Sterling Silver: A number of project experiences for start-up companies to compress process costs

Intermediary transaction http://www.aliyun.com/zixun/aggregation/6858.html ">seo diagnose Taobao guest cloud host technology Hall

I have mentioned that the Cicada team is more effective in compressing the process costs. This part can be talked about, perhaps to others a little help.

About the project

So far, although we have only made a cicada travels, in fact, also made another four app design, just do not have time to research and development, first put.

The process of the project is this, usually by me first to mention an idea, to talk to you, and if you do not encounter strong objections, and a few friends and relatives to chat. When I feel a little bit of a bottom, I look at the same products and axure prototypes-this will not take more than two days. But it is also possible to find the idea in the process is not reliable, then give up.

Low-income real prototype after painting, Cicada team will squeeze into a lump, listen to me to explain, to make suggestions. If you do not encounter a strong objection, ask the UI designer to take out the PSD, usually only the main page, minor small pages are not tube.

Then I copy the visual copy to the phone, I met acquaintances to take out to show him, listen to outsiders how to evaluate.

For new projects, I would prefer to "put some time" after the design, rather than immediately on the horse. Of course, we don't have time to go into research and development right away. Putting one aside will make the idea mature and get more feedback from outsiders to improve the design. An app may have changed its prototype several times before it started coding, correcting a lot of details.

Cicada travels the same way of handling, often put in six months after the module design, put six months, repeated changes. This will offset the recklessness of some of the quick designs. To a set of molded things, there will be more careful thinking. Ideas need to be quickly converted into persuasive prototypes, otherwise it's not up to the idea of a head shot to be discussed or done.

About a complete version process

A normal app version of the Cicada's travels, usually in a 3-5-week period.

The small function point in the version plan is determined by me directly, the big function point wants to consult the engineer's opinion in advance, go through the prototype review. Say is the review, in fact, we crowded into a lump, listen to me to explain the prototype. Finish speaking no objection, I then tower the visual task in the schedule, with the UI designer to confirm the time, the designer in black and white on the schedule out of the PSD (Tower true collaboration artifact also).

I'll have all the PSD ready before the new version is started, and the version plan will be labeled with a blue label on the tower. In advance with the engineer agreed to a few weeks, each research and development finished a function point, in the tower hook off, every day to see tower progress, morning will not open.

If the 4-week version of the iteration to calculate, the first 2 weeks I was preparing for the next version, starting from the 3rd week, I stepped in the test, to the tower of the blue tag off the check, and then the debugging requirements with a red label record. The 3rd week will complete the function development, reserve 1 weeks half time debugging. The new version of the demand all fixed, I spent the whole day full regression test round, the Cicada team then the whole test round, after the submission. After submitting the engineer to come to me, when the blue label has been sorted out, to tower listen to me about the next version of the plan.

The use of tower.im can make the project orderly, clear and transparent to advance.

About documents

I have a point of view, large companies in the complete product documentation in fact, few people see, mainly for the wrangling. "I wrote very clearly in the document that you did not achieve it!" Your requirements and documentation are inconsistent and the duration must be postponed! "

Cicada team set up in the early days, but also with Doc to record demand, later cooked, tacit understanding, to greatly simplify the process. The UI designer axure the PSD with the low fidelity, and the engineer encodes the PSD. Interactive effect mainly by oral communication, simple functional algorithm also verbal communication, complex as a function point, in the tower on the list of detailed remarks.

Of course, this is a particularly tacit approach.

Still in the running-in period, I will be a more complete sketch prototype, interactive effect, functional algorithm, design ideas are labeled above, to facilitate the engineer to establish the overall impression of the product. But I am too lazy to do the prototype dynamic effect, the annotation is full of text, like a yellow sticky paper as pasted next to the sketch, accompanied by oral explanation. Later, the engineer to the product design to know, to throw off the prototype directly to see more intuitive PSD, corresponding to the function point on the tower, a look at it.

The disadvantage of doing this is that the lack of complete product records, once forgotten what immediately dumbfounded. Wait a minute ... I also have a particularly detailed test case where app lists nearly 500 finely tested points. Although it is written with MindManager, the readability is particularly poor, forcing only I can read, but covers the vast majority of functional points. When the product enters the stable period, after the normal period, it is easy to sort out the specification product document against the test case. Prior to that, it was much easier to revise directly on test cases because of changing requirements.

About tower.im

Writing to the back, I found that this article basically turned into a tower ad post. Yes, I am Tower's brain residue powder. Some say it replicates who, and then I have to endure the very slow speed with the English interface to use the original, to highlight my moral superiority? Stop it.

There are several advantages to using tower. The first is a strong sense of reason, the task was broken out, corresponding to the head, the corresponding date, very clear.

Second, the flexibility is particularly good, with the sub-item, grouping and labeling, you can record version requirements, debugging opinions, difficult problems, visual scheduling, business plan, no matter what can be loaded.

Third, the progress is particularly transparent, who assigned the number of tasks, the degree of completion, everyone can see.

After adjusting to tower for a while, everyone brushes tower every day, knowing what they do and what others do. Can accept the task that I arrange, also can give oneself under the task. Because all the work arrangements are tiled above, there is a clear cut-off time, it is easy to make a reasonable time plan.

But adapting to tower has a premise that you have to be a very organized person. Tower doesn't change the way you work, it makes your original work clearer, more transparent, and more efficient. If you rarely 1, 2, 3 to their own task, to others under the task, rarely break down the task into fine particles, but also rarely strict time planning, it is not to play tower. And I don't think that kind of person has the will and ability to improve efficiency.

Finally, the color design of old Shen will be 5 cents into my alipay account.