Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
mattjung
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
28 ms
·
241.
▲
by
mattjung
18y ago
What about an outsourcing company like http://www.b2kcorp.com ?
242.
▲
by
mattjung
18y ago
He was obviously looking for the right people for the right OS and programming languages at that time for highly scalable communication systems. Would be interesting to compare that with job offers of competitors of that time (who were thos
243.
▲
by
mattjung
18y ago
The sentence I liked most: "Just knowing what to search for is now as difficult as the search itself."
244.
▲
Nearly Infinite
(sethgodin.typepad.com)
3 points
by
mattjung
18y ago
|
1 comments
245.
▲
by
mattjung
18y ago
How many iterations took it from the initial idea to the current product? How did you get your first customers?
246.
▲
by
mattjung
19y ago
May design patterns may help you somehow: http://cssdesignpatterns.com
247.
▲
Hundreds of Thousands of Microsoft Web Servers Hacked
(blog.washingtonpost.com)
5 points
by
mattjung
19y ago
|
4 comments
248.
▲
Yahoo! Web Services in Ruby
(rubyflow.com)
1 points
by
mattjung
19y ago
|
0 comments
249.
▲
by
mattjung
19y ago
The separation is generally not so clear as the article suggests: take Google search as an example. A powerful algorithm is used to rank the search results by relevance. But relevance is to a large degree determined how often pages are refe
250.
▲
by
mattjung
19y ago
It's more than a lottery. It is also a great adventure of which you can learn a lot. We need those visionary "crazy" people with big ideas that change the world.
251.
▲
by
mattjung
19y ago
The 37Signals model is a valid model and it is honorable that they evangelize it. I would say, it is the "30-something-model" for people you can't affort to deal with high risks and prefer to slidely glitch into professional independence.
252.
▲
The power of links and the value of global knowledge
(paulbuchheit.blogspot.com)
25 points
by
mattjung
19y ago
|
6 comments
253.
▲
by
mattjung
19y ago
A large number of Mac users are also professional designers or music producers and not part of that typical Internet user. So Mac users may pay for Mac software, but not for web-services. My point is: the consumer market offer less niche ma
254.
▲
by
mattjung
19y ago
You are right. I used the term "enterprise software" only as a contrast to "consumer software". The "Fortune 5,000,000" is probably the best target market for a startup. Why? Targeting large enterprises (what the company I work for does) is
255.
▲
by
mattjung
19y ago
One characteristic is that they all provide enterprise software. The end-consumer market is very difficult since internet-users are used to get everything for free.
256.
▲
by
mattjung
19y ago
One reason more to create a startup that sweeps away the colossus ;-)
257.
▲
by
mattjung
19y ago
The title of the blog-post is "Aren’t There Real Problems To Solve?", which is something completely differnet than "Aren't There Any Real Problems Left To Solve?". The latter sounds like a rhethorical question that suggests that progress ha
258.
▲
by
mattjung
19y ago
Usage, technology, resources, bottlenecks - all have been changing all along the time and further will change. The ideal solution of today will not be a good solution for tomorrow. What you call mess, is the result of an evolutionary develo
259.
▲
by
mattjung
19y ago
You are right, i was unclear here. I didn't mean "urgency is poisonous" is counter-intuitive, but rather the advice in the blog-posting "we work less (4 days a week) to be more productive". The issue with urgency is that you often have no c
260.
▲
by
mattjung
19y ago
More than just a counter-intuitive thesis to raise discussions?
261.
▲
Urgency is poisonous?
(37signals.com)
29 points
by
mattjung
19y ago
|
22 comments
262.
▲
by
mattjung
19y ago
Two little stories: A friend of mine has been trying since 10 years to transform his consulting company into a product company - without success. The projects he has implemented were too expensive in terms of development and maintenance to
263.
▲
by
mattjung
19y ago
... lack of sense for a good cost-benefit ratio? Given that test code also needs to be maintained, blowing up the whole code base will make maintenance probably more expensive that needed.
264.
▲
by
mattjung
19y ago
I miss Facebook in the article!
265.
▲
2011 Tipping Point For Web Apps In The Enterprise (Read Write Web)
(readwriteweb.com)
1 points
by
mattjung
19y ago
|
0 comments
266.
▲
by
mattjung
19y ago
The number for inactive blogs (or inactive users in web-applications) is surely much higher because its even easier to create a blog. People play around with something new, find something better, etc.
267.
▲
by
mattjung
19y ago
Blocking ads is a more and more widely required feature by users that advertisers as well as product companies will have to deal with. Demonizing doesn't seem to be the best strategy...
268.
▲
by
mattjung
19y ago
What about a simple traffic light feedback: green for the accepted ones, yellow for the pretty good rejected ones, red for the rest. At least it would help to see if someone is on the right track...
269.
▲
by
mattjung
19y ago
Since it is much more expensive and difficult to create traction on a product and to win customers than to port an existing product on another platform or technology, it should not be a show stopper for a large aquirer. Things may be diffe
270.
▲
Hacker's paradise? Google Offices Zurich
(swissmiss.typepad.com)
2 points
by
mattjung
19y ago
|
0 comments
More ›