most part, i.e. overlooking multimedia) to thump vendor-specific solutions.

However, MSFT has a much greater ability than either Adobe or Sun to cram things down the user's throat. They have their 'update' services, their ability to bundle, and their presence on the desktop with which to nag. This could help them w/ Silverlight market penetration, a prerequisite for making that platform a competitor to the standards-based web .

Question: Is silverlight really part of "the web"? For that matter, is Flash? (Were applets?) It seems to me that "the web" consists of things built around W3C standards, to a first approximation.

Just because something plugs into a browser doesn't make it part of "the web"; that's just a sneaky way to put a thick client on the user's computer.

I'd argue that Silverlight _might_ be the future, but it's not the web. (Whether that would suit MSFT just fine is, of course, open to speculation.)

FYI, 'x' is Jessica Livingston, according to the YC site. TLB presumably lives in the Bay Area; his Anybots company is here, anyway, and hosts the WC YC offices.

Which raises an interesting question as to how the YC partners handle the EC/WC split between the SFP and WFP. Do different partners play bigger roles in different programs? Does PG rack up the airmiles/spend 3mo of the year away from home?

Is cloning involved?

As a practical matter, if I read the article correctly, VBA is only going away for the Mac. Is that the platform you're working on? If not, no problem. If so:

1.) Not upgrading sounds good .... it's cheap! 2.) You could move to a PC. 3.) You could learn the beauties of CSV format, and the language of your choice. I like Python, but something else might be better for your domain.

The Larger Issue here, IMO, is the peril of using proprietary technologies. It seems unwise to build any long-term project in something a vendor can too easily end-of-life.

"and 99% of what we program"

That seems consistent w/the "95% of what we teach" number. As Douglas Adams wrote, a computer "...was, on the other hand, very good at being a slow and dim-witted pupil. ... And the more slow, and dim-witted the pupil, the more you have to break things down into more and more simple ideas. And that's really the essence of programming."